JP4321261B2 - Engine starter - Google Patents

Description

  The present invention relates to an engine starting device that automatically starts an engine that has been automatically stopped when idling or the like in response to a restart request, and more particularly, correction of the amount of fuel supplied to be injected into a predetermined cylinder during the restart. It belongs to the technical field of control.

2. Description of the Related Art Conventionally, an engine control system (idle stop system) is known in which an engine is automatically stopped during idling for the purpose of reducing fuel consumption and suppressing CO ₂ emission. In such a system, the engine must be started immediately in response to an engine restart request such as a start operation. However, in a general starting method in which the engine is started through cranking by the starting motor, the starting time is short. There is a tendency that it becomes slightly longer, and there is also a problem that vibration and noise accompanying cranking make the driver feel uncomfortable.

  In addition, if the engine is stopped and restarted every time the engine is in an idle state, the number of times of starting is significantly increased compared to a normal system that starts only when the ignition switch is operated. The starter motor is required to have extremely high durability, and there is a problem that unnecessary cost increases.

Therefore, in recent years, for example, as in a direct injection engine disclosed in Patent Literatures 1 and 2, fuel is injected and supplied into a cylinder that is in an expansion stroke in a stopped state, and then a starting motor is ignited and burned. The one that started the engine with its own power without borrowing the power of has been developed.
Japanese Utility Model Publication No. 60-128975 Special table 2003-517134 gazette

  By the way, in the direct injection engine as in the conventional example (Patent Documents 1 and 2), the fuel is directly injected into the high-pressure cylinder, so the pressure of the fuel supplied to the fuel injection valve for each cylinder is considerably high. Therefore, after stopping the fuel supply to each cylinder in order to stop the engine, a small amount of fuel leaks from the nozzle hole of the fuel injection valve facing the cylinder. This fuel is vaporized in the high-temperature cylinder to form an air-fuel mixture, but when the engine is subsequently started in response to the ignition operation, it is discharged from the cylinder along with cranking and harmless by the catalytic converter in the exhaust passage. Will be converted.

  However, when the engine is started on its own without relying on the ignition operation, such as when restarting after the automatic stop, cranking is not performed, so that the fuel mixture leaked into the cylinder remains as described above. In addition to this, if fuel for starting is injected and supplied, the air-fuel ratio in the cylinder becomes excessively rich, and the combustion state may deteriorate and start-up may fail. is there.

  Therefore, it is conceivable to correct the amount of fuel injected into the cylinder at the time of restart by the amount of fuel remaining in the cylinder. However, the amount of fuel leaking into each cylinder varies depending on individual variations of the fuel injection valve. Therefore, in order to appropriately correct the fuel injection amount, it is necessary to detect the fuel leakage amount for each cylinder.

  The present invention has been made in view of such a point, and an object of the present invention is to restart the engine from the stop in the engine starter which automatically stops the engine and then restarts automatically. In the meantime, a method for obtaining the amount of fuel leaking from the fuel injection valve into the cylinder is devised so that the fuel supply amount at restart can be appropriately corrected with a simple configuration.

  In order to achieve the above object, in the present invention, after the fuel supply to each cylinder is stopped and the engine is stopped, the air-fuel ratio in the cylinder is gradually increased by the fuel leaking little by little from the injection hole of the fuel injection valve. Focusing on the approach to the state where ignition is possible, a predetermined cylinder after being stopped is repeatedly ignited at very short time intervals, so that the time change of the fuel leakage amount in the cylinder from the time until ignition occurs is determined. Estimated.

  Specifically, the invention of claim 1 includes an automatic stop means for automatically stopping the engine by stopping fuel supply to each cylinder when a predetermined stop condition is satisfied during operation of the multi-cylinder engine, When a predetermined restart condition is satisfied after the automatic stop, fuel is injected and supplied into at least the cylinder stopped in the expansion stroke, and ignited and burned to automatically restart without using a starter motor. And an automatic engine starting device.

  An operation means for performing an operation for manually stopping the engine, a manual stop means for stopping the engine by stopping fuel supply to each cylinder when the operation means is operated, and the engine When the engine is stopped by the manual stop means, the post-stop ignition means for repeatedly igniting the predetermined cylinder, the determination means for determining that the air-fuel mixture in the predetermined cylinder has been ignited by the ignition, and the determination from the stop of the engine Learning means for measuring a time until ignition is determined by the means, and estimating and storing a time change characteristic of the fuel leakage amount in the cylinder based on at least the measurement time.

  With the above-described configuration, when the operation unit is operated during the operation of the engine, the fuel supply to each cylinder is stopped by the manual stop unit in response to this, and when the engine stops, the post-stop ignition unit repeats the predetermined cylinder. Ignition is performed. At this time, in the cylinder, the fuel leaking little by little from the nozzle hole of the fuel injection valve is vaporized to form an air-fuel mixture, and the air-fuel ratio gradually changes to the rich side as time passes. When the air-fuel ratio exceeds the lean ignition limit, the air-fuel mixture is ignited, and this is determined by the determining means.

  Then, based on at least the time from the stop of the engine to the determination of the ignition, the characteristic of the time variation of the fuel leakage amount in the cylinder is estimated and stored by the learning means. In other words, since it is considered that the air-fuel ratio in the cylinder corresponds to the lean ignition limit of the air-fuel mixture at the time of the ignition discrimination, the amount of remaining fuel is calculated from the cylinder volume at this time. Based on the amount of the fuel and the elapsed time from the engine stop, it is possible to estimate the characteristics of the time variation of the fuel leakage amount.

Specifically, for example, a detection unit that detects the amount of air in the cylinder that is ignited by the post-stop ignition unit after the engine is stopped is provided in advance, and the learning unit associates in advance with the lean ignition limit of the air-fuel mixture. Based on the set air-fuel ratio and the detected air amount, a fuel leakage amount at the time of ignition determination by the determination means is calculated, and based on this fuel leakage amount and the measurement time until the ignition determination, the fuel leakage amount is calculated. What is necessary is just to estimate the time change characteristic of the leakage amount (the invention of claim 6).
In this way, if the characteristics of the time variation of the fuel leakage amount are learned for each cylinder, even when the engine is automatically restarted after the automatic stop, the fuel leaked into the cylinder during the stop Therefore, the amount of fuel newly injected and supplied into the cylinder for starting may be corrected accordingly.

  That is, for example, when the automatic start means is to inject and supply fuel into a cylinder stopped in the expansion stroke and ignite, the time from the engine stop to the fuel supply by the automatic start means, And correction means for correcting the fuel supply amount to the expansion stroke cylinder by the automatic start means based on the time variation characteristic of the fuel leakage amount of the cylinder stored by the learning means. Invention of 2). In this way, according to the amount of fuel remaining in the cylinder that is in the expansion stroke at the time of restart, the amount of fuel newly injected and supplied into this cylinder is appropriately corrected so that the air-fuel ratio becomes excessively rich, etc. The engine startability can be stably ensured by good combustion.

  Alternatively, the automatic starting means injects fuel into the cylinder stopped in the compression stroke and ignites it, thereby causing the engine to operate once in reverse and in the expansion stroke cylinder compressed accordingly. In the case where the engine is started by injecting and supplying fuel and igniting after reverse operation, the time from the stop of the engine to the fuel supply to each cylinder by the automatic starting means, and learning means And a correction means for correcting the amount of fuel supplied to each cylinder by the automatic start means based on the time variation characteristic of the fuel leakage amount for each cylinder stored in (5). ).

  In this case, when the engine is restarted, the piston of the expansion stroke cylinder moves to the top dead center side by the first reverse rotation operation, so that the effective stroke of this cylinder becomes long, and the density and pressure in the cylinder are increased. Combustion is performed at a high state. Thereby, a large combustion pressure can be obtained, and the startability is improved by increasing the start torque.

  In both the compression stroke cylinder and the expansion stroke cylinder in which combustion is performed in this manner, the air-fuel ratio fluctuation caused by the fuel leakage into the cylinder is suppressed as in the invention of claim 2, and good combustion is achieved. As a result, the engine startability improved as described above can be stably secured.

  In order to ignite a relatively lean air-fuel mixture formed by fuel leaking into the cylinder as described above and determine the ignition, for example, a rise in the cylinder pressure due to the ignition may be detected by a pressure sensor. Alternatively, the movement of the piston due to this may be detected by a crank angle sensor. For this purpose, since the inside of the cylinder must be sealed, the post-stop ignition means preferably ignites a cylinder in which both the intake valve and the exhaust valve are closed (invention of claim 4). In other words, the cylinder stopped in either the compression stroke or the expansion stroke may be ignited (invention of claim 5).

  Here, it is not only when the engine is manually stopped that the amount of fuel leakage can be estimated from the elapsed time until ignition by repeatedly igniting the predetermined cylinder after the engine is stopped as described above. It may be at the time of automatic stop such as idle stop. However, if the air-fuel mixture is ignited in the expansion stroke cylinder of the engine after an automatic stop such as an idle stop, the air in the cylinder is consumed and burned gas is generated, and then the engine is automatically started. Even so, it is difficult to perform good combustion in the expansion stroke cylinder, and the startability may be impaired.

  In view of this point, the invention of claim 7 is directed to the engine starting device having the same premise structure as that of the invention of claim 1, and once in a preset number of times when the engine is stopped by the automatic stop means. The post-stop ignition means for repeatedly igniting the predetermined cylinder at a ratio of the above, the determination means for determining that the air-fuel mixture in the predetermined cylinder has been ignited by the ignition, and until the ignition is determined by the determination means from the stop of the engine Learning means for measuring time and estimating and storing a time change characteristic of the amount of fuel leakage in the cylinder based on at least the measurement time. As in the first aspect of the invention, it is possible to learn the fuel leakage amount of each cylinder, and after that, after the ignition is determined by the determination means, the engine When the start condition is satisfied, and to reliably start the engine using the starter motor by the automatic start means.

  This makes it possible to learn the amount of fuel leakage not only when the engine is manually stopped, but also at the time of automatic start, which is much more frequent than that. Variations and changes over time can be learned early and reflected in fuel injection control at the time of automatic start.

  As described above, according to the engine starting device according to the first to fifth aspects of the present invention, the engine is automatically stopped at the time of idling or the like, and then the engine is automatically and restarted automatically. In the apparatus, when the engine is stopped in response to manual operation, a predetermined cylinder is repeatedly ignited, so that the time variation characteristic of the fuel leakage amount in the cylinder is learned from the time until ignition occurs. Based on the learning result, the fuel supply amount at the time of automatic start is appropriately corrected, the fluctuation of the air-fuel ratio in the cylinder due to fuel leakage is suppressed, and stable startability of the engine can be ensured by good combustion.

According to the invention of claim 6, since the learning of the fuel leakage amount can be performed not only at the time of manual stop of the engine but also at the time of automatic stop, the effect of learning can be obtained at an early stage. Moreover, when the engine is automatically started after learning, a reliable start can be performed using the start motor.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. It should be noted that the following description of the preferred embodiment is merely illustrative in nature and is not intended to limit the present invention, its application, or its use.

-Schematic configuration of engine control system-
1 and 2 show an embodiment of an engine control system including an engine starter according to the present invention. The engine system E controls an engine 1 including a cylinder head 10 and a cylinder block 11, and controls the engine 1. ECU2 (engine controller) for carrying out. As shown in FIG. 2, the engine 1 is provided with four cylinders 12A to 12D. Inside each of the cylinders 12A to 12D, pistons connected to the crankshaft 3 as shown in FIG. Thus, a combustion chamber 14 is formed above the piston 13 in each of the cylinders 12A to 12D.

  Here, in general, in a multi-cylinder four-cycle engine, each cylinder performs a combustion cycle composed of intake, compression, expansion, and exhaust strokes with a predetermined phase difference. In the case of a cylinder engine, when the first cylinder 12A, the second cylinder 12B, the third cylinder 12C, and the fourth cylinder 12D are called from one end in the cylinder row direction, as shown in FIG. 6 (e), the first cylinder (# 1 ) Combustion is performed with a phase difference of 180 degrees in crank angle in the order of the third cylinder (# 3), the fourth cylinder (# 4), and the second cylinder (# 2).

  A spark plug 15 for igniting and burning the air-fuel mixture in the combustion chamber 14 is provided at the top of each combustion chamber 14 of each of the cylinders 12A to 12D. An electrode is disposed so as to face the combustion chamber 14. Further, a fuel injection valve 16 is disposed on the side of the combustion chamber 14 (right direction in FIG. 1) with the tip injection hole facing the combustion chamber 14. This fuel injection valve 16 incorporates a needle valve and a solenoid (not shown) and is driven to open for a time corresponding to the pulse width in response to the input of a pulse signal from the ECU 2 so that an amount of fuel corresponding to the driving time is supplied. The cylinders 12A to 12D are configured to inject directly. The fuel injection direction is adjusted so as to be directed to the vicinity of the electrode of the spark plug 15.

  Although not shown, fuel is supplied to the fuel injection valve 16 via a fuel supply passage or the like by a fuel pump, and the fuel supply pressure is in the middle of the compression stroke of each cylinder 12A to 12D. Thereafter, the pressure is set higher than the pressure in the combustion chamber 14 so that fuel can be injected into the high-pressure in-cylinder combustion chamber 14.

  An intake port 17 and an exhaust port 18 that open toward the combustion chamber 14 are provided in the upper part of the combustion chamber 14 of each of the cylinders 12A to 12D. An intake valve 19 and an exhaust valve are provided in these ports 17 and 18, respectively. 20 are arranged respectively. These intake valve 19 and exhaust valve 20 are driven by a valve operating mechanism including a camshaft (not shown), and as described above, each of the cylinders 12A to 12D performs a combustion cycle with a predetermined phase difference. The opening / closing timing of the intake / exhaust valves 19 and 20 for each cylinder is set.

  Further, an intake passage 21 and an exhaust passage 22 are provided so as to communicate with the intake port 17 and the exhaust port 18, respectively. As shown in FIG. An independent branch intake passage 21a is provided for each of the cylinders 12A to 12D, and the upstream end of each branch intake passage 21a communicates with the surge tank 21b. The intake passage 21 upstream of the surge tank 21b is a common intake passage 21c common to the cylinders 12A to 12D. The throttle valve 23 adjusts the cross-sectional area of the passage 21c by, for example, a butterfly valve to restrict the intake flow. 2 and an actuator 24 for driving the air flow sensor. Further, as shown only in FIG. 2, on the upstream side and the downstream side of the throttle valve 23, an air flow sensor 25 and an intake pressure for detecting the intake air amount, respectively. An intake pressure sensor 26 for detecting (negative pressure) is provided.

  On the other hand, a catalyst 29 for purifying the exhaust is disposed downstream of the collection portion of the exhaust passage 22 where exhaust from the cylinders 12A to 12D collects. This catalyst 29 is, for example, a so-called three-way catalyst in which the purification rate of HC, CO, and NOx is extremely high when the air-fuel ratio state of the exhaust is in the vicinity of the theoretical air-fuel ratio, and this is a relatively high oxygen concentration in the exhaust. It has an oxygen storage capacity for storing it in a high oxygen excess atmosphere. When the oxygen concentration is relatively low, the stored oxygen is released and reacted with HC, CO and the like. The catalyst 29 is not limited to a three-way catalyst, and may be any catalyst having an oxygen storage capacity as described above. For example, the catalyst 29 may be a so-called lean NOx catalyst that can purify NOx even in an oxygen-excess atmosphere.

  The engine 1 is provided with an alternator 28 that is drivingly connected to the crankshaft 3 by a belt or the like. Although not shown in detail, the alternator 28 includes a regulator circuit 28a that changes the output voltage by controlling the current of the field coil and thereby adjusts the amount of power generation, and the regulator circuit 28a includes the ECU 2. By inputting a control command (for example, voltage) from the vehicle, the amount of power generation is basically controlled in accordance with the electric load of the vehicle electrical component and the in-vehicle battery voltage. When the power generation amount of the alternator 28 is changed in this way, the driving force, that is, the magnitude of the external load of the engine 1 changes accordingly.

  Further, the engine system E is provided with two crank angle sensors 30 and 31 for detecting the rotation angle of the crankshaft 3, and the engine rotation speed is mainly based on a signal from one crank angle sensor 30. As will be described in detail later, the rotation direction and the rotation angle of the crankshaft 3 are detected by the crank angle signals output from the two crank angle sensors 30 and 31 that are out of phase with each other. Yes. In addition, the engine system E includes a cam angle sensor 32 that detects a specific rotational position of the camshaft and outputs it as a cylinder identification signal, and an operation unit that performs an on / off operation for manually switching between operation and stop of the engine 1. An ignition switch 33 (IG switch), an accelerator opening sensor 34 for detecting an accelerator operation amount, and the like are provided.

  The ECU 2 receives signals from the sensors and switches 25, 26, 30 to 34, outputs a signal for controlling the fuel injection amount and the injection timing to the fuel injection valve 16, and controls the ignition plug 15. A signal for controlling the ignition timing is output to the ignition device 27, and a signal for controlling the throttle opening is output to the actuator 24 of the throttle valve 23. As will be described in detail below, the ECU 2 stops the fuel supply to each cylinder 12A to 12D (fuel cut) and automatically stops the engine when a predetermined engine stop condition is satisfied during idling. After that, when a predetermined engine restart condition is satisfied by the driver's accelerator operation or the like, the engine 1 is automatically restarted.

  That is, when the engine 1 is automatically restarted, the engine 1 is started by itself without borrowing the power of the starting motor. In this embodiment, as schematically shown in FIG. The first combustion is performed in the cylinder 12 in which the piston 13 is stopped in the middle of the compression stroke (# 1 cylinder 12A in the example in the figure, hereinafter also referred to as the compression stroke cylinder when stopped), and the piston 13 is pushed down. Thus, the crankshaft 3 is slightly rotated reversely ((a) in the figure), so that the cylinder 12 in the expansion stroke (# 2 cylinder 12B in the example in the figure, hereinafter also referred to as the stop expansion stroke cylinder). The piston 13 is raised to compress the air-fuel mixture in the cylinder 12B ((b) in the figure). Then, by igniting and burning the air-fuel mixture in the expansion stroke cylinder 12B that has been compressed and thus increased in temperature and pressure, a torque in the forward rotation direction is given to the crankshaft 3, and the engine 1 is I try to start it.

  In order to start the engine 1 by itself, the torque in the forward rotation direction as large as possible is applied to the crankshaft 3 by the combustion of the expansion stroke cylinder 12B at the time of stop, and as shown in FIG. Subsequently, the cylinder 12A that reaches compression top dead center (hereinafter abbreviated as TDC) must overcome the compression reaction force (compression pressure) and exceed TDC. Therefore, in order to start the engine 1 reliably, it is necessary to ensure sufficient air for combustion in the stop-time expansion stroke cylinder 12B.

  Therefore, in this embodiment, when the engine 1 is automatically stopped during idling, first, the fuel is cut at a predetermined rotational speed slightly higher than the idling rotational speed so that the scavenging of each cylinder 12A to 12D is sufficiently performed. At the same time, the throttle valve 23 is opened for a predetermined period of time to control the opening degree to be set in advance. Then, by closing the throttle valve 23 at an appropriate preset timing, the amount of air sucked into the stop expansion stroke cylinder 12B and the stop compression stroke cylinder 12A is sufficiently increased, and the expansion stroke cylinder The air amount of 12B is set to be slightly larger than that of the compression stroke cylinder 12A.

  By doing so, the piston 13 of the expansion stroke cylinder 12B is somewhat suitable for restarting near the bottom dead center (BDC) from the center of the stroke due to the balance of the compression pressure of the air in the two cylinders 12A and 12B. It stops within a range R (described later).

-Automatic engine stop control-
Next, control for automatically stopping the engine 1 by the ECU 2 will be described mainly with reference to FIGS. 4 and 5 are flowcharts showing the procedure of the stop control, and FIG. 6 shows the engine rotation during the period from the fuel cut to the inertial rotation of the engine 1 (hereinafter also referred to as a stop operation period). Changes in speed, crank angle, and stroke of each cylinder 12A to 12D are shown in association with each other, and the control of the throttle opening performed during that time and the change in intake pressure (intake pipe negative pressure) thereby are schematically shown. It is explanatory drawing shown.

  FIG. 7 is a diagram showing a correlation between the TDC rotational speed (described later) of the engine 1 whose rotation gradually decreases during the stop operation period, and the piston stop position in the expansion stroke cylinder 12 after the stop.

  First, as shown in FIG. 6 (a), when a fuel cut is performed at a predetermined set rotational speed (800 rpm in the illustrated example) during operation of the engine 1 (time t0), the crankshaft 3 and fly The kinetic energy of a moving part such as a wheel is consumed by mechanical friction and pumping work of each cylinder 12A to 12D, so that the engine rotation speed gradually decreases, and the engine 1 stops after several revolutions due to inertia. It will be. In detail, while the engine 1 rotates by inertia, the engine rotation speed temporarily decreases greatly every time the compression top dead center (TDC) of each of the cylinders 12A to 12D is reached. When it exceeds, it rises again, and it goes down while repeating up and down. For example, when the fuel is cut at about 800 rpm as shown in the figure, the TDC is usually exceeded 8 or 9 times, and after the last TDC (time t3), the next TDC cannot be exceeded. To stop (time t4 to t6).

  That is, as described above, in the cylinder 12 (# 1 cylinder 12A in the figure) that cannot exceed TDC and remains in the compression stroke, the air pressure increases as the piston 13 rises due to inertial force, and the compression reaction force causes the piston 13 to move. Once stopped (time t4), it is pushed back toward the BDC. As a result, the crankshaft 3 reverses and the engine speed becomes a negative value as shown in FIG. 5A. Then, this time, the cylinder 12 in the expansion stroke (transition to the expansion stroke beyond the last TDC) is performed. In this example, the air pressure of the # 2 cylinder 12B) rises, and a compression reaction force to the BDC side acts on the piston 13, and the piston 13 of the expansion stroke cylinder 12 is temporarily caused by this compression reaction force. After stopping (time t5), it is pushed back toward the BDC. Thus, the crankshaft 3 rotates forward again, and the engine speed returns to a positive value.

  As described above, the pistons 13 of the cylinders 12 </ b> A to 12 </ b> D are reciprocated several times and stopped after the compression reaction forces acting in the opposite directions on the pistons 13 of the compression stroke cylinder 12 and the expansion stroke cylinder 12. (Time t6), the stop position is roughly determined by the balance between the compression reaction forces of the compression and expansion stroke cylinders 12, and finally exceeds TDC before the stop due to the influence of the friction of the engine 1 and the like. It changes according to the rotational inertia of the engine 1 at the time, that is, the level of the engine speed when the TDC is finally exceeded.

  Therefore, in order to stop the piston 13 of the cylinder 12 in the expansion stroke when the engine is stopped within the predetermined range R suitable for restarting, first, the compression reaction of the expansion stroke cylinder 12 at the stop time and the compression stroke cylinder 12 at the stop time is first. It is necessary to adjust the amount of intake air to both cylinders 12 so that the force is sufficiently large and the compression reaction force of the expansion stroke cylinder 12 is in a suitable balance larger than the compression stroke cylinder 12 by a predetermined amount or more. There is. Therefore, in this embodiment, as shown in FIG. 6 (c), the throttle valve 23 (time t1) opened immediately after the fuel cut is closed after the lapse of a predetermined period (time t2). As shown, the intake pipe negative pressure is temporarily reduced (intake amount is increased), so that the required amount of air is sucked into the compression and expansion stroke cylinders 12 when stopped.

  However, the actual engine 1 has individual variations in the shape of the throttle valve 23 itself, the intake port 17, the branch intake passage 21a, and the like, and the behavior of the air flow that circulates there may change. Since the amount of air flowing into each of the cylinders 12A to 12D varies to some extent during the period, even if the opening / closing control of the throttle valve 23 as described above is performed, it is only that when the engine is stopped, the compression stroke or the expansion stroke It is difficult to accurately set the piston stop position of the cylinder 12 to be within the target range R.

  In this regard, in this embodiment, as shown in an example in FIG. 7, the engine rotation speed when each of the cylinders 12 </ b> A to 12 </ b> D sequentially passes TDC in the process of gradually decreasing the engine rotation speed during the stop operation period. Focusing on the fact that there is a clear correlation between (hereinafter also referred to as TDC rotational speed) and the piston stop position of the cylinder 12 in the expansion stroke after the engine is stopped, as shown in FIG. By detecting the TDC rotational speed at every 180 ° CA in the process of decreasing the engine rotational speed, and controlling the power generation amount of the alternator 28 and the opening degree of the throttle valve 23 according to the detected value, the degree of decrease in the engine rotational speed is detected. To adjust.

  Specifically, FIG. 7 shows that the engine 1 that rotates by inertia is cut off when the engine speed is approximately 800 rpm as described above, and the throttle valve 23 is kept open for a predetermined period thereafter. Each time the cylinders 12A to 12D exceed TDC, the engine rotational speed (TDC rotational speed) at that time is measured, and the piston position of the expansion stroke cylinder 12 after being stopped is examined, and this piston position is determined. The relationship between the two is shown with the vertical axis and the TDC rotational speed on the horizontal axis. By repeating such an operation a predetermined number of times, a distribution diagram showing a correlation between the TDC rotational speed during the engine stop operation period and the piston stop position in the expansion stroke cylinder 12 after the stop is obtained.

  In the example in the figure, the rotation speed when the last TDC before the engine stop is not shown, but the last one from the TDC rotation speed immediately after the fuel cut (the ninth in the example counted from the end) is not shown. Data up to the previous TDC rotational speed (second one from the end) is shown. The 9th to 2nd TDC rotation speeds from the last are distributed in a lump, and as is apparent in the 6th to 2nd ones shown in the figure, the TDC rotation speed is within a certain range (see FIG. If it is within a range indicated by hatching, it can be seen that the piston stop position falls within a range R suitable for restart (ATDC 100 to 120 ° CA in the example in the figure).

  As described above, the specific range of the TDC rotational speed at which the piston 13 of the expansion stroke cylinder 12 stops within the predetermined range R suitable for restarting the engine 1 is hereinafter referred to as an appropriate rotational speed range in this specification. And In this embodiment, as will be described in detail below, when the engine speed decreases while repeating up and down as shown in FIG. 6 (a), the TDC rotation speed for each cylinder 12A to 12D is detected. Then, the detected value is compared with the appropriate rotational speed range, and the power generation amount of the alternator 28 and the opening degree of the throttle valve 23 are controlled according to the speed deviation between them.

  That is, first, during a predetermined period immediately after the fuel cut, as described above, the throttle valve 23 is opened relatively large for scavenging of each of the cylinders 12A to 12D. Even if the throttle opening is further adjusted, the cylinder Since the pump work of 12 does not change so much, it is difficult to adjust the engine speed. Therefore, during this period, the alternator 28 is intentionally operated to generate power, and the amount of power generation is controlled by changing the magnitude of the power generation driving force for that purpose, thereby adjusting the degree of decrease in engine rotation speed. At this time, the power generation amount of the alternator 28 is controlled to be large so that the TDC rotation speed is close to the lower limit of the appropriate rotation speed range, that is, the engine rotation is slightly lowered.

  Further, after the predetermined period has elapsed, the degree of decrease in the engine rotation speed is adjusted by adjusting the pump work amount of the engine 1 by controlling the opening degree of the throttle valve 23. However, in the case of the throttle valve 23 arranged upstream of the surge tank 21b, the change in the intake air amount of each of the cylinders 12A to 12D is slow even if the throttle valve 23 is controlled to close, so that the engine is previously controlled by the alternator 28 as described above. Only when the TDC rotational speed is lower than the appropriate rotational speed range, the throttle opening is controlled to the open side so that the decrease in the engine rotational speed is moderated. ing.

  As described above, the degree of decrease in the engine rotation speed is adjusted by the power generation control of the alternator 28 and the control of the opening degree of the throttle valve 23, and the TDC rotation speed is within the proper rotation speed range before passing the last TDC at the latest. If so, the kinetic energy of the moving parts such as the crankshaft 3, the flywheel, or the piston 13 and the connecting rod at this time, or the potential energy of the high pressure air of the compression stroke cylinder 12 acts thereafter. This is commensurate with friction and the like, and the piston 13 of the cylinder 12 in the expansion stroke when the engine 1 is stopped can be stopped within a predetermined range R suitable for the restart.

  Next, the specific procedure of the above-described engine automatic stop control will be described with reference to the flowcharts of FIGS. 4 and 5. This flow starts at a predetermined timing during engine operation (START), and in step SA1, idle is performed. Judges whether or not the stop condition (IDL Stop condition) is satisfied. This determination is made based on the vehicle speed, the operating condition of the brake, the engine water temperature, etc. For example, the vehicle speed is smaller than a predetermined speed, the brake is operating, the engine water temperature is within a predetermined range, and the engine 1 is If there is no particular inconvenience for stopping, it is assumed that the idle stop condition is satisfied.

  When the idle stop condition is satisfied in step SA1 of FIG. 4 (in the case of YES), in the subsequent step SA2, any one cylinder 12 (for example, the first cylinder 12A) is specified and the engine is stopped. It is determined whether or not the above condition is satisfied. That is, it is determined whether or not the engine rotation speed is a fuel cut set rotation speed (approximately 800 rpm in this embodiment), whether or not the specified cylinder 12 is in a preset stroke (for example, an intake stroke).

  If all the conditions are satisfied and it is determined YES, the process proceeds to step SA3 to stop the fuel injection to each cylinder 12A to 12D (fuel cut), and in the subsequent step SA4, the throttle valve 23 is set to the set opening degree. Open (throttle open). As a result, as shown in FIGS. 6C and 6D, the amount of intake air to each of the cylinders 12A to 12D is increased, sufficient scavenging is performed, and the catalyst 29 disposed in the exhaust passage 22 has a large amount. As fresh air is supplied, the amount of oxygen stored in the catalyst 29 is sufficiently increased.

  Subsequently, in step SA5, it is determined whether or not the TDC rotational speed obtained from the signal from the crank angle sensor 30 is within an appropriate rotational speed range (is the rotational speed at TDC within a predetermined range?). If this determination is YES and the TDC rotational speed is in the appropriate rotational speed range, the process proceeds to step SA6, where it is determined whether the engine rotational speed is equal to or lower than the predetermined rotational speed. This predetermined rotational speed takes into account the intake transport delay, and as shown in FIGS. 6 (c) and 6 (d), the intake amount to the stop expansion stroke cylinder 12 (# 2 cylinder 12B in the example) is compressed when stopped. This is for closing the throttle valve 23 at a timing higher than that of the stroke cylinder 12 (# 1 cylinder 12A in the figure), and corresponds to the time t2 in the figure, and in this embodiment, for example, about 500 to 600 rpm. Is set in the range. If the engine rotational speed is equal to or lower than the predetermined rotational speed (YES in step SA6), the process proceeds to step SA9 described later. On the other hand, if the engine rotational speed is higher than the predetermined rotational speed (in the case of NO), the process proceeds to step SA5. Return.

  If it is determined in step S5 that the TDC rotational speed is not within the proper rotational speed range (in the case of NO), the process proceeds to step SA7, and based on the rotational speed deviation between the TDC rotational speed and the proper rotational speed range. The power generation amount of the alternator 28 is calculated. This power generation amount is read from, for example, a map set in advance according to the engine rotation speed, the speed deviation from the appropriate rotation speed range, and the current power generation amount. For example, the TDC rotation speed is higher than the upper limit of the appropriate rotation speed range. Sometimes, the power generation amount of the alternator 28 is increased so that the load on the engine 1 is increased, while when the TDC rotational speed is lower than the lower limit of the appropriate rotational speed range, the power generation amount is decreased so that the load on the engine 1 is decreased. is there. In the map, the target value of the power generation amount is set to be large so that the TDC rotational speed is near the lower limit of the appropriate rotational speed range.

  In step SA8 following step SA7, a control command is output to the regulator circuit 28a of the alternator 28 according to the calculation result (alternator power generation). By adjusting the load of the engine 1 by the power generation operation of the alternator 28, the trajectory of the rotational speed of the engine 1 rotating by inertia is shifted to either the high rotation side or the low rotation side, and gradually becomes the target. Approach the trail. If the engine rotational speed becomes equal to or lower than the predetermined rotational speed in step SA6 (YES), the process proceeds to step SA9, the throttle valve 23 is closed (throttle close), and the process proceeds to step SA10 in FIG.

  In step SA10 of the flow of FIG. 5, it is determined whether the TDC rotational speed is in the proper rotational speed range as in step SA5. If the determination is YES and the TDC rotational speed is in the proper rotational speed range, the process proceeds to step SA11. On the other hand, when it is determined that the TDC rotational speed is not within the appropriate rotational speed range (in the case of NO), the process proceeds to step SA12, and the opening degree of the throttle valve 23 is determined based on the deviation of the TDC rotational speed from the appropriate rotational speed range. Is calculated. The throttle opening is read from a map set in advance according to, for example, the engine speed, the speed deviation from the appropriate speed range, and the current opening, and the TDC speed is lower than the lower limit of the proper speed range. Sometimes, the throttle opening is increased so that the pump work of the engine 1 is reduced (TVO in FIG. 6 (c)). On the other hand, when the TDC rotational speed is higher than the upper limit of the appropriate rotational speed range, the throttle control is not performed. Is set to

  In other words, when the throttle valve 23 upstream of the surge tank 21b is used, the response delay toward the throttle side of the intake air becomes large and sufficient controllability cannot be obtained. Is increased, the degree of decrease in engine rotational speed is increased (lower rotational speed), and the throttle valve 23 is driven to open only when the TDC rotational speed is lower than the lower limit of the appropriate rotational speed range. As a result, the degree of decrease in engine speed is moderated. In step SA13, the actuator 24 of the throttle valve 23 is driven (throttle driving), and the process proceeds to step SA11.

  By controlling the alternator 28 and the throttle valve 23 as described above, by adjusting the degree of decrease in the engine speed after the fuel cut, the engine speed gradually decreases while repeating up and down as shown in FIG. The speed trajectory is gradually corrected so that it can be within the proper rotational speed range by the last TDC at the latest.

  In step SA11, it is determined whether or not the detected TDC rotational speed is equal to or less than a predetermined value A. This predetermined value A is set in association with the last TDC rotational speed before the engine stop determined experimentally in advance. If the TDC rotational speed determined in step SA10 is higher than the predetermined value A (determination) When NO is NO), the engine 1 has not yet passed through the last TDC, so the routine returns to step SA10 to continue the control of the throttle valve 23 described above.

  On the other hand, if the TDC rotational speed is equal to or less than the predetermined value A (when the determination is YES), the engine 1 has already passed through the last TDC, and thereafter, as described above, 2 in each of the compression stroke and the expansion stroke. Due to the compression reaction force of the two cylinders 12, 12, the cylinder is rotated several times in the forward direction and the reverse side and then stopped. Therefore, the process proceeds to step SA14, where the stop (complete stop) of the engine 1 is confirmed based on the signals from the crank angle sensors 30, 31, and if the stop of the engine 1 is confirmed as YES, the process proceeds to step SA22. The piston stop position of the cylinder 12 in the expansion stroke is detected based on the crank angle signals output from the two crank angle sensors 30 and 31 and out of phase by a subroutine (see FIGS. 8 and 9) described later. This is stored in the memory of the ECU 2 and the engine stop control is completed (END).

  That is, as described above, the crankshaft 3 rotates several times in both forward and reverse directions immediately before the engine 1 is stopped, so that the piston stop position is detected only by counting the signal from the crank angle sensor 30. I can't. Therefore, in this embodiment, the rotation direction and the rotation angle of the crankshaft 3 are detected as follows based on the crank angle signals output from the two crank angle sensors 30 and 31 and shifted from each other. A crank angle with respect to TDC or BDC of each cylinder 12A to 12D, that is, a piston stop position is detected.

  Specifically, FIG. 8 is a flowchart showing a subroutine for detecting the stop position of the piston. When this flow starts, in step SC1, the first crank angle signal CA1 (the output signal from the first crank angle sensor 30) is shown. ) And the second crank angle signal CA2 (output signal from the second crank angle sensor 31), when the ECU 2 rises the first crank angle signal CA1, the second crank angle signal CA2 is either Low or High. It is also determined whether the second crank angle signal CA2 is High or Low when the first crank angle signal CA1 falls. That is, it is determined whether the phase relationship between these signals CA1 and CA2 is as shown in FIG. 9A or FIG. 9B, and the normal rotation and inversion of the engine 1 are thereby performed. Determine.

  More specifically, during forward rotation of the engine, as shown in FIG. 9A, the second crank angle signal CA2 causes a phase delay of about a half pulse width with respect to the first crank angle signal CA1, and the first The second crank angle signal CA2 becomes Low when the first crank angle signal CA1 rises, and the second crank angle signal CA2 becomes High when the first crank angle signal CA1 falls. On the other hand, at the time of reverse rotation of the engine, as shown in FIG. 9B, the second crank angle signal CA2 has a phase advance of about a half pulse width with respect to the first crank angle signal CA1. Contrary to the forward rotation, the second crank angle signal CA2 becomes High when the first crank angle signal CA1 rises, and the second crank angle signal CA2 becomes Low when the first crank signals CA1 fall. It is.

  When it is determined in step SC1 of the flow that the engine 1 is in the normal rotation state (in the case of YES), the count number of the CA counter for measuring the crank angle change in the normal rotation direction of the engine 1 is set. On the contrary, when it is determined that the reverse rotation state is established (in the case of NO), the count number of the CA counter is decreased. Here, the rise and fall of the first crank angle signal CA1 and the second crank angle signal CA2 are caused by rotation of the crankshaft 3 for each predetermined angle (in this embodiment, the interval between the rise and fall is approximately 10 degrees). Therefore, the forward / reverse rotation of the engine 1 is determined as described above according to the state of the second crank angle signal CA2 when the first crank angle signal CA1 rises and falls. In addition, the rotation angle of the crankshaft 3 can be obtained from the number of rising or falling edges of the first crank angle signal CA1 and the second crank angle signal CA2. In this way, even when the crankshaft 3 rotates in the forward and reverse directions as described above when the engine is stopped, the crank angle can be accurately detected regardless of this and the piston stop position can be obtained.

  When the predetermined stop condition is satisfied during the operation of the engine 1 by the steps SA1 to SA3 of the engine automatic stop control flow shown in FIGS. 4 and 5, the fuel supply to each of the cylinders 12A to 12D is stopped, An automatic stop means 2a for automatically stopping the engine 1 is configured.

  As described above, according to the engine automatic stop control described in detail, when the engine 1 is automatically stopped by the fuel cut when idling, the throttle valve 23 is opened for the first predetermined period, and after the stop, the cylinders 12, which respectively become the expansion stroke and the intake stroke, The required amount of air is sucked into each of the cylinders 12 and the degree of decrease in the engine speed is adjusted by controlling the alternator 28 and the throttle valve 23, so that the piston 13 is moved in the expansion stroke cylinder 12 after the engine is stopped. It can be stopped within a predetermined range R suitable for restart.

  Further, as described above, when the throttle valve 23 is opened for a predetermined period in the engine stop operation period, almost all the burned gas in each of the cylinders 12A to 12D is scavenged out of the cylinder and filled with fresh air. A large amount of oxygen is stored in the catalyst 29 in the exhaust passage 22. However, after the engine 1 is stopped, the air pressure leaks immediately even in the expansion stroke cylinder 12 and the compression stroke cylinder 12 in which the intake and exhaust valves 19 and 20 are closed, so that each of the cylinders 12A to 12D has a piston. There is a state where fresh air (air) at approximately atmospheric pressure exists in the volume corresponding to the stop position.

−Automatic engine start control−
Next, the case where the engine 1 that has been automatically stopped at the time of idling as described above is automatically restarted will be described with reference to FIGS. 3 and 10 to 13. FIGS. 10 and 11 are flowcharts showing the procedure of start control. FIG. 12 shows the timing of fuel injection and ignition for each cylinder 12A to 12D at the start, and the change of the stroke of each cylinder 12A to 12D. FIG. 3 is a stroke diagram corresponding to open / close states of intake and exhaust valves. FIG. 13 shows a state in which the in-cylinder pressure of each cylinder 12A to 12D, the starting torque of the engine 1, and the engine rotation speed change due to fuel injection and ignition for each cylinder 12A to 12D at the time of the start. 6 is a time chart showing together with a decrease in the intake pressure of the engine and a change in the ignition timing that is changed in accordance therewith.

  The automatic start control of the engine 1 of this embodiment is to start the engine 1 by itself as described above. As shown in the example in FIGS. 3 and 12, the compression stroke cylinder 12A (# The engine 1 is temporarily operated reversely (Fig. 3 (a), a1 and a2 in Fig. 12), thereby compressing the expansion stroke cylinder 12B (# 2 cylinder) when stopped. Then, the air-fuel mixture in the expansion stroke cylinder 12B whose temperature and pressure have increased is ignited and burned (FIG. 3 (b), a3, a4 in FIG. 12). By doing so, the combustion pressure of the expansion stroke cylinder 12B becomes sufficiently high, and the effective stroke becomes long, so that a large starting torque can be obtained.

  However, when combustion is initially performed in the compression stroke cylinder 12A at the time of stop as described above, the burned gas resulting from this combustion is filled in the cylinder 12A, so that the cylinder 12A is started as shown in FIG. 3 (c). When the first TDC of the time is reached, the compression reaction force acting on the piston 13 becomes considerably large, and there is a possibility that the start fails because the compression reaction force cannot be exceeded. Even if the compression reaction force can be overcome and TDC can be exceeded, the engine rotational speed at that time becomes large, and the starting torque cannot be obtained by combustion in the cylinder 12A filled with burned gas. It is difficult to increase the engine speed smoothly.

  The # 3 cylinder 12C in the next ignition order of the # 1 cylinder 12A is stopped in the intake stroke (hereinafter also referred to as a stop-time intake stroke cylinder), and the air in the cylinder 12C is stopped in the engine. In addition, since the air is warmed by heat radiation from the cylinder wall surface and is in a considerably high temperature state, the air warmed in the intake passage 21 is sucked into a full filling state. The temperature and pressure become considerably high, and the engine rotation drop due to the compression reaction force becomes large, and the auto-ignition easily occurs. If self-ignition occurs, a large reverse torque acts on the engine 1 and the engine fails to start.

  In view of such a problem, in the start control of this embodiment, first, the compression stroke cylinder 12A at the time of stop is burned in a rich state of the air-fuel ratio, and the reverse rotation torque of the engine 1 is ensured. The expansion stroke cylinder 12B is sufficiently compressed, and the starting torque in the forward rotation direction due to the combustion of the cylinder 12B is greatly increased. In addition, additional fuel injection is performed when the stop-time compression stroke cylinder 12A is compressed in accordance with the forward rotation operation of the engine 1 (a5 in FIG. 12), and the compression pressure is reduced by the cooling effect due to the latent heat of vaporization. Thus, as shown in FIG. 3 (c), the rotation of the engine 1 in the forward rotation direction is maintained so as to surely exceed the TDC that the cylinder 12A first meets when starting.

  Further, with respect to the stop-time intake stroke cylinder 12C (# 3 cylinder) in the next ignition sequence of the stop-time compression stroke cylinder 12A, the cylinder 12C is shifted to the compression stroke and the temperature and pressure are increased. The fuel is injected (a6 in FIG. 12), and the compression pressure is reduced by the cooling effect by the latent heat of vaporization, and the occurrence of self-ignition is prevented. Further, by retarding the ignition timing of the cylinder 12C after the TDC (a7 in FIG. 12), it is avoided that the cylinder internal pressure increases due to the combustion before the TDC, and the drop in the engine rotation is reduced. As shown in d), after passing through the second TDC at the time of start, the engine rotation is started by igniting and burning.

  Next, the specific procedure of the start control will be described with reference to the flowcharts of FIGS. 10 and 11. This flow starts from the engine stop state (START), and first, predetermined engine restart is performed at step SB1 of FIG. It is determined whether the condition is satisfied. This restart condition is when the brake is released to start from the stop state, when the accelerator operation is performed, when the engine is required to operate the air conditioner, etc. If such a condition is not satisfied, the process waits until the condition is satisfied. If the restart condition is satisfied (YES in step SB1), the process proceeds to step SB2.

  In step SB2, the stop compression stroke cylinder 12 (# 1 cylinder 12A in FIG. 3) and the stop expansion stroke cylinder 12 (see FIG. 3) are determined based on the stop position of the piston 13 obtained by the above-described subroutine (see FIGS. 8 and 9). In FIG. 3, the amount of air in the # 2 cylinder 12B) is calculated. That is, the volumes in the stop compression stroke cylinder 12 and the stop expansion stroke cylinder 12 are determined from the stop position of the piston 13, and the cylinders 12A to 12D of the engine 1 are almost in the atmospheric pressure state as described above. Assuming that the above is satisfied, the air amounts of the cylinders 12 and 12 are calculated.

  Subsequently, in step SB3, a fuel injection amount is calculated such that a predetermined air-fuel ratio (A / F for the first compression stroke cylinder) is obtained with respect to the air amount of the compression stroke cylinder 12 at the time of stop calculated in step SB2. The fuel is injected into the compression stroke cylinder 12. Specifically, the air-fuel ratio is obtained from an air-fuel ratio map set in advance in association with the piston stop position or the like when the engine is stopped, so that the air-fuel ratio of the compression stroke cylinder 12 is greater than the stoichiometric air-fuel ratio. It is set to a rich value (A / F is preferably in the range of approximately 11 to 14, more preferably approximately 13).

  Here, since fuel leaks from the injection hole of the fuel injection valve 16 between the stop and restart of the engine 1 in the stop-time compression stroke cylinder 12, the fuel injection amount is reduced and corrected accordingly. There is a need to. Therefore, in this embodiment, as will be described in detail later, the amount of fuel leaking into each of the cylinders 12A to 12D after the engine 1 is stopped is learned in association with the passage of time after the stop, and each cylinder 12A to 12D is learned. Is stored as a fuel leakage characteristic table (time variation characteristic of the fuel leakage amount). In step SB13, the fuel leakage amount corresponding to the time from the engine stop to the fuel injection into the compression stroke cylinder 12 is stored in the table. The fuel injection amount is corrected to decrease accordingly.

  Next, in step SB4, after elapse of a predetermined time set in consideration of the fuel vaporization time from the fuel injection to the stop-time compression stroke cylinder 12, the spark plug 15 of the cylinder 12 is energized to become the air-fuel mixture. Ignite. In step SB5, whether or not the piston 13 has moved depending on whether or not the edge of the signal from the crank angle sensors 30 and 31 (rising or falling of the crank angle signal) is detected within a predetermined time from the ignition in step SB4. If the piston 13 does not move due to misfire or the like (in the case of NO), the process returns to step SB6 to the compression stroke cylinder 12. Repetitively ignite.

  On the other hand, when the edge of the crank angle signal is detected in step SB5 (in the case of YES) and it is determined that the piston 13 has moved, that is, the engine 1 has started reverse rotation, the routine proceeds to step SB7, where Fuel is injected into the expansion stroke cylinder 12 so as to be a predetermined air-fuel ratio (A / F for the expansion stroke cylinder 12) with respect to the air amount of the expansion stroke cylinder 12 at the time of stop calculated in step SB2. At this time, the air-fuel ratio is obtained from a map set in advance in association with the piston stop position at the time of engine stop in the same manner as the compression stroke cylinder 12 at the time of stop (see step SB3). A slightly rich value is set, and the fuel injection amount is corrected in consideration of the amount of fuel leakage while the engine is stopped.

  In the subsequent step SB8, the expansion stroke cylinder 12 is ignited and combusted after a predetermined time (ignition delay) has elapsed since the reverse rotation operation of the engine 1 was detected. During this ignition delay time, the piston 13 of the expansion stroke cylinder 12 at the time of stop rises due to the reverse rotation operation of the engine 1, the air-fuel mixture in the cylinder 12 is sufficiently compressed, and the piston 13 is almost stopped by the compression reaction force. This corresponds to the time until the engine is stopped, and is obtained from a map set in advance in association with the piston stop position when the engine is stopped. By igniting and burning the air-fuel mixture sufficiently compressed in the expansion stroke cylinder 12 in this way, the engine 1 starts to rotate in the forward direction with a sufficiently large torque.

  Subsequently, at step SB9, the second-time fuel injection (additional fuel injection) for the cylinder 12 is performed on the compression stroke cylinder 12 at the time of stopping when the engine 1 reaches the first TDC as the engine 1 rotates forward. It is executed at a predetermined timing considering the vaporization time. When the fuel injected in this way is vaporized, it takes heat from the surrounding gas (vaporization latent heat), thereby lowering the temperature in the compression stroke cylinder 12 and lowering the in-cylinder pressure. Even if is filled, the compression reaction force can be reduced, and the piston 13 can reliably exceed TDC. Thus, the forward rotation operation of the engine 1 started by the combustion of the stop-time expansion stroke cylinder 12 in step SB8 is continued, the stop-time compression stroke cylinder 12 exceeds TDC, and each of the cylinders 12A to 12D follows. It will proceed to the process of.

  Subsequent to step SB9, in step SB10 of FIG. 11, the air charged in the intake stroke cylinder 12 (# 3 cylinder 12C in FIG. 3) at the time of stop by the forward rotation operation of the engine 1 started as described above. Calculate the amount of In other words, the intake stroke cylinder 12 at the time of stop is the second TDC at the start time following the compression stroke cylinder 12 at the time of stop, but as described above, relatively high temperature air is present in the cylinder 12. Is fully filled at approximately atmospheric pressure and is very prone to self-ignition.

  Therefore, in step SB10, the air density in the intake stroke cylinder 12 at the time of stop is estimated based on the in-cylinder temperature and the atmospheric pressure estimated from the engine water temperature, the engine stop time, the intake air temperature, and the like. The air filling amount in the cylinder 12 is calculated based on the value. In the subsequent step SB11, a correction value to the rich side of the air-fuel ratio for preventing self-ignition is calculated mainly based on the estimated value of the in-cylinder temperature, and in the subsequent step SB12, the correction value is considered. Based on the air-fuel ratio determined in this way and the air charge amount in the cylinder 12 calculated in step SB10, the fuel injection amount to the stop-time intake stroke cylinder 12 is calculated.

  Subsequently, in step SB13, after the stop-time intake stroke cylinder 12 shifts to the compression stroke, fuel is injected into the cylinder 12 after the middle stage of the compression stroke. By doing so, the temperature and pressure in the cylinder 12 are reduced due to the latent heat of vaporization of the fuel, as in the case of the compression stroke cylinder 12 at the time of stop, so that the cylinder 12 is filled with relatively high temperature air as described above. Even if it is, it can suppress the raise of the temperature and pressure accompanying compression, and can prevent generation | occurrence | production of self-ignition. In addition, since the compression pressure of the cylinder 12 is also reduced, the drop in engine rotation during TDC ignition is reduced accordingly.

  In step SB14, after the stop-time intake stroke cylinder 12 exceeds the TDC from the compression stroke and shifts to the expansion stroke, the spark plug 15 is energized to ignite the air-fuel mixture in the cylinder 12. If this ignition timing is also at the time of engine start using a normal starter motor, it is set to an advance side from TDC (for example, about 6 ° CA before TDC). When starting the engine on its own without using it, the ignition before the TDC and the increase in the cylinder pressure due to combustion may become a torque in the reverse direction, which may hinder the engine start. The timing is delayed until after TDC.

  Subsequently, in step SB15, based on the signal from the intake pressure sensor 26, it is determined whether or not the intake pressure (intake pipe negative pressure) in the intake passage 21 downstream of the throttle valve 23 is higher than a preset value. This set value is a value such that if the intake pressure is lower than this value (the negative pressure is increased), the amount of intake air charged into the cylinder 12 does not increase so much that the engine 1 does not blow up significantly. Specifically, it is obtained by an experiment or the like, and is set slightly higher than the intake pressure state during idling after the engine 1 is warmed up.

  Then, when the intake pressure is higher than the set value (YES in step SB15), the process proceeds to step SB16, and the cylinder 12 that reaches TDC following the intake stroke cylinder 12 at the time of stop (# 4 cylinder in FIG. 3). 12D), that is, the ignition timing of the stop-time exhaust stroke cylinder 12 is also retarded after TDC, and the process returns to step SB15. Until the intake pressure becomes equal to or lower than the set value, the determination at step SB15 is YES, the process proceeds to step SB16, and the ignition timings of the cylinders 12B, 12A,. Each is retarded after TDC. On the other hand, if the intake pressure is lower than the set value, NO is determined in step SB15, the process proceeds to step SB17, and the ignition timing is set before TDC (transition to normal control).

  In other words, in this flow, the stop exhaust stroke cylinder 12D that shifts to the intake stroke with the start of the engine 1 and the cylinders 12B, 12A,... Considering that air in a state close to atmospheric pressure is sucked and the intake charge amount of those cylinders 12 increases, the ignition timing of each cylinder 12A to 12D is retarded after TDC while the intake pressure is relatively high. By doing so, an increase in combustion torque is suppressed, and a sudden increase in the engine 1 is prevented.

  When a predetermined restart condition is established after the engine 1 is automatically stopped by the steps SB1 to SB8 of the start control flow shown in FIG. 10, fuel is injected into the stop-time compression stroke cylinder 12 and ignited. The engine 1 is once reversely operated, and fuel is also injected into the expansion stroke cylinder 12 at the time of stop, which is compressed accordingly, and is ignited after the reverse operation, thereby automatically restarting without using the starting motor. The automatic starting means 2b which performs is comprised.

  In particular, in steps SB3 and SB7, the control procedure of correcting the fuel injection amount for starting in consideration of the amount of fuel leakage while the engine is stopped corresponds to the correcting means 2c. This correcting means 2c The fuel injection amount is corrected based on the time from the stop of the engine 1 to the fuel supply to each cylinder 12 by the automatic start means 2b and the characteristic of the fuel leakage amount learned for each cylinder 12. It has become.

  As described above, according to the engine automatic start control described in detail, the engine 1 automatically stopped at the time of idling can be restarted by itself without using a start motor or the like in response to a restart request. That is, as shown mainly in FIG. 13, when a restart request is made while the engine 1 is stopped (time t0), first, as shown by reference numeral a1 in FIG. The fuel injection valve 16 of the # 1 cylinder 12 </ b> A is operated, and fuel is injected into the cylinder 12 </ b> A, thereby forming a rich air-fuel mixture in the cylinder 12. When the rich air-fuel mixture is ignited by the spark plug 17 and burned (a2), a negative torque (reverse torque) is generated as indicated by T1 in FIG. It rotates in the left direction of FIG. 12 (time t1). For this reason, the engine speed temporarily becomes a negative value as shown in FIG.

  When the reverse rotation operation of the engine 1 is detected by signals from the crank angle sensors 30, 31, the fuel injection valve 16 of the stop expansion stroke cylinder 12 (# 2 cylinder 12B) is operated as shown in FIG. (A3), an air-fuel mixture is formed in the cylinder 12B, and the air-fuel mixture is compressed by the upward movement of the piston 13 due to the reverse rotation operation. At this time, since the reverse torque T1 is sufficiently large, the piston 13 of the expansion stroke cylinder 12B at the time of stop rises to the vicinity of the TDC, and the air-fuel mixture in the cylinder 12B is sufficiently compressed so that the temperature and pressure are high. become. When the engine 1 is rotated from the reverse direction to the normal direction by the compression pressure, that is, immediately after the engine speed returns from the negative value to zero as shown in FIG. (A4 in FIG. 4B), a large combustion pressure is generated thereby, the starting torque rises as shown by T2 in FIG. 4E, and the engine speed increases as shown in FIG. Time t2: Forward rotation start).

  When the inside of the stop compression stroke cylinder 12A is compressed along with the forward rotation operation, fuel injection (additional addition) into the cylinder 12A again after the middle stage of the compression stroke, as shown in FIG. Fuel injection) is performed (a5), and the inside of the cylinder 12A is cooled by the latent heat of vaporization of the fuel, so that the temperature in the cylinder 12A is higher than that in the case where this additional fuel injection is not performed (shown by a broken line in the figure). And the increase in pressure is greatly suppressed. As a result, the engine 1 can surely exceed the first TDC (first TDC) that is first met at the time of start-up, and the drop in engine rotation at that time is reduced (f). Further, the fuel injected for cooling the compression stroke cylinder 12A subsequently reacts with oxygen stored in the catalyst 29 in the exhaust passage 22 and is rendered harmless, so that no problem occurs. .

  FIG. 6 (c) shows the stop-time intake stroke cylinder 12C (# 3 cylinder) that has shifted to the compression stroke after the stop-time compression stroke cylinder 12A has exceeded the initial first TDC at the start as described above. As shown in FIG. 4, fuel is injected after the middle of the compression stroke (a6), and the inside of the cylinder 12 is cooled by the latent heat of vaporization. As a result, an increase in temperature and pressure due to compression is suppressed, the occurrence of self-ignition is prevented, and the compression reaction force of the cylinder 12C is reduced. In addition, the ignition timing for the cylinder 12C is until after TDC. By retarding, an increase in cylinder pressure due to ignition and combustion before TDC is avoided. As a result, the drop in engine rotation is reduced, and the engine 1 can reliably exceed the second TDC.

  Then, when the stopped intake stroke cylinder 12C that has shifted to the expansion stroke beyond the second TDC is ignited (a7: about 20 ° CA after TDC, for example) and combustion is performed, the engine 1 is rotated in the forward rotation direction. Torque is added, and the starting torque rises as indicated by T3 in FIG. 5 (e), and as a result, the engine speed increases to near the idle speed (650 rpm in this example) as shown in FIG. Subsequently, as shown in FIG. 6 (d), fuel is injected into the exhaust stroke cylinder 12D at the stop time in the intake stroke (a8), and ignition is performed after TDC (third TDC) (a9). In addition, fuel injection (a10, a12) and ignition (a11, a13) are also applied to the stop expansion stroke 12B and the stop compression stroke cylinder 12A, respectively. As shown in f), the engine speed gradually converges to the idle speed.

-Learning of fuel leakage-
As described above, when the engine 1 is automatically started after an idle stop or the like, fuel is injected and supplied into the cylinders 12 and 12 stopped in the compression stroke or the expansion stroke, and this is ignited and burned. In this case, the fuel leaked from the nozzle hole of the fuel injection valve 16 is vaporized and mixed in each of the cylinders 12A to 12D while the engine 1 is stopped. Qi is formed. Since the cranking is not performed when the engine 1 is started by itself, the air-fuel mixture exists in the compression stroke cylinder 12 and the expansion stroke cylinder 12 where the intake and exhaust valves 19 and 20 are closed as described above. If fuel for start-up is injected there, the air-fuel ratio in the cylinder 12 becomes excessively rich, and startability may be impaired due to deterioration of the combustion state.

  Considering this point, in the engine system E of this embodiment, after the engine 1 is manually stopped, the amount of fuel leaking into each of the cylinders 12A to 12D is learned as time elapses. Based on the result, the amount of fuel newly injected into each of the cylinders 12A to 12D at the time of automatic start can be corrected. That is, when the engine 1 is stopped (manually stopped) according to the operation of the IG switch 33, the cylinder 12 that is stopped in the expansion stroke is repeatedly ignited at a preset very short time interval, and the cylinder 12 The time until the mixture of fuel leaking into the fuel is ignited is measured, and thereby the characteristics of the time variation of the fuel leakage amount are estimated.

  Such estimation of the fuel leakage characteristic can be performed not only when the engine 1 is manually stopped but also when the engine 1 is automatically stopped. In this case, mixing is performed in the expansion stroke cylinder 12 of the engine 1 after the automatic stop. Since the air in the cylinder 12 is consumed due to ignition, burned gas is generated, and thereafter, the combustion cannot be performed in the expansion stroke cylinder 12 and the engine 1 is There is a possibility that the engine cannot be started on its own. Therefore, in this embodiment, the estimation of the fuel leakage characteristic is performed only when the engine 1 is manually stopped, and the fuel leakage characteristic table for each cylinder 12A to 12D is updated based on the estimation result (learning). I am doing so.

  Hereinafter, a control procedure for learning the fuel leakage characteristic of the predetermined cylinder 12 after the engine 1 is manually stopped will be described mainly based on the flowchart of FIG. This flow starts from the operating state of the engine 1 (START), and first, at step SD1, it is determined whether or not the IG switch 33 is turned off (IG. Off?). If it is not turned off, it waits until it is operated. If it is turned off (YES), the process proceeds to step SD2, and each cylinder is similar to the case of the automatic stop described above (see the flow of FIG. 4). Fuel injection to 12A to 12D is stopped (fuel cut), and in the subsequent step SD3, the throttle valve 23 is opened for a predetermined period (throttle opening / closing). As a result, as in the case of the automatic stop, sufficient scavenging of each of the cylinders 12A to 12D is performed, and a large amount of fresh air is supplied to the catalyst 29 in the exhaust passage 22.

  Subsequently, also in step SD4, as in the case of the automatic stop, the stop (complete stop) of the engine 1 is confirmed based on the signals from the crank angle sensors 30, 31, and until the engine 1 stops (NO ) While waiting, if the stop of the engine 1 is confirmed (YES), the process proceeds to step SD5 to detect the cylinder 12 stopped in the expansion stroke from the result of cylinder identification during engine operation, and to the crank angle signal. Based on this, the stop position of the piston in the cylinder 12 is detected.

  Subsequently, in step SD6, based on the piston stop position of the expansion stroke cylinder 12, the air amount in the cylinder 12 is calculated in the same manner as in step SB2 (see FIG. 10) of the automatic start flow described above. In step SD7, as schematically shown in FIG. 15 (a), the ignition plug 17 of the expansion stroke cylinder 12 is energized (ignited) at a set time interval (for example, 25 millisecond intervals), and in subsequent step SD8. Similarly to step SB5 of the automatic start flow, whether or not the piston 13 has moved is determined based on whether or not the edge of the signal from the crank angle sensors 30 and 31 has been detected.

  That is, after the engine 1 is stopped, the fuel leaking little by little from the injection hole of the fuel injection valve 16 in each of the cylinders 12A to 12D is vaporized in the high-temperature cylinder 12 to form an air-fuel mixture. In the compression stroke cylinder 12 and the expansion stroke cylinder 12 in which the intake valve 19 and the exhaust valve 20 are closed, as shown in FIG. 15B, the air-fuel ratio in the cylinder 12 increases as the amount of fuel leakage increases. Gradually changes to the rich side, and when ignition is possible, the air-fuel mixture is ignited by repeated ignition as described above, and combustion causes the piston 13 to move and the crank angle signal to rise. (Edge detection in (c) of the figure).

  Therefore, if the amount of fuel leaked into the expansion stroke cylinder 12 has not yet increased so much and the air-fuel ratio in the cylinder 12 is leaner than the ignition limit, no matter how much the air-fuel mixture is ignited, Without ignition, the determination in step SD8 is NO, the process returns to step SD7, and ignition is continued. When the amount of fuel leaked into the expansion stroke cylinder 12 increases with time and the air-fuel ratio in the cylinder 12 becomes richer than the ignition limit, the air-fuel mixture is ignited and the piston 13 moves. Then, when the edge of the crank angle signal is detected (YES in step SD8), the process proceeds to step SD9, and the value of the timer that measures the elapsed time from the stop of the engine 1 is read.

  Subsequently, in step SD10, the air amount in the cylinder 12 obtained in step SD6 and an air-fuel ratio (for example, about 40 at A / F) set in advance corresponding to the lean side ignition limit of the air-fuel mixture are set. The amount of fuel leaking into the cylinder 12 is calculated based on the above, and the time change of the fuel leakage amount in the cylinder 12 is estimated based on the amount of fuel and the measurement time read in step SD9. To do. Subsequently, in step SD11, the fuel leakage characteristic table of the expansion stroke cylinder 12 is corrected and updated based on the estimation result (learning the fuel leakage characteristic of the corresponding cylinder), and then the control ends (END).

  Here, as shown in FIG. 15 (b), the fuel leakage characteristic table indicates that the amount of fuel leakage increases with the passage of time. The amount of fuel leakage is taken on the shaft, and is represented as a straight or curved graph. In other words, if it is approximated that the amount of fuel leakage per hour is substantially constant, the fuel leakage characteristic is represented by a straight line graph as shown by a solid line in the figure, and the slope of the graph is changed according to the estimation result. Will be. In addition, since it can be considered that the amount of fuel leakage per hour increases as the fuel pressure increases, the amount of fuel leakage may be represented by a curve graph as shown by a broken line in the figure.

  If the fuel leakage amount corresponding to the elapsed time after the engine stop is read from the fuel leakage characteristic table corrected and updated for each cylinder 12A to 12D in this way, the amount of fuel leaked into each cylinder 12 can be obtained. The amount of fuel injected into the stop-time compression stroke cylinder 12 and the stop-time expansion stroke cylinder 12 in steps SB3 and SB7 of the above-described automatic start control flow (see FIG. 4) can be accurately estimated. Can be corrected appropriately.

  In the above flow, the cylinder 12 in the expansion stroke is repeatedly ignited after the engine 1 is stopped, and the fuel leakage characteristics in the cylinder 12 are estimated. This estimation is based on the intake valve 19 and the exhaust valve. Since it can be performed if the cylinder 12 is closed, 20 may be performed in the cylinder 12 in the compression stroke.

  Step SD2 of the learning control flow shown in FIG. 14 constitutes a manual stop means 2d for stopping the engine 1 by stopping the fuel supply to the cylinders 12A to 12D when the IG switch 33 is turned off. Has been.

  Further, the detection means 2e for detecting the amount of air in the cylinder 12 in the expansion stroke after the engine 1 is stopped is configured by the step SD6 of the flow, and the cylinder 12 in the expansion stroke after the engine 1 is stopped is configured by the step SD7. An after-stop ignition means 2f that repeatedly ignites is configured, and further, at step SD8, a determination means 2g that determines that the air-fuel mixture in the expansion stroke cylinder 12 has been ignited by the ignition is configured.

  Then, in steps SD9 to SD11 of the flow, based on the air amount in the expansion stroke cylinder 12 and the set air-fuel ratio corresponding to the lean side ignition limit, the fuel leakage in the cylinder 12 at the ignition determination time point The amount of fuel is calculated, and the time from the stop of the engine 1 to the ignition determination time is measured. Based on the measured time and the fuel leakage amount, the fuel leakage characteristic of the cylinder 12 (time variation characteristic of the fuel leakage amount) ) Is estimated and stored.

  Therefore, according to the engine system E of this embodiment, the IG switch 33 is turned off during operation of the engine 1, and the learning control (FIGS. 14 and 15) is executed when the engine 1 is stopped accordingly. Thus, the cylinder 12 in the expansion stroke is repeatedly ignited, whereby the time change of the fuel leakage amount in the cylinder 12 is estimated from the elapsed time until the mixture is ignited. Based on this estimation result, the fuel leakage characteristic table of the cylinder 12 is corrected and updated.

  On the other hand, when the engine 1 automatically stops during idling, the throttle valve 23 is opened for a predetermined period after the fuel cut by the above-described stop control (FIGS. 4 to 6 and the like), and thereafter, the alternator 28 and the throttle valve 23 are controlled. By adjusting the degree of decrease in the engine rotation speed by the control, the burned gas of each of the cylinders 12A to 12D is scavenged, and the predetermined range R suitable for restarting the piston 13 in the expansion stroke cylinder 12 after the engine is stopped. Can be stopped.

  Then, when the engine 1 is restarted thereafter, the fuel is injected into the cylinder 12 in the compression stroke and ignited by the above-described start control (FIGS. 10 to 12, etc.), whereby the engine 1 is once reversely operated. Along with this, the fuel is also injected into the expansion stroke cylinder 12 that is compressed, and is ignited after the reverse operation, whereby the torque in the forward rotation direction due to the combustion of the expansion stroke cylinder 12 becomes sufficiently large. As a result, the engine 1 can be started by itself without using a starting motor.

  In addition, at that time, the fuel injection amount into the compression stroke cylinder 12 and the expansion stroke cylinder 12 is corrected based on the fuel leakage characteristic table according to the passage of time from the stop of the engine 1 to the restart. Even if fuel leaks into each cylinder 12, the amount of newly injected fuel is corrected accordingly, and the air-fuel ratio in each cylinder 12 becomes appropriate as intended. As a result, the combustion state of each cylinder 12 is always good, so that the self-starting of the engine 1 can be made more reliable.

  In the engine system E of this embodiment, the fuel leakage characteristic is estimated only when the engine 1 is manually stopped. However, the present invention is not limited to this, and the fuel leakage characteristic is also obtained during automatic stop such as idle stop. Can be estimated. However, in this case, as described above, the air in the estimated cylinder 12 is consumed and the burned gas is generated, which makes subsequent self-starting difficult. However, it is preferable to perform cranking by the starter motor and thereby reliably start the engine 1.

  Further, if cranking is performed during the automatic start of the engine 1 as described above, there is a possibility that vibrations and noises associated with the cranking may give the driver a sense of incongruity. Therefore, it is preferable to estimate the fuel leakage characteristics during automatic stop. For example, it should be performed at a rate of about once every 50 to 100 times (once a preset number of times). Even if the fuel leakage characteristics are only occasionally estimated during automatic starting in this way, the frequency of learning of the amount of fuel leakage is significantly higher because the number of automatic stops is originally much higher than that of manual stopping. Thus, the individual variation of the fuel injection valve 16 and its secular change can be learned at an early stage and reflected in the fuel injection control at the time of automatic start.

  Furthermore, the present invention is not limited to the engine system E in which the engine 1 is first slightly rotated reversely as in this embodiment, and the mixture in the expansion stroke cylinder 12 at the time of stop is compressed and then ignited. The present invention can also be applied to an engine in which the expansion stroke cylinder 12 at the time of stop is ignited so that the engine 1 is restarted.

1 is a schematic configuration diagram of an engine control system according to an embodiment of the present invention. It is a schematic diagram which shows the structure of the intake system and exhaust system of an engine. It is explanatory drawing which shows typically the procedure of reverse rotation start of an engine. It is a flowchart which shows the control procedure of the first half of an engine automatic stop. It is a flowchart which shows the control procedure of the second half of an engine automatic stop. It is explanatory drawing which shows the change of the engine speed in the engine stop operation period, a crank angle, the throttle opening, and the intake pipe negative pressure with the change of the stroke of each cylinder. It is a distribution map which shows the correlation with the piston stop position after an engine stop and the TDC rotational speed in an engine stop operation period. It is a flowchart of the subroutine for detecting the piston position at the time of an engine stop. It is explanatory drawing which shows the crank angle signal output from two crank angle sensors, (a) is an engine normal rotation, (b) is a crank angle signal at the time of engine reverse rotation. It is a flowchart which shows the control procedure of the first half of an engine automatic start. It is a flowchart which shows the control procedure of the second half of an engine automatic start. FIG. 5 is a stroke diagram showing fuel injection and ignition timing for each cylinder at the start, together with stroke change of each cylinder and the open / closed state of the intake and exhaust valves. 3 is a time chart showing changes in in-cylinder pressure, engine torque and rotational speed for each cylinder at the time of starting, together with changes in intake pressure and ignition timing. It is a flowchart which shows the procedure of the learning control of a fuel leak amount. It is explanatory drawing which shows a mode until a fuel leak amount increases after an engine stop and it ignites an air-fuel | gaseous mixture by repeated ignition along time passage.

Explanation of symbols

E Engine system (engine starter)
1 Engine 2 ECU (Engine Controller)
2a Automatic stop means 2b Automatic start means 2c Correction means 2d Manual stop means 2e Detection means 2f Post-stop ignition means 2g Discrimination means 2h Learning means 12A to 12D Cylinder 19 Intake valve 20 Exhaust valve 33 Ignition switch (operation means)

Claims (7)

Automatic stop means for automatically stopping the engine by stopping fuel supply to each cylinder when a predetermined stop condition is satisfied during operation of the multi-cylinder engine;
When a predetermined restart condition is satisfied after the automatic stop of the engine, fuel is injected and supplied into at least the cylinder stopped in the expansion stroke, and is ignited and burned. Automatic starting means for starting; and
An engine starter comprising:
Operation means for performing an operation for manually stopping the engine;
Manual stop means for stopping fuel supply to each cylinder when the operation means is operated, and stopping the engine;
After-stop ignition means for repeatedly igniting a predetermined cylinder when the engine is stopped by the engine manual stop means;
Determining means for determining that the air-fuel mixture in the predetermined cylinder is ignited by the ignition;
Learning means for measuring a time from when the engine is stopped until ignition is discriminated by the discriminating means, and estimating and storing a time change characteristic of the fuel leakage amount in the cylinder based on at least the measurement time;
An engine starting device comprising: The automatic starting means is for injecting and supplying fuel into a cylinder that is stopped in the expansion stroke, and igniting.
Based on the time from the stop of the engine to the fuel supply by the automatic start means and the time change characteristic of the fuel leakage amount of the cylinder stored by the learning means, the fuel to the expansion stroke cylinder by the automatic start means 2. The engine starting device according to claim 1, further comprising correction means for correcting the supply amount. The automatic starting means injects and supplies fuel into the cylinders stopped in the compression stroke, and ignites them to temporarily reverse the engine, and also injects fuel into the expansion stroke cylinders compressed accordingly. The engine is started by supplying an injection and igniting after reverse rotation,
Based on the time from the engine stop to the fuel supply to each cylinder by the automatic starting means and the time variation characteristic of the fuel leakage amount for each cylinder stored by the learning means, the automatic starting means 2. The engine starting device according to claim 1, further comprising correction means for correcting the amount of fuel supplied to each cylinder.   The engine starting device according to any one of claims 1 to 3, wherein the post-stop ignition means ignites a cylinder in which both the intake valve and the exhaust valve are closed.   The engine starting device according to any one of claims 1 to 3, wherein the post-stop ignition means ignites a cylinder stopped in either the compression stroke or the expansion stroke. A detection means for detecting the amount of air in the cylinder that is ignited by the ignition means after stopping after the engine is stopped;
The learning means calculates the amount of fuel leakage at the time of ignition discrimination by the discrimination means based on the air-fuel ratio set in advance corresponding to the lean side ignition limit of the air-fuel mixture and the detected air amount, 6. The time variation characteristic of the fuel leakage amount is estimated on the basis of the fuel leakage amount and the measurement time until the ignition determination. The fuel leakage amount according to any one of claims 1 to 5, Engine starter. Automatic stop means for automatically stopping the engine by stopping fuel supply to each cylinder when a predetermined stop condition is satisfied during operation of the multi-cylinder engine;
When a predetermined restart condition is satisfied after the automatic stop of the engine, fuel is injected and supplied into at least the cylinder stopped in the expansion stroke, and is ignited and burned. Automatic starting means for starting; and
An engine starter comprising:
Post-stop ignition means for repeatedly igniting a predetermined cylinder at a rate of once per preset number of times when the engine is stopped by the automatic stop means;
Determining means for determining that the air-fuel mixture in the predetermined cylinder is ignited by the ignition;
Learning means for measuring a time from when the engine is stopped until ignition is determined by the determination means, and estimating and storing a time change characteristic of the fuel leakage amount in the cylinder based on at least the measurement time. ,
The automatic starter is configured to start the engine using a starter motor when an engine restart condition is satisfied after ignition is determined by the determination unit. .

Images