JP2000130246A - Knock detecting method of internal combustion engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve knock detecting accuracy by setting a plurality of learning regions for learning a reference signal level according to an operating condition of an internal combustion engine, and detecting generation of knock on the basis of the reference signal level which is set in the learning region in an operating condition according to the set learning region. SOLUTION: It is calculated on the basis of an engine load by an electronic control device 6 whether a present operating condition is appropriate to whichever learning zones, and a back ground level in the learning zone is learnt and held. In one learning zone, a plurality of learning values are memorized for a set load at every engine rotating speed. A knock signal is A/D converted from ion current detected by a detecting circuit 25, a coefficient is multiplied by a learning value memorized in the learning zone, and a learning value is learned in the case where the knock signal is not superposed. A signal level detected on the basis of the learning value is judged, and generation of knock is judged in the case where the signal level exceeds a judged level.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a knock detection method for an internal combustion engine, which mainly detects the occurrence of knock in an automobile engine by passing an ionic current into a combustion chamber after ignition.

[0002]

2. Description of the Related Art Conventionally, as a device for detecting knock generated in an internal combustion engine, a device using a vibration type knock sensor and detecting the occurrence of knock based on a signal output from the knock sensor is known. Are known. Such a knock detection device is configured to extract a signal in a predetermined frequency band from an output signal from a knock sensor and process the extracted signal to detect knock.

As a method of detecting knock using an ion current, for example, as disclosed in Japanese Patent Application Laid-Open No. 58-7536, the knock and the knock are detected based on the amplitude and width of an ion signal corresponding to the detected ion current. A device that detects (determines) knock is known. Also in this example, in order to prevent the spark noise from being superimposed on the ion signal, the detection of the ion signal is performed after a predetermined time delay from the ignition. And even in the case of the knock detection method using such an ion current, similar to the case of the knock sensor,
There is known an ion current in which a signal in a predetermined frequency band (knock frequency component) is peak-held and subjected to signal processing.

[0004]

By the way, in the knock detection method using the vibration type knock sensor, an output signal of the knock sensor is measured while the engine is rotating, and the average value of the measured output signals is knocked. There is known a method of setting a reference level for determination, comparing the reference level with the signal level of an output signal measured after ignition, and detecting occurrence of knock based on the comparison result. Such a method for detecting knock has been attempted in a method for detecting knock using ion current.

In this case, the reference level is set by storing the signal level of the ion current on which the knock signal due to knock is not superimposed. Generally, it is known that the signal level of the ion current on which the knock signal is not superimposed differs depending on the operating state, for example, the engine speed, the size of the load, and the like. Therefore, in order to set the reference level, the signal level of the ion current under a certain operating condition of the engine is stored, but when the operating condition becomes different from the operating condition at that time, the setting under the different operating condition is performed. Since a deviation from the reference level to be performed has occurred, the occurrence of knock may be erroneously detected or may not be detected.

An object of the present invention is to solve such a problem.

[0007]

In order to achieve the above object, the present invention takes the following measures. That is, the knock detection method for an internal combustion engine according to the present invention focuses on the fact that the reference signal level for knock detection differs depending on the operating conditions of the internal combustion engine, and adjusts the reference signal level in accordance with the operating conditions of the internal combustion engine. A plurality of learning regions for learning are set, and in a driving state corresponding to the set learning region, occurrence of knock is detected based on a reference signal level set in the learning region.

[0008]

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an internal combustion engine for measuring a signal level of an ionic current generated at each ignition in a combustion chamber of the internal combustion engine and detecting knock generated in the internal combustion engine based on the measured signal level. A knock detection method, wherein a plurality of learning regions for learning a reference signal level for detecting knock are set based on the operating state of the internal combustion engine, and when the operating state of the internal combustion engine corresponds to the set learning region. The reference signal level in the learning area is learned based on the signal level of the ion current on which the knock signal is not superimposed, and the signal level of the ion current is compared with the reference signal level in the operating state of the internal combustion engine corresponding to the learning area. A knock detection method for an internal combustion engine, characterized by detecting occurrence of knock.

With such a configuration, even if the operating state of the internal combustion engine changes, the occurrence of knock is detected using the reference signal level set in the learning region corresponding to the changed operating state. This makes it possible to prevent the occurrence of knocking from being detected despite the occurrence of knocking, or the erroneous detection that knocking has occurred when knocking has not occurred.

Note that the reference signal level learning region may be set, for example, using at least the rotation speed of the internal combustion engine as a parameter. Further, the parameter may be set as a parameter in combination with the magnitude of the load. Further, the reference signal level is obtained by learning a value under a central driving condition of each learning region and performing an interpolation calculation based on the central driving condition when other driving conditions are used. May be.

[0011]

An embodiment of the present invention will be described below with reference to the drawings. The engine 100 schematically shown in FIG. 1 is of a three-cylinder type for an automobile, and has a throttle valve 2 that opens and closes in response to an accelerator pedal (not shown) in an intake system 1 thereof.
The surge dunk 3 is provided on the downstream side. In the vicinity of one end communicating with the surge tank 3, a fuel injection valve 5 is further provided. The fuel injection valve 5 is controlled by the electronic control unit 6 so as to be independently injected for each cylinder. ing. A spark plug 18 that generates sparks and serves as an electrode for ion current is attached to a cylinder head 31 that forms the combustion chamber 30. Further, an O ₂ sensor 21 for measuring the oxygen concentration in the exhaust gas is attached to the exhaust system 20 at a position upstream of a three-dimensional catalyst 22 disposed in a pipe leading to a muffler (not shown). I have.

The electronic control unit 6 includes a central processing unit 7
, A storage device 8, an input interface 9, an output interface 11, and an A / D converter 10. The input interface 9 has an intake pressure signal a output from an intake pressure sensor 13 for detecting the pressure in the surge tank 3 and a cylinder output from a cam position sensor 14 for detecting the rotation state of the engine 100. Vehicle speed sensor 1 for detecting discrimination signal G1, crank angle reference position signal G2, engine speed signal b, and vehicle speed
5, an IDL signal d output from an idle switch 16 for detecting the open / close state of the throttle valve 2, and a water temperature signal e output from a water temperature sensor 17 for detecting the engine coolant temperature. , The current signal h output from the O ₂ sensor 21 and the like are input. On the other hand, the output interface 11 outputs a fuel injection signal f to the fuel injection valve 5 and an ignition pulse g to the spark plug 18. The spark plug 18 is connected via a high voltage diode 23 to a bias power supply 24 for measuring an ion current. This bias power supply 24
Various methods known in the art can be used for the circuit 25 for measuring the ion current and the measurement method itself. In this embodiment, the ion current can be detected for each cylinder, so that knock can be detected for each cylinder.

The electronic control unit 6 includes an intake pressure signal a output from the intake pressure sensor 13 and a cam position sensor 14.
The basic injection time is corrected by various correction coefficients determined according to the operating state of the engine 100 to determine the fuel injection valve opening time, that is, the injector final energization time T, with the rotation speed signal b output from the main information as main information. A program for controlling the fuel injection valve 5 based on the determined energizing time and injecting fuel corresponding to the engine load into the intake system 1 is built in. Further, the electronic control unit 6 is programmed so as to control the operation of the engine 100 and detect the knock by detecting the ionic current flowing for each ignition. That is, in order to detect knock, the signal level of the ionic current generated for each ignition in the combustion chamber 30 of the engine 100 is measured, and the knock generated in the engine 100 is detected based on the measured signal level. A plurality of learning regions for learning a reference signal level for detecting knock are set based on the operating state of the engine 100, and a knock signal is superimposed when the operating state of the engine 100 corresponds to the set learning region. The reference signal level in the learning region is learned based on the signal level of the non-existing ion current, and the occurrence of knock is detected by comparing the signal level of the ion current with the reference signal level in the operating state of the engine 100 corresponding to the learning region. The program to be executed is built in the electronic control unit 6.

An outline of the knock detection program based on the ion current is as shown in FIG. In this embodiment, the learning zone, which is the learning area, is set according to the engine speed and the magnitude of the load, as shown in FIG. Specifically, three regions are set according to the level of the engine speed, and two regions are set according to the magnitude of the load. Therefore, six learning zones are set as learning zones according to the combination of the engine speed and the load. These learning zones are stored in the storage device 8 of the electronic control unit 6, and a background level, which is a reference signal level described in detail below, is stored for each learning zone.

In step S1, a learning zone to which the current operating state corresponds is calculated based on the engine speed and the intake pressure that changes depending on the magnitude of the load. When the learning zone is determined, in step S2, learning of the background level in the learning zone is performed, and the learning value KNKKG is stored. Learning value KNKKG
Are stored for each set engine load for each engine speed in one learning zone. Learning value KN
KKG learns, for example, as follows.

In learning the learning value KNKKG, first,
A / D conversion of the knock signal is performed from the detected ion current, and the conversion result is set as a signal level NIONKNK.
Since it is known that the knock signal generally occurs near the time when the combustion starts and the maximum combustion pressure is passed, a period including before and after the time is set, and the knock signal is obtained from the signal level of the ion current in the period. It is to be detected. Next, the learning value KNKK stored at this time in the learning zone
G is multiplied by a coefficient to determine whether or not the signal level is such that the knock signal is not superimposed. As a result of the determination, when the knock signal is not superimposed, the signal level NION obtained this time is obtained.
KNK is learned as a background level learning value KNKKG.

In such a configuration, the occurrence of knock is determined by the learning value KNK of the learning zone corresponding to a certain driving state.
Based on KG, the signal level NIONKNK detected in this operation state is determined and detected. That is, first, a corresponding learning zone is determined based on the engine speed and the magnitude of the load in the operating state at this time. Next, the learning value KNKKG of the corresponding learning zone is read, and the learning value KNKKG is multiplied by a coefficient for knock determination to calculate a knock determination level. If the detected signal level NIONKNK exceeds the calculated knock determination level, it is determined that knock has occurred.

As described above, the learning value KNKKG is learned and updated in each of the learning zones corresponding to the operating state even when the operating state of the engine 100 changes suddenly, so that the knock determination level can be calculated quickly. The knock determination level for knock determination can be appropriately changed according to the driving state. Further, since the knock determination level corresponds to the driving state, it is possible to prevent detection of knock that has not occurred, and also to prevent detection of occurrence of knock by mistake.

The present invention is not limited to the embodiment described above. In each learning zone, the learning value KNKKG learns a value of 1 for the engine speed and load at the center of the learning zone, and the engine speed and engine speed other than the center engine speed and load at the same learning zone. The background level for the load may be obtained by interpolation calculation.

More specifically, as shown in FIG. 4, it is determined whether or not knock has occurred in each learning zone (step S11), and a signal of an ion current corresponding to a knock signal when there is no knock is determined. Average value KNKA from level
V is calculated (step S12). This average value KNKA
V is obtained by averaging the signal levels NIONKNK detected in a plurality of operating states included in the learning zone. Next, the learning value KNKKG learned last time
and a _n-1 and the current obtained average value KNKAV, computed by the following equation the current learning value KNKKG _n (step S1
3).

KNKKG _n = KNKKG _n-1- (KNK
KG _n−1 −KNKAV) / N Therefore, learning of the learning value KNKKG is performed by the smoothing process according to the above equation, based on the stored learning value KNKKG _{n−1 and} the magnitude of the average value KNKAV obtained this time. It will be increased or decreased. On the other hand, when the occurrence of knock is detected, as shown in FIG. 5, interpolation is calculated from the learning value KNKKG based on the engine speed and the magnitude of the load (step S21). The learned value of the ground level (step S22) is used for knock determination.

Therefore, the background level can be changed without abrupt change for a plurality of operating states corresponding to the same learning zone. Also,
As described above, in the case of detecting the occurrence of knock based on the learning value of the background level obtained by interpolation calculation from one learning value, the data amount of the learning value can be reduced. Can be reduced. Therefore, the storage device 8 can be used efficiently.

In addition, the configuration of each section is not limited to the illustrated example, and various modifications can be made without departing from the spirit of the present invention.

[0024]

As described above, according to the present invention, when the knock is detected based on the ion current, even if the operating state of the internal combustion engine changes, the learning area corresponding to the changed operating state is set. Knock is detected using the reference signal level that was set, so that knock could not be detected even though knock had occurred, or it was erroneously determined that knock had occurred when knock had not occurred. Detection can be prevented. As a result, knock detection can be performed under substantially the same conditions irrespective of the presence or absence of a signal level change, and knock detection accuracy and reliability can be improved.

[Brief description of the drawings]

FIG. 1 is a schematic structural explanatory view showing one embodiment of the present invention.

FIG. 2 is a flowchart showing a schematic control procedure of the embodiment.

FIG. 3 is a configuration explanatory view showing a schematic configuration of a learning zone according to the present invention.

FIG. 4 is a flowchart showing a schematic control procedure of a routine for calculating a learning value according to another embodiment of the present invention.

FIG. 5 is a flowchart illustrating a schematic control procedure for determining a learning value according to another embodiment.

[Explanation of symbols]

 6 Electronic control unit 7 Central processing unit 8 Storage unit 9 Input interface 10 A / D converter 11 Output interface 30 Combustion chamber

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Teruhito Ichihara 2-1-1, Taoyuan, Ikeda-shi, Osaka Daihatsu Kogyo Co., Ltd. F-term (reference) 2G087 AA01 BB12 CC35 FF23 3G019 AA05 DB02 DB07 DB14 DC07 FA05 GA00 GA05 GA08 GA14 3G084 AA03 CA03 CA04 CA09 DA04 DA27 EA05 EA11 EB08 EB18 EB20 EB22 EB25 FA00 FA11 FA25 FA33 FA35

Claims (1)

[Claims] 1. A knock detection method for an internal combustion engine, comprising: measuring a signal level of an ion current generated at each ignition in a combustion chamber of the internal combustion engine; and detecting knock generated in the internal combustion engine based on the measured signal level. A plurality of learning regions for learning a reference signal level for detecting knock are set based on the operating state of the internal combustion engine, and a knock signal is superimposed when the operating state of the internal combustion engine corresponds to the set learning region. The reference signal level in the learning area is learned based on the signal level of the ion current that has not been detected, and the knock level is generated by comparing the signal level of the ion current with the reference signal level in the operating state of the internal combustion engine corresponding to the learning area. A knock detection method for an internal combustion engine, wherein the knock is detected.