JP2005048747A - Combustion control device for internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To smoothly change combustion modes without causing a fire. <P>SOLUTION: In changing a normal combustion mode into an air-fuel ratio enriching combustion mode by raising exhaust temperature in regeneration of an exhaust emission control device or the like, fuel injection is executed in fuel injection quantity for preliminary combustion at fuel injection timing for preliminary combustion. It is then changed into fuel injection timing for main combustion with retard. Specifically, they are gradually changed, taking several cycles till target main injection timing is reached. For the fuel injection quantity for main injection, generation torque is reduced as the fuel injection timing is retarded. For maintaining the torque, increasing quantity is corrected in accordance with retard. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a combustion control device for an internal combustion engine.

Conventionally, as disclosed in Patent Document 1, in a fuel injection device for a diesel engine, when a temperature rise of a catalyst is urged, fuel of a basic fuel injection amount corresponding to a required torque of the engine is supplied by a fuel injection valve. In addition, it is known to divide and inject three times near the compression top dead center of each cylinder. In conjunction with this, it is also known to increase the fuel injection amount.
JP 2000-320386 A

However, in the apparatus described in Patent Document 1, since the fuel is injected so that the combustion of the separately injected fuel continues, the fuel is injected into the flame of the initially injected fuel. The fuel injected after the second time becomes the combustion mainly of diffusion combustion. If the air-fuel ratio is enriched in such a combustion state, a significant deterioration of smoke is inevitable.
In view of such circumstances, the present invention has an object of realizing combustion that does not cause smoke deterioration even when the air-fuel ratio is enriched, for example, when the exhaust gas temperature is increased. Another object of the present invention is to realize a target air-fuel ratio without causing a torque change when switching combustion.

  For this reason, the present invention provides a main combustion for generating a main torque under a predetermined condition based on the state of an exhaust gas purification device provided in an exhaust passage of an internal combustion engine, and at least one time performed prior to the main combustion. Pre-combustion is performed, and at least one of the pre-combustion occurs near the compression top dead center, and the main combustion is injected after the pre-combustion is completed. While the combustion control to be controlled is performed, when the combustion control is switched from the combustion performed previously, the switching of the main combustion is delayed after the switching of the preliminary combustion.

According to such a configuration, the main combustion starts after the pre-combustion is completed, whereby the main combustion can be made pre-mixed combustion main combustion, and the deterioration of smoke due to enrichment can be suppressed.
In addition, since the in-cylinder temperature is increased by the preliminary combustion, it is possible to retard the generation timing of the main combustion, thereby increasing the exhaust temperature.
Therefore, by switching to the combustion mode, the exhaust air-fuel ratio can be enriched and / or the exhaust purification device can be regenerated by raising the temperature without deteriorating the smoke.

  Here, when switching to the combustion when the predetermined condition is satisfied, the preliminary combustion performed in the vicinity of the compression top dead center is hardly affected by the intake system fluctuations even if the intake system fluctuations occur. Although it can be switched to the value, the main combustion after the compression stroke is easily affected by fluctuations in the intake system, and if it is immediately switched to the target value, the ignition becomes unstable and misfires and the combustion is not established.

  Therefore, by switching the preliminary combustion first to ensure the compression end temperature that enables the main combustion, the main combustion switching is delayed, so that the intake system fluctuation is reduced and then the main combustion can be switched to. Therefore, the main combustion can be established stably, and the smooth combustion control can be switched.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a system configuration diagram showing a combustion control device of a diesel engine 1 as an internal combustion engine for a vehicle.
A compressor 3a of the turbocharger 3 is disposed upstream of the intake passage 2 of the engine 1, and the intake air is supercharged by the compressor 3a, then cooled by the intercooler 4, and after passing through the intake throttle valve 5, It flows into the combustion chamber of each cylinder through the collector 6. The fuel is increased in pressure by the fuel injection pump 7 and sent to the common rail 8 and is directly injected into the combustion chamber from the fuel injection valve 9 of each cylinder. The air that has flowed into the combustion chamber and the injected fuel are combusted by compression ignition, and the exhaust gas flows out into the exhaust passage 10. The fuel injection pump 7, the common rail 8, and the fuel injection valve 9 constitute a common rail fuel injection device.

Part of the exhaust gas flowing into the exhaust passage 10 is recirculated to the intake side via the EGR valve 12 through the EGR passage 11 as EGR gas. The remainder of the exhaust passes through the exhaust turbine 3b of the turbocharger 3 and drives it.
Here, downstream of the exhaust turbine in the exhaust passage 10, for exhaust purification, NOx in the exhaust is trapped when the exhaust air-fuel ratio is lean, and the trapped NOx is desorbed and purified when the exhaust air-fuel ratio is rich. A possible NOx trap catalyst 13 is arranged. Further, the NOx trap catalyst 13 carries an oxidation catalyst (noble metal) and has a function of oxidizing the exhaust components (HC, CO) flowing in.

  Further, a diesel particulate filter (hereinafter referred to as DPF) 14 for collecting PM (Particulate Matter), which is particulate matter in the exhaust gas, is disposed downstream of the NOx trap catalyst 13. The DPF 14 also has a function of supporting an oxidation catalyst (noble metal) and oxidizing exhaust components (HC, CO) flowing in. Note that the NOx trap catalyst 13 and the DPF 14 may be disposed in reverse, or may be configured integrally by supporting the NOx trap catalyst on the DPF.

The control unit 20 includes a rotational speed sensor 21 for detecting the engine rotational speed Ne, an accelerator opening sensor 22 for detecting the accelerator opening APO, an air flow meter 23 for detecting the intake air amount Qac, and an engine for controlling the engine 1. A signal is input from the water temperature sensor 24 for detecting the cooling water temperature Tw.
Further, a catalyst temperature sensor 25 for detecting the temperature of the NOx trap catalyst 13 (catalyst temperature), an exhaust pressure sensor 26 for detecting the exhaust pressure on the DPF 14 inlet side of the exhaust passage 10, and a DPF for detecting the temperature of the DPF 14 (DPF temperature). A temperature sensor 27 and an air / fuel ratio sensor 28 for detecting an exhaust air / fuel ratio (hereinafter referred to as exhaust λ, which is expressed as an excess air ratio) are provided on the outlet side of the DPF 14 in the exhaust passage 10. 20 is input. However, the temperature of the NOx trap catalyst 13 and the temperature of the DPF 14 may be indirectly detected from the exhaust temperature by providing an exhaust temperature sensor on the downstream side thereof.

Based on these input signals, the control unit 20 controls the fuel injection amount of the fuel injection by the fuel injection valve 9 and the fuel injection command signal to the fuel injection valve 9 for injection timing control, and the opening to the intake throttle valve 5. A command signal, an opening command signal to the EGR valve 12, and the like are output.
Here, in the control unit 20, regeneration of the DPF 14 by combustion removal of the PM collected and deposited in the DPF 14, desorption purification of NOx trapped in the NOx trap catalyst 13, S (sulfur) poisoning of the NOx trap catalyst 13. Exhaust gas purification control including cancellation is performed, and the exhaust gas purification control will be described in detail below.

2 to 12 are flowcharts of exhaust purification control executed by the control unit 20.
First, a description will be given along the flow of FIG.
In S1, various sensor signals are read, and the engine speed Ne, the accelerator opening APO, the intake air amount Qac, the catalyst temperature, the DPF inlet side exhaust pressure, the DPF temperature, and the DPF outlet side exhaust λ are detected.

  In S2, a warm-up (active) / cool-down (inactive) state of the NOx trap catalyst of the exhaust system is determined. When the catalyst temperature is equal to or lower than the activation temperature T5, it is determined that the engine is in the cold state (inactive state), and the control proceeds to the warm-up promotion mode control in FIG. If it is determined that the warm-up state (the active state after completion of warm-up), the process proceeds to S3. The active state of the NOx trap catalyst may be determined based on the concentration of at least one of HC and CO at the catalyst outlet. That is, when the concentration of at least one of HC and CO is larger than the threshold value, it is determined that the catalyst is in an inactive state. The determination of the HC and CO concentrations can also be performed based on the output level of the air-fuel ratio sensor 28 (it is estimated that the higher the rich degree, the higher the HC and CO concentrations).

  In S3, the NOx deposition amount trapped and deposited on the NOx trap catalyst is calculated. For example, it may be estimated from the integrated value of the engine speed as in the calculation of the NOx absorption amount described in Japanese Patent No. 2600492, page 6, or may be estimated from the travel distance. When the integrated value is used, the integrated value is reset at the time when NOx desorption purification is completed (including the time when NOx desorption purification is simultaneously performed by performing the S poison removal).

In S4, the S deposition amount deposited on the NOx trap catalyst due to S poisoning is calculated. Here, similarly to the calculation of the NOx accumulation amount, it may be estimated from the engine rotation speed integrated value and the travel distance. When the integrated value is used, the integrated value is reset when the S poisoning release is completed.
In S5, the PM deposition amount collected and deposited in the DPF is calculated as follows. If the amount of accumulated PM in the DPF increases, the exhaust pressure on the DPF inlet side naturally increases, so the exhaust pressure sensor detects the exhaust pressure on the DPF inlet side, and the reference exhaust in the current operating state (engine speed, load). The amount of PM deposition is estimated by comparison with the pressure. Note that the PM accumulation amount may be estimated by combining the engine rotation speed integrated value or travel distance from the previous DPF regeneration and the exhaust pressure.

In S6, it is determined whether or not a reg flag indicating that the DPF regeneration mode is in effect is set. When the reg flag = 1, the process proceeds to the DPF regeneration mode control shown in FIG.
In S7, it is determined whether or not a desul flag indicating that the NO poisoning release mode of the NOx trap catalyst is being set is set. When the desul flag = 1, the process proceeds to the control of the S poison release mode of FIG. 4 described later.

In S8, it is determined whether or not the sp flag indicating that the NOx trap catalyst is in the rich spike mode for NOx desorption purification is set. If the sp flag = 1, the control proceeds to the rich spike mode control of FIG.
In S9, it is determined whether or not a rec flag indicating that the melting prevention mode is in effect after the DPF regeneration and the S-poisoning release is set. When the rec flag = 1, the process proceeds to the control of the flaw prevention mode shown in FIG.

In S10, it is determined whether or not an rq-DPF flag indicating that a DPF regeneration request has been issued is set. When the DPF regeneration request is issued and the rq-DPF flag = 1, the process proceeds to the flow of FIG. 7 described later, and the priority of regeneration when the DPF regeneration request is issued is determined.
In S11, it is determined whether or not an rq-desul flag indicating that an S poison release request has been issued is set. When the S poison removal request is issued and the rq-desul flag = 1, the process proceeds to the flow of FIG. 8 described later, and the priority of regeneration when the S poison removal request is issued is determined.

In S12, it is determined whether or not the PM accumulation amount of the DPF calculated in S5 has reached the predetermined amount PM1 and the DPF regeneration time has come.
If it is determined that the PM accumulation amount> PM1 and the DPF regeneration timing is reached, the flow proceeds to the flow of FIG. 9, and the rq-DEF flag is set to 1 in S701 to issue a DPF regeneration request.
In S13, it is determined whether or not the S accumulation amount of the NOx trap catalyst calculated in S4 has reached a predetermined amount S1 and the S poisoning release time has come.

If the S accumulation amount> S1 and it is determined that the NO poisoning release timing of the NOx trap catalyst is reached, the flow proceeds to the flow of FIG. 10, and the rq-desul flag is set to 1 in S801, and an S poison removal request is issued.
In S14, it is determined whether or not the NOx trap amount of the NOx trap catalyst calculated in S3 has reached a predetermined amount NOx1 and the NOx desorption purification time has come.

If NOx accumulation amount> NOx1 and NOx desorption purification time of the NOx trap catalyst is determined, the process proceeds to the flow of FIG. 11 and the rq-sp flag is set to 1 in S901 to request NOx desorption purification (rich spike request). ).
Next, control of the DPF regeneration mode in FIG. 3 will be described. When the PM accumulation amount reaches the predetermined amount PM1, the rq-DPF flag = 1 is set, and when the reg flag = 1 is set by the flow of FIG. 7 described later, the flow of FIG. 3 is started.

In S101, the engine combustion is switched from the normal lean combustion to the split retard combustion according to the present invention for regeneration of the DPF.
Here, the split retard combustion according to the present invention will be described. In addition to DPF regeneration, this combustion is used for S poison release, NOx desorption purification (rich spike), and warm-up promotion.
When the DPF is regenerated, the exhaust λ needs to be controlled between 1 and 1.4 and the temperature of the DPF needs to be 600 ° C. or higher. In addition, when performing S poisoning cancellation, it is necessary to realize λ ≦ 1 and exhaust temperature ≧ 600 ° C.

In the normal operation region under lean conditions, pilot injection is usually performed, the pilot injection timing is 40 to 10 ° BTDC, the pilot injection amount is 1 to ³ mm ³ / st, and the main injection timing is about 10 to −5 ° BTDC. The interval between the pilot injection and the main injection is set to about 10 to 30 ° CA.
In order to achieve low λ and high exhaust temperature, such as DPF regeneration and S-poisoning release, from normal operation, it is necessary to reduce the intake air amount. However, if the intake air amount is reduced, the compression end temperature in the cylinder will decrease, and the combustion will become unstable. With the pilot injection setting similar to normal lean combustion, the injection timing of the main injection will be advanced. It is necessary to make an angle (FIG. 13; Reference Example (1)). With such fuel injection amount and injection timing settings, even if you want to retard the injection timing in order to raise the exhaust temperature, the combustion becomes unstable, so there is a limit to retard, and low λ, high emission It is difficult to achieve temperature.

Therefore, in Patent Document 1, the main injection is divided to widen the retard limit of the injection timing to achieve low λ and high exhaust temperature (FIG. 14; Reference Example (2)).
However, since the next fuel is injected while the combustion of the fuel blown before is active, the combustion is continuous as shown in FIG. That is, since the fuel divided for main combustion is injected into the flame of the previously injected combustion, combustion starts as soon as it is injected, the diffusion combustion ratio increases, and the partial equivalent ratio Will be very rich and smoke will be significantly worse.

Therefore, in the present invention, as shown in FIG. 15, the main combustion for generating the main torque and the preliminary combustion performed prior to the main combustion are performed, and the preliminary combustion is performed near the compression top dead center (TDC). Wake up and the main combustion controls the fuel injection (a, b) to start after the pre-combustion ends.
That is, fuel is first injected in the compression stroke (a), and preliminary combustion for increasing the in-cylinder temperature (compression end temperature) in the vicinity of TDC is performed. Depending on the operating conditions, the amount of pre-combustion heat generation differs, but at least the pre-combustion heat generation can be confirmed, and the in-cylinder temperature during fuel injection for main combustion exceeds the temperature at which self-ignition is possible. The required amount of fuel is injected. Further, the stability of the preliminary combustion can be improved by changing the fuel injection amount and the timing for the preliminary combustion according to the compression end temperature expected in each operation condition.

Subsequently, after the preliminary combustion is completed, fuel for main combustion is injected after TDC so that main combustion starts (b).
In other words, by increasing the in-cylinder temperature by pre-combustion, the retard limit of main combustion is expanded and the controllability to the target temperature is improved, while the fuel of main combustion is injected after the pre-combustion is finished reliably. The ignition delay period for main combustion is secured, the premixed combustion ratio of main combustion is increased, and smoke emission is suppressed.

  Although the interval from the start timing of the pre-combustion to the start timing of the main combustion depends on the engine rotation speed, the pre-combustion (heat generation by the pre-combustion) is not completely completed unless it is separated by at least 20 ° CA or more. . By setting such an interval, deterioration of main combustion can be suppressed, and deterioration of smoke can be prevented. In addition, since the main combustion starts in the expansion stroke, the combustion speed is very slow, and the main combustion ends after 50 ° ATDC. By making the end timing of the main combustion as late as possible, the main combustion becomes slow and the deterioration of the combustion noise can be suppressed.

If the split retard combustion according to the present invention is realized, as shown in (3) of FIG. 16, compared with the reference examples (1) and (2), even when the rich condition is realized, the exhaust gas temperature is high. In addition, low smoke combustion can be realized. Further, HC also shows a very low value.
In addition, since the retard limit of main combustion is widened by pre-combustion, combustion under a low λ condition is stable even when the injection timing of main injection is retarded, and a high exhaust temperature can be realized.

Referring to FIG. 17, if the timing of main combustion is retarded, the premixing ratio of main combustion is increased, so that smoke is suppressed only by retarding even when λ is small. Further, if the timing of the main combustion is retarded, a higher exhaust temperature can be realized, and the exhaust temperature can be controlled by changing the fuel injection timing for the main combustion.
FIG. 18 shows the target fuel injection timing for the preliminary combustion using the engine operating conditions (engine rotational speed Ne, load Q) as parameters.

FIG. 19 shows the target fuel injection amount for the preliminary combustion using the engine operating conditions (engine rotational speed Ne, load Q) as parameters.
FIG. 20 shows a target fuel injection timing (main injection timing) for main combustion for realizing a certain target exhaust temperature using engine operating conditions (engine rotation speed Ne, load Q) as parameters.

  Note that when the load is low, the combustion timing of the main combustion for achieving the target exhaust temperature is very retarded, so that the in-cylinder temperature at the injection timing of the main combustion may not be maintained high only once in the preliminary combustion. . In that case, as shown in FIG. 21, preliminary combustion is performed a plurality of times so that the respective heat generations do not overlap, so that both low smoke and high exhaust temperature can be achieved even under low load conditions.

From the above, when low λ and high exhaust temperature are required for DPF regeneration, S poison removal, etc., switching to split retard combustion according to the present invention is performed.
Specifically, this is performed as shown in the flowchart of FIG. In the embodiment shown in FIG. 22A, in S1101, fuel injection is performed at the fuel injection timing for preliminary combustion (FIG. 18) at the fuel injection amount for preliminary combustion (FIG. 19).

  In S1102, the fuel injection for main combustion is performed by retarding the fuel injection timing. However, the fuel injection timing for the preliminary combustion is switched at once, whereas the fuel injection timing for the main combustion is switched with a delay. Specifically, the switching is gradually performed in several cycles until the target main injection timing shown in FIG. The fuel injection amount of the main injection is gradually increased and corrected in accordance with the retard amount so as to maintain the torque because the generated torque decreases as the fuel injection timing is retarded.

  That is, when switching from normal combustion to split retard combustion according to the present invention for the purpose of regeneration of DPF or NOx trap catalyst, the amount of fresh air and EGR are also changed simultaneously by changing the target λ, etc. Accompanied by. As described above, the preliminary combustion performed near the compression top dead center is not easily affected by the intake system fluctuation, so it can be immediately switched to the target value, but the main combustion after the compression stroke is the intake system fluctuation. It is easily affected, and if it is immediately switched to the target value, ignition becomes unstable and misfire is likely to occur.

  More specifically, the change in the fresh air amount and the behavior of the EGR gas when the throttle operation is performed according to the change in the target λ accompanying the combustion switching affects the main combustion. In other words, while the throttle opening is throttled immediately following the target, a delay occurs in the development of the differential pressure across the throttle to a differential pressure corresponding to the target throttle opening. In this case, the compression end temperature is excessively lowered, and in the worst case, the main combustion is misfired. Similarly, since the differential pressure across the EGR valve is not stable during the transition, the EGR may increase excessively. In this case, the ignition delay increases, and in the worst case, the main combustion is misfired. Furthermore, there may be a case where the misfire tendency of the main combustion increases due to the adverse effects of the fresh air amount and EGR.

Therefore, as described above, first, the pre-combustion is switched to ensure the compression end temperature at which the main combustion is possible, and then the main combustion is gradually switched, so that the main system is changed after the intake system fluctuation becomes small. Be able to switch to combustion.
Thereby, main combustion can be established stably and smooth combustion control can be switched.

The main combustion switching speed may be set according to the delay element of the intake system (collector capacity: 2 to 3 cycles) actual λ (new air amount, fluctuation of EGR). The setting is made such that the retarded amount until the main injection timing is retarded by 1/3 and the switching is completed in 3 cycles.
Another embodiment is shown in FIG.

In S1111, a delay time until the transient fluctuation of the intake system (new air amount or EGR amount) converges is set in accordance with the target λ after the combustion switching (searching may be performed by setting a map or the like).
In S1112, pre-combustion is controlled as in S1101.
In S1113, the process waits for the time until the new air amount and the EGR gas transient fluctuations due to the operation of the intake system converge. This waiting time may be set to be equal to or longer than the convergence time obtained experimentally in advance.

In step S1114, the main combustion is controlled after the convergence time has elapsed.
Even in this case, the main combustion is switched after the transient fluctuations of the fresh air amount and the EGR gas are converged and the differential pressure across the throttle valve and the EGR valve is stabilized, so that the main combustion can be established stably. Smooth combustion control can be switched.
Returning to FIG. 3, after switching the engine combustion from the normal lean combustion to the split retard combustion according to the present invention for DPF regeneration in S101, the process proceeds to S102.

In S102, the exhaust λ is controlled to a target value. In the regeneration of the DPF, the target value of the exhaust λ varies depending on the PM accumulation amount. Accordingly, the DPF inlet side exhaust pressure is detected, and the PM accumulation amount is estimated by comparison with the reference exhaust pressure in the current operation state (engine speed, load), and the target corresponding to the PM accumulation amount shown in FIG. λ is set and controlled.
Here, after switching to the split retard combustion in S101, the target air-fuel ratio is controlled by adjusting the amount of fresh air by the intake throttle valve 6 and / or the EGR valve 19. Specifically, in order to obtain the target intake air amount, the intake throttle is set so that the target intake air amount is obtained by multiplying the target λ by the value of the map shown in FIG. 24 (target intake air amount for the operation of λ = 1). If the air-fuel ratio deviates from the target value after being controlled by the valve 6, the intake throttle valve 6 and / or the EGR valve 19 adjusts the target λ.

However, when switching to the above-described split retard combustion, the fuel injection timing is significantly retarded. Therefore, in addition to the control of the intake air amount, the target injection timing (main injection timing) shown in FIG. 24), the target intake air amount and the fuel injection amount in FIG. 24 are corrected.
Further, when the target air-fuel ratio decreases to the stoichiometric air-fuel ratio or a value close thereto, a pumping loss due to the intake throttle occurs, so that the target intake air amount and the torque correction coefficient K2 corresponding to the target λ as shown in FIG. The fuel injection amount for main combustion is corrected.

In S103, it is determined whether or not the DPF temperature has exceeded the target upper limit value T22 being regenerated.
If the DPF temperature is greater than T22, the upper limit value has been exceeded during regeneration, so the routine proceeds to S111 to advance the fuel injection timing of main combustion to lower the exhaust temperature, and then at S112, advance the fuel injection timing. Torque correction (decrease correction of the main combustion injection amount) is performed to suppress the torque increase due to.

When DPF temperature ≦ T22, the routine proceeds to S104, where it is determined whether or not the DPF temperature has fallen below the target lower limit value T21 being regenerated.
If the DPF temperature is less than T21, the lower limit value has been fallen during regeneration. Therefore, the process proceeds to S109, the main fuel injection timing is retarded to increase the exhaust temperature, and then the fuel injection timing is retarded in S110. Torque correction (increase correction of the main combustion injection amount) is performed to compensate for the torque drop due to.

In S105, it is determined whether or not a predetermined time tdpfreg has elapsed from the start of DPF regeneration. If the predetermined time has elapsed, the PM deposited on the DPF is surely burned and removed, and the process proceeds to S106.
In S106, since the regeneration of the DPF is completed, the divided retard combustion according to the present invention is switched to the normal combustion, and the heating of the DPF is stopped.

In S107, since regeneration of the DPF is completed, the reg flag is set to 0. In S108, regeneration of the DPF is completed, but when there is PM unburned in the DPF, if the exhaust λ is suddenly increased, PM in the DPF Since there is a risk of burning at once and there is a risk of melting, the rec flag is set to 1 to enter the melting prevention mode.
Next, the control in the S poisoning release mode of FIG. 4 will be described. When the S accumulation amount of the NOx trap catalyst reaches the predetermined value S1 and becomes rq-desul flag = 1, and when the desul flag is set to 1 in the flow of FIG. 8 described later, the flow of FIG. 4 is started.

In S201, the engine combustion is switched from normal lean combustion to split retard combustion according to the present invention in order to release S poisoning of the NOx trap catalyst.
In S202, the exhaust λ is controlled to stoichiometric. That is, control is performed by setting the target λ to stoichiometric.
In S203, it is determined whether or not the catalyst temperature is higher than a predetermined value T4. For example, in the case of a Ba-based NOx trap catalyst, T4 is set to 600 ° C. or higher because it is necessary to set it to 600 ° C. or higher in a rich to stoichiometric atmosphere.

If the catalyst temperature is lower than the predetermined value T4, the process proceeds to S210, the fuel injection timing of main combustion is retarded, and the exhaust temperature is raised.
In S204, it is determined whether or not a predetermined time tdesul has elapsed in the S poisoning release mode. If the predetermined time has elapsed, S poisoning is released, and the process proceeds to S205.
In S205, since the S poison release has been completed, the split retard combustion according to the present invention is switched to the normal combustion, and the heating of the NOx trap catalyst is stopped. Of course, at the same time, the stoichiometric operation is canceled. In S206, since the S poisoning cancellation is completed, the desul flag is set to 0.

In S207, S poison release is completed, but if PM is accumulated in the DPF under such a high temperature condition, if the exhaust λ is suddenly increased, the DPF may burn at once and melt down. For this reason, the rec flag is set to 1 to enter the melt prevention mode.
In S208, the rq-sp flag is set to 0. When the S poisoning release is performed, the NOx trap catalyst is exposed to stoichiometry for a long time, so that NOx desorption purification is performed at the same time. This is to withdraw.

Next, the control in the rich spike mode (NOx desorption purification mode) in FIG. 5 will be described. When the NOx accumulation amount of the NOx trap catalyst reaches the predetermined value NOx1 and becomes rq-sp flag = 1, and when the sp flag is set to 1 by the flow of FIG. 7 or FIG. 8 described later, the flow of FIG. 5 is started. The
In S301, the engine combustion is switched from normal lean combustion to split retard combustion according to the present invention for NOx desorption purification of the NOx trap catalyst.

In S302, the exhaust λ is controlled to be rich. The exhaust λ is controlled by the intake throttle valve 6 so as to be the target intake air amount in the rich spike operation shown in FIG. 27, and then the fresh air amount is adjusted by the intake throttle valve 6 and the EGR valve 19 as described above. Done in
In S303, it is determined whether or not a predetermined time tspike has elapsed in the rich spike mode. If it has elapsed, it is considered that NOx desorption purification has been completed, and the process proceeds to S304.

In S304, since NOx desorption purification is completed, the split retard combustion according to the present invention is switched to normal combustion. Of course, at the same time, the rich operation is canceled. In S305, since NOx desorption purification is completed, the sp flag is set to zero.
Next, the control of the melt damage prevention mode of FIG. 6 will be described. When DPF regeneration or S-poisoning release is completed and the rec flag is set to 1 in the flow of FIG. 3 or FIG. 4, the flow of FIG. 6 is started.

  In S401, immediately after the DPF regeneration or the like, it is still in a high temperature state, and if the exhaust λ is suddenly leaned, the unburned PM in the DPF may burn at once and melt, so the exhaust λ is set to a target value, for example, λ Control to ≦ 1.4. In the melt prevention mode, it is desirable that the exhaust gas temperature is low. Therefore, the exhaust λ is controlled to the target value by the normal combustion instead of the divided retard combustion according to the present invention.

In S402, it is determined whether or not the DPF temperature has become lower than a predetermined temperature T3 (for example, 500 ° C.) at which there is no risk of rapid oxidation of PM. If it is higher than T3, the exhaust λ control is continued. If it is lower than T3, the DPF can be prevented from being melted even when the oxygen concentration is the same as the atmosphere, so the process proceeds to S403.
In S403, the exhaust λ control is stopped because there is no fear of melting of the DPF.

In S404, since the melting prevention mode has ended, the rec flag is set to zero.
Next, the reproduction priority order determination flow (1) in FIG. 7 will be described. When a DPF regeneration request (rq-DPF flag = 1) is issued, the flow of FIG. 6 is started. In addition, this flow prescribes | regulates the priority when a DPF regeneration request | requirement, a S poison removal cancellation request | requirement, or a NOx desorption purification request | requirement occur simultaneously.

In S501, after the DPF regeneration request is issued, it is determined by the same method as S13 whether or not the S accumulation amount reaches the predetermined value S1 and the S poisoning release timing is reached.
When the S accumulation amount> S1, the process proceeds to S801 in the flow of FIG. 10, the rq-desul flag = 1 is set, and an S poisoning release request is issued. In this case, the priority order is determined by the flow of FIG. 8 described later.

If S accumulation amount <S1, the process proceeds to S502.
In S502, it is determined whether rq-sp flag = 1, that is, whether a NOx desorption purification request (rich spike request) has been issued. If not, the process proceeds to S503.
In S503, after the DPF regeneration request is issued, it is determined by the same method as in S14 whether or not the NOx accumulation amount reaches the predetermined value NOx1 and the NOx desorption purification time is reached.

If NOx accumulation amount> NOx1, the process proceeds to S901 in the flow of FIG. 11, and the rq-sp flag = 1 is set, and a NOx desorption purification request (rich spike request) is issued.
In the determination in S503, if NOx accumulation amount <NOx1, only the DPF regeneration request is issued. In this case, the process proceeds to S504.
In S504, the DPF regeneration and S poisoning release possible region shown in FIG. 28 (region other than low rotation / low load; a region where the temperature increase is relatively small and the exhaust performance deterioration does not exceed the allowable value). It is determined whether or not. In the case of the reproducible area, the process proceeds to S505, where the reg flag = 1 is set and the process proceeds to DPF regeneration.

In the determination at S502, when the rq-sp flag = 1, the DPF regeneration request and the NOx desorption purification request are issued simultaneously. In this case, the process proceeds to S506.
In step S506, it is determined whether or not the engine operating condition is a condition (for example, a steady condition) with a small NOx emission amount. If the amount of NOx emission is small, even if the regeneration of the NOx trap catalyst is somewhat delayed, there is almost no deterioration of the exhaust in the tailpipe. Therefore, it is desirable to prioritize the regeneration of the DPF that greatly affects the operability. Accordingly, in this case, the process proceeds to S507.

It is desirable to prioritize regeneration of the NOx trap catalyst in order to prevent exhaust deterioration in the tail pipe under conditions where the amount of NOx emission is large (for example, acceleration conditions). Therefore, in this case, the process proceeds to S508, where the sp flag = 1 is set, and the process proceeds to NOx desorption purification (rich spike).
In S507, it is determined whether or not the DPF temperature is higher than a temperature T6 at which the oxidation catalyst supported on the DPF is activated. When starting the temperature rise, if the oxidation catalyst supported on the DPF is lower than the activation temperature T6, it takes time to reach a reproducible temperature even if the temperature rise is started. Therefore, it is desirable to give priority to regeneration of the NOx trap catalyst. Accordingly, also in this case, the process proceeds to S508, where the sp flag = 1 is set, and the process proceeds to NOx desorption purification (rich spike).

If it is determined in S507 that DPF temperature> T6, the process proceeds to S504 and 505 described above in order to prioritize the regeneration of the DPF.
Next, the reproduction priority order determination flow (2) in FIG. 8 will be described. When an S poisoning release request (rq-desul flag = 1) is issued, the flow of FIG. 8 is started. In addition, this flow prescribes | regulates the priority when an S poisoning cancellation | release request | requirement and a NOx desorption purification request | requirement occur simultaneously.

In S601, after the S poisoning release request is issued, it is determined by the same method as in S12 whether or not the PM accumulation amount reaches the predetermined value PM1 and the DPF regeneration time is reached.
When PM accumulation amount> PM1, the process proceeds to S701 in the flow of FIG. 9, and the rq-DPF flag = 1 is set, and a DPF regeneration request is issued. In this case, the priority order is determined by the flow of FIG.

If PM deposition amount <PM1, the process proceeds to S602.
In S602, it is determined whether or not the catalyst temperature is higher than a predetermined temperature T1, and if it is higher, the process proceeds to S603.
In S603, the DPF regeneration and S-poisoning release possible region shown in FIG. 24 (region other than low rotation / low load; a region where the temperature increase is relatively small and the exhaust performance deterioration does not exceed the allowable value). It is determined whether or not. In the case of the S poisoning releasable area, the process proceeds to S604, the desul flag = 1 is set, and the process proceeds to S poisoning cancellation.

If the catalyst temperature is less than T1 in the determination in S602, it takes time to reach the temperature at which S poisoning can be released even if the temperature increase is started, and there is a concern that the NOx in the tail pipe may deteriorate during the temperature increase. Therefore, it is desirable to give priority to NOx desorption purification. For this reason, it progresses to S605.
In S605, it is determined whether or not the rq-sp flag = 1, that is, whether or not a NOx desorption purification request has been issued, and if so, the process proceeds to S607, where the sp flag = 1 is set and NOx desorption purification (rich spike) is performed. Migrate to

If the rq-sp flag is not 1, the process proceeds to S606.
In S606, after the S poisoning release request is issued, it is determined by the same method as in S14 whether or not the NOx accumulation amount reaches the predetermined value NOx1 and the NOx desorption purification time is reached.
If NOx accumulation amount> NOx1, the process proceeds to S901 in the flow of FIG. 11, and the rq-sp flag = 1 is set.

Next, the control in the warm-up promotion mode of FIG. 12 will be described. This is performed when the catalyst temperature is T5 or less.
In S1001, it is determined whether or not it is a warm-up promoting operation possible region. Since the warm-up promoting operation here is performed by split retard combustion according to the present invention, it is determined whether or not this combustion is possible. Specifically, the region where DPF regeneration and S-poisoning can be canceled shown in FIG. 24 is set as a warm-up promotion operation possible region. In this region, the process proceeds to S1002.

In S1002, the engine combustion is switched from the normal lean combustion to the split retard combustion according to the present invention in order to promote warm-up. By switching to the main combustion, the exhaust temperature becomes high and the warm-up of the catalyst can be promoted.
Again, the target λ is set and controlled. As for the control to the target λ, the torque is reduced by the retarded combustion. Therefore, the control to the target λ is performed while performing the torque correction.

  In S1003, it is determined whether or not the catalyst temperature has become higher than its activation temperature T5, and if the catalyst temperature> T5, the process proceeds to S1004 to switch from split retard combustion according to the present invention to normal combustion. End warm-up promotion operation.

Engine system diagram showing an embodiment of the present invention Flow chart of exhaust purification control (part 1) Flow chart of exhaust purification control (part 2) Flow chart of exhaust purification control (part 3) Flow chart of exhaust purification control (part 4) Flow chart of exhaust purification control (part 5) Flow chart of exhaust purification control (Part 6) Flow chart of exhaust purification control (7) Flow chart of exhaust purification control (8) Flow chart of exhaust purification control (9) Flow chart of exhaust purification control (10) Flow chart of exhaust purification control (11) The combustion form of the reference example 1 is shown. The figure which shows the combustion form of the reference example 2 The figure which shows the combustion form of this invention Figure comparing the exhaust gas states of Reference Examples 1 and 2 and the present invention Diagram showing the relationship between main combustion timing and exhaust gas status Diagram showing target fuel injection timing for pre-combustion Diagram showing target fuel injection amount for pre-combustion Diagram showing target fuel injection timing for main combustion The figure which shows another combustion form of this invention Flow chart for switching to split retard combustion A diagram showing the relationship between the amount of accumulated PM and the target λ being regenerated The figure which shows the target intake air amount calculation map for (lambda) = 1 driving | operation The figure which shows the relationship between the main injection timing and the torque correction coefficient K1 The figure which shows the relationship between target (lambda) and the torque correction coefficient K2. The figure which shows the map for target intake air amount calculation for rich spike operation The figure which shows the possible area | region of DPF reproduction | regeneration and S poisoning cancellation | release

Explanation of symbols

1 Diesel engine
2 Air intake passage
5 Inlet throttle valve
9 Fuel injection valve
10 Exhaust passage
11 EGR passage
12 EGR valve
13 NOx trap catalyst
14 DPF
20 Control unit
21 Rotational speed sensor
22 Accelerator position sensor
23 Air Flow Meter
24 Water temperature sensor
25 Catalyst temperature sensor
26 Exhaust pressure sensor
27 DPF temperature sensor
28 Air-fuel ratio sensor

Claims (13)

In an internal combustion engine provided with an exhaust purification device in an exhaust passage,
Under a predetermined condition based on the state of the exhaust gas purification device, main combustion for generating main torque and at least one preliminary combustion performed prior to main combustion are performed, and at least one of the preliminary combustion is compressed While performing near the top dead center, and performing the combustion control to control the fuel injection to the engine so that the main combustion starts after the preliminary combustion ends,
The combustion control device for an internal combustion engine, wherein when switching the combustion control from the combustion performed before that, the switching of the main combustion is delayed after the switching of the preliminary combustion.   2. The combustion control apparatus for an internal combustion engine according to claim 1, wherein the main combustion switching is controlled so as to gradually approach a target value. 3.   3. The combustion control apparatus according to claim 2, wherein the fuel injection timing for main combustion is gradually retarded to a target injection timing after completion of switching.   2. The combustion control apparatus for an internal combustion engine according to claim 1, wherein the main combustion is switched after the fluctuation of the intake system at the time of switching to the combustion control is converged.   2. The internal combustion engine according to claim 1, wherein the fuel injection amount of the preliminary combustion is a fuel injection amount necessary for an in-cylinder temperature at the time of fuel injection of the main combustion to exceed a temperature at which self-ignition is possible. Combustion control device.   The internal combustion engine according to any one of claims 1 to 5, wherein the combustion start timing of the main combustion is a timing separated by 20 degrees or more in crank angle from the combustion start timing of the preliminary combustion. Combustion control device.   The combustion control device for an internal combustion engine according to any one of claims 1 to 6, wherein the end time of the main combustion is a time separated by 50 degrees or more in crank angle from the compression top dead center. .   The combustion control apparatus for an internal combustion engine according to any one of claims 1 to 7, wherein the main combustion controls an exhaust gas temperature by changing a fuel injection timing for the main combustion. . As an exhaust purification device, equipped with a filter to collect PM in the exhaust,
The predetermined condition based on the state of the exhaust gas purification device is at least the time of regeneration of a filter that raises the exhaust gas temperature and burns and removes PM accumulated on the filter. 6. The combustion control device for an internal combustion engine according to any one of 5 above. As an exhaust purification device, provided with a NOx trap catalyst for trapping NOx in exhaust when the exhaust air-fuel ratio is lean,
The predetermined condition based on the state of the exhaust purification device is a time when at least the exhaust air-fuel ratio is enriched to desorb and purify NOx trapped in the NOx trap catalyst. Item 10. The combustion control device for an internal combustion engine according to any one of Items 9 to 9. As an exhaust purification device, provided with a NOx trap catalyst for trapping NOx in exhaust when the exhaust air-fuel ratio is lean,
The predetermined condition based on the state of the exhaust gas purification device is a time when at least the exhaust gas temperature is raised to remove the poisoning of sulfur accumulated in the NOx trap catalyst. The combustion control device for an internal combustion engine according to claim 10.   The predetermined condition based on the state of the exhaust gas purification device is at least when the exhaust gas purification device is rapidly warmed up when the exhaust gas purification device is cold. A combustion control device for an internal combustion engine according to claim 1.   The combustion before switching to the combustion control is a combustion mainly based on diffusion combustion in which main injection of fuel is performed during combustion by pilot injection of fuel. A combustion control device for an internal combustion engine according to claim 1.

Images