JP6089624B2 - Control device and control method for internal combustion engine - Google Patents

Description

  The present invention detects an occurrence or a sign of pre-ignition, which is one of abnormal combustion in an internal combustion engine, and executes a predetermined pre-ignition avoidance control in response to the detection. Regarding the method.

  In a spark ignition type internal combustion engine, if so-called pre-ignition occurs in which the air-fuel mixture in the combustion chamber ignites before ignition by the spark plug, the durability and performance of the internal combustion engine are significantly lowered, which is not preferable.

  In Patent Document 1, in addition to detecting the occurrence of pre-ignition from the ion current in the combustion chamber or the like, a pre-ignition sign that occurs before the occurrence of pre-ignition is detected. A technology is disclosed in which the temperature in the combustion chamber is lowered by increasing the fuel injection amount, retarding the ignition timing, and the like, thereby preventing the occurrence of pre-ignition.

  In Patent Document 2, the occurrence of pre-ignition is detected from the in-cylinder pressure change detected by the in-cylinder pressure sensor, and when pre-ignition occurs, the fuel injection timing into the cylinder is retarded to reduce the combustion pressure. A technique for suppressing pre-ignition is disclosed. Here, the retard correction amount of the injection timing with respect to the basic injection timing includes a correction amount that gradually increases by a predetermined amount every cycle from the occurrence of pre-ignition until the pre-ignition is not detected, the engine speed and the suction. Correction amount learning values respectively assigned to learning regions determined by the air amount. The correction amount learning value is sequentially learned and updated based on the injection timing at the stage where the pre-ignition actually occurs when the injection timing is gradually advanced in the state where the pre-ignition has not occurred. .

JP 2006-46140 A JP 2010-281300 A

  In Patent Document 1, an attempt is made to avoid the occurrence of actual pre-ignition based on the detection of the sign. However, for example, the operating conditions in which pre-ignition occurs relatively easily, such as when the intake air temperature or the cooling water temperature of the internal combustion engine is high, are described. Below, it is easy to shift from the normal combustion state to the pre-ignition occurrence state immediately, and it is difficult to avoid the occurrence of pre-ignition in terms of time before the occurrence of the pre-ignition is detected.

  In other words, in the situation where the transition period from the normal combustion state to the occurrence of pre-ignition is very short, it is difficult to detect the sign, and even if the sign can be detected, the fuel injection amount is increased or the ignition timing is retarded. The pre-ignition avoidance due to the above is not in time. Eventually, after the pre-ignition occurs, the subsequent pre-ignition is merely suppressed by the increase in the fuel injection amount or the retard of the ignition timing.

  Further, Patent Document 2 allows the actual occurrence of pre-ignition itself, and uses the correction amount learning value to retard the injection timing relatively early from the beginning, thereby achieving early elimination of pre-ignition. However, it is not possible to prevent the occurrence of pre-ignition. That is, the learning value of Patent Document 2 is a learning value of a correction amount given when pre-ignition occurs, and does not contribute to avoiding the occurrence of pre-ignition.

  An object of the present invention is to provide a control device and control method for an internal combustion engine that suppresses the occurrence of pre-ignition in a condition where pre-ignition is likely to occur.

  In one aspect of the control device or control method for an internal combustion engine according to the present invention, the pre-ignition in the combustion chamber, the occurrence of pre-ignition, the pre-ignition that occurs in the combustion chamber, and the occurrence of these pre-ignition At least three states of normal combustion in which no sign is detected are identified, and when the occurrence of pre-ignition is detected, predetermined pre-ignition avoidance control is executed.

  Here, in the present invention, a plurality of learning regions are set corresponding to at least the driving conditions that may cause pre-ignition.

Then, after the pre-ignition avoidance control is executed, when the normal combustion is restored, the control amount of the pre-ignition avoidance control is gradually decreased, and the control amount immediately before the pre-ignition sign is detected is determined. Learning is performed as a learning value corresponding to the learning region at that time. Thereafter, in the learning area, a control amount based on the learning value is given in advance. In addition, the same learning value is reflected in one or a plurality of learning areas that are determined in advance as learning areas that are more likely to cause pre-ignition than the learning area.

  The pre-ignition avoidance control includes the reduction of the effective compression ratio of the internal combustion engine, the suppression of the intake air amount, the increase of the fuel injection amount, the retardation of the injection timing of the fuel injection into the cylinder, and the increase of the rotational speed of the internal combustion engine. However, the present invention is not limited thereto, and known means that contribute to suppression of pre-ignition can be used.

  In the second aspect of the present invention, the control device or control method for an internal combustion engine monitors a pre-ignition sign generated in the combustion chamber, and detects a pre-ignition sign when the pre-ignition sign is detected. Execute.

  Here, in the present invention, a plurality of learning regions are set corresponding to at least the driving conditions that may cause pre-ignition.

Then, the control amount of the pre-ignition avoidance control is gradually increased with respect to the detection of the sign, and the control amount when returning to normal combustion is learned as a learning value corresponding to the learning region at that time. Thereafter, in the learning area, a control amount based on the learning value is given in advance. In addition, the same learning value is reflected in one or a plurality of learning areas that are determined in advance as learning areas that are more likely to cause pre-ignition than the learning area.

  That is, in any aspect, the occurrence of pre-ignition during driving in a certain learning region or the increase or decrease of the control amount of the pre-ignition avoidance control accompanying the detection of the pre-sign is minimal because no pre-pre-ignition sign occurs. A limited control amount is obtained as a learning value in the learning region. Then, after the learning value is given in this way, a control amount based on the learning value is given regardless of whether or not there is an actual pre-ignition or its sign when driving in the same learning region.

  For example, if the pre-ignition avoidance control is a decrease in the effective compression ratio, when an occurrence or a sign of pre-ignition is actually detected during driving in a certain learning region, in the process of pre-ignition avoidance, A learning value indicating a control amount that provides an effective compression ratio that does not cause a sign of pre-ignition in the learning region is obtained. Thereafter, in the learning area, even if there is no sign of pre-ignition, the effective compression ratio is limited to the learning value. Therefore, under the driving conditions where the occurrence of pre-ignition is a concern, the margin until the occurrence of pre-ignition is expanded, and the occurrence of pre-ignition is more reliably prevented.

  According to the present invention, for each learning region corresponding to the driving condition in which the occurrence of pre-ignition is concerned, a control amount that does not cause a pre-ignition sign is given in advance based on the learning value. Occurrence can be avoided in advance.

The system block diagram of the internal combustion engine of one Example to which this invention is applied. The flowchart which shows the first half of the learning process of one Example. The flowchart which shows the latter half part of a learning process. The time chart which shows the effect | action of an Example when generation | occurrence | production of pre-ignition is detected. The time chart which shows the effect | action of an Example when the sign of pre-ignition is detected. Explanatory drawing of a learning area | region. Explanatory drawing of the addition / extension process of a learning area. The flowchart which shows the principal part of 2nd Example of a learning process. The time chart which shows the effect | action of a 2nd Example when generation | occurrence | production of pre-ignition is detected.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows the system configuration of an automotive internal combustion engine 1 to which the present invention is applied. The internal combustion engine 1 is an in-cylinder direct injection type spark ignition internal combustion engine, and includes a fuel injection valve 2 for injecting fuel into the cylinder for each cylinder, and the generated air-fuel mixture. Is equipped with a spark plug 3 for igniting. The spark plug 3 is provided with an ion current sensor 4 for measuring an ion current in the combustion chamber as a pre-ignition detection means. The ion current sensor 4 uses the spark plug 3 itself as a detection probe. However, an ion current sensor independent of the spark plug 3 may be used.

  Each cylinder includes an intake valve 5 and an exhaust valve 7. The intake valve 5 includes a variable valve device 6 that can variably control the opening / closing timing (at least the closing timing) of the intake valve 5. The valve 7 includes a variable valve device 8 that can variably control the opening / closing timing of the exhaust valve 7. These variable valve gears 6 and 8 are controlled by the engine controller 10.

  An electronically controlled throttle valve 12 whose opening degree is controlled by a control signal from the engine controller 10 is interposed upstream of the intake passage 11, and an exhaust recirculation passage 14 extending from the exhaust passage 13 to the intake passage 11 is provided. Is provided with an exhaust gas recirculation control valve 15.

  In addition to the ion current sensor 4 described above, the engine controller 10 includes a crank angle sensor 16 for detecting the engine rotational speed, an air flow meter 17 for detecting the intake air amount, an intake air temperature sensor 18 for detecting the intake air temperature, and a cooling function. Detection signals of sensors such as a water temperature sensor 19 that detects a water temperature, an oil temperature sensor 20 that detects a lubricating oil temperature, and an accelerator opening sensor 21 that detects a depression amount of an accelerator pedal operated by a driver are input. Yes. Based on these detection signals, the engine controller 10 determines the fuel injection amount and injection timing by the fuel injection valve 2, the ignition timing by the spark plug 3, the opening and closing timing of the intake valve 5 and the exhaust valve 7, the opening of the throttle valve 12, and the like. Is optimally controlled.

  In the illustrated example, the output shaft of the internal combustion engine 1 is configured to drive the drive wheels 27 via a torque converter 25 and a belt type continuously variable transmission (CVT) 26. The transmission ratio of the continuously variable transmission 26 is controlled by the CVT controller 28. The CVT controller 28 is connected to the engine controller 10 via CAN communication.

  In the above configuration, based on the ion current measured by the ion current sensor 4, a pre-ignition occurrence state in which pre-ignition has actually occurred, and a pre-prediction in which pre-ignition has not occurred yet have occurred. It is possible to discriminate between three states: an ignition sign state and a normal combustion state where there is no pre-ignition or its sign. The ionic current generated in the combustion chamber correlates with the combustion state or combustion pressure in the combustion chamber.For example, the occurrence of pre-ignition can be detected from the generation start time or peak time, and before the actual pre-ignition occurs. , You can grasp the signs.

  On the other hand, when the occurrence of a pre-ignition or a sign thereof is detected, predetermined pre-ignition avoidance control is executed. In the above configuration, pre-ignition avoidance control includes, for example, a reduction in the effective compression ratio by correction of the retard of the intake valve closing timing, which is generally after bottom dead center, a reduction in the intake amount by correction of the throttle valve opening reduction, and injection into the cylinder. It is possible to reduce the in-cylinder temperature by correcting the increase in the amount of fuel that is generated. The occurrence of pre-ignition greatly depends on the in-cylinder temperature and the time during which it is exposed to the in-cylinder temperature (that is, the engine rotation speed). The lower the rotation speed and the higher the load, the easier the occurrence of pre-ignition. The higher the temperature (cooling water temperature, oil temperature, etc.) and intake air temperature, the more likely it is.

  2 and 3 are flowcharts showing learning value learning processing, which is the main part of the present invention. Here, reduction correction of the throttle valve opening is used as preignition avoidance control.

  In the present embodiment, the plurality of learning regions are set by four parameters: the rotational speed of the internal combustion engine, the load of the internal combustion engine (accelerator pedal depression amount), the coolant temperature, and the intake air temperature. Specifically, as schematically shown in FIG. 6, a learning value map in which a learning region is divided with respect to the engine rotation speed and the load is provided for each temperature condition (combination of cooling water temperature and intake air temperature). For example, a learning value map is provided for each combination of “the cooling water temperature Tw is Tw1 ≧ Tw> Tw2 and the intake air temperature Ta is Ta1 ≧ Ta> Ta2”. In the initial state, each region of these learning value maps is not learned and is blank. Of course, it is possible to use the lubricating oil temperature instead of the cooling water temperature as a representative of the engine temperature.

  The processing shown in FIG. 2 and FIG. 3 is repeatedly executed during operation of the internal combustion engine. First, in step 1, it is determined whether or not a pre-ignition sign is detected. For example, in step 2, it is determined whether or not the actual occurrence of pre-ignition is detected. If “NO” in the step 2, the key OFF determination in the step 25 of FIG. 3 is performed, and the series of routines are ended. In other words, while operating in a normal combustion state, pre-ignition signs and occurrence of pre-ignition are continuously monitored.

  As will be described later, when a learning value has already been written in the learning region corresponding to the driving condition at this time (that is, when the learning region has already been learned), the throttle valve is moved along this learning value. The operation is performed with the opening degree corrected for reduction.

  For example, when a pre-ignition occurs suddenly when the intake air temperature is high, for example, the detection of the sign and the avoidance of the pre-ignition based on the detection of the sign may not be substantially in time, and the pre-ignition may occur. When the occurrence of pre-ignition is detected in step 2, the process proceeds to step 3 where throttle control is performed by adding a predetermined control amount (opening reduction correction amount of a predetermined magnitude) to the normal target opening as pre-ignition avoidance control. Close the valve opening relatively large. The corrected target opening at this time is a throttle valve opening that is an intake air amount that can avoid pre-ignition, and may be a predetermined fixed value, or may be variable according to operating conditions. You may make it set.

  After the throttle valve opening is corrected for reduction in this way, the routine proceeds to step 4 where it is repeatedly determined whether a normal combustion state has been reached. If it is detected that the normal combustion state has been achieved by the reduction correction of the throttle valve opening, the routine proceeds to step 5 where the throttle valve opening is increased by a predetermined minute amount in order to learn the learning value. In other words, the control amount for avoiding pre-ignition (throttle valve opening decrease correction amount) is decreased by a certain amount. In step 6, it is determined whether or not a pre-ignition sign has been detected, and steps 5 and 6 are repeated until a pre-ignition sign is detected. Therefore, the throttle valve opening gradually increases (the control amount for avoiding preignition gradually decreases) until a sign is detected.

  If a pre-ignition sign is detected in step 6, the process proceeds to step 7, and the throttle valve opening (that is, the previous value) immediately before the sign is detected is stored as a learning value in the learning region.

  On the other hand, in a situation where pre-ignition is about to occur relatively slowly, the sign can be detected before the actual occurrence of pre-ignition. When a sign is detected in step 1, the process proceeds to step 8 where the throttle valve opening is decreased by a predetermined minute amount as preignition avoidance control (in this case, it also serves as learning control of the learning value). In other words, the control amount for avoiding pre-ignition (throttle valve opening decrease correction amount) is increased by a certain amount. Then, in step 9, it is determined whether or not the pre-ignition sign has been resolved and the normal combustion state has been reached, and steps 8 and 9 are repeated until it is detected that the normal combustion state is present. Accordingly, the throttle valve opening gradually decreases (the control amount for avoiding preignition gradually increases) until the normal combustion state is reached.

  When a normal combustion state is detected in step 9, the process proceeds to step 10 to determine whether or not a learning value already exists in the learning region (that is, has been learned). If it has already been learned, the learning value is not updated and the process proceeds to the key OFF determination in step 25 of FIG. If it is determined in step 10 that the engine has not been learned, the process proceeds to step 11, and the throttle valve opening at that time (that is, the throttle valve opening when the normal combustion state is restored) is used as a learning value in the learning region. save.

  FIG. 3 is a flowchart relating to the learning value writing process after obtaining the learning value in Step 7 or Step 11 above. In Step 21, four parameters for specifying the learning region when the learning value is obtained; That is, the rotational speed, load (accelerator pedal depression amount), cooling water temperature, and intake air temperature are read.

  Next, in step 22, it is determined whether or not the driving condition specified by these four parameters is outside the existing learning region in the learning value map. If the operation condition is included in the existing learning region, the process proceeds from step 22 to step 23, and the step is performed in the learning region corresponding to the parameters (rotation speed, load, cooling water temperature, intake air temperature) read in step 21. 7 or the learning value obtained in step 11 is written.

  Further, the process proceeds to step 24, where the same learning value is written if there is an unlearned area in a learning area where the possibility of occurrence of pre-ignition is higher than that in the learning area. Specifically, with respect to the rotational speed Ne0, load TH0, cooling water temperature Tw0, and intake air temperature Ta0 when actual learning is performed, the rotational speed is Ne0 or lower, the load is TH0 or higher, the cooling water temperature is Tw0 or higher, and the intake air temperature. The same learning value is stored in an unlearned learning region that satisfies the four conditions that is equal to or higher than Ta0. FIG. 6 illustrates an example of this processing. For example, if a learning value is obtained in the region 101 in a learning value map under a certain temperature condition, the rotational speed is lower and the load is higher than this. The same learning value is reflected in all of the regions 102 to 124.

  In one embodiment, as described above, the learning value is written only in the unlearned area among the areas 102 to 124. However, as another embodiment, the existing value is written in the already learned area. A comparison is made between the learning value and the new learning value, and if a new learning value is obtained in a direction that avoids pre-ignition more (a smaller opening if the throttle valve is open), it is updated. You may do it.

  Then, the process proceeds to step 25, where it is determined whether or not the key is OFF. If the key is not OFF, the series of routines described above are repeated. When the key is OFF, the process proceeds to step 26, and all the learning values written in each learning area are reset.

  Therefore, in one trip from the key ON to the key OFF, there is no learning value in any learning region at the initial stage when the key is ON, and the throttle valve opening is not particularly limited. When preignition itself or a sign thereof is detected during operation under certain operating conditions, preignition avoidance control for reducing the throttle valve opening is executed as described above, and the learning value is obtained during the process. Is set. As described above, this learning value is reflected not only in the learning area but also in other learning areas having a higher possibility of pre-ignition. In the state where the learning value is written in a part of the learning region in this way, thereafter, when the operating condition enters the learning region, the restriction of the throttle valve opening according to the learning value is executed in advance. The Therefore, in the learning area where the occurrence or sign of pre-ignition has been detected (and the learning area where the possibility of pre-ignition is higher), the pre-ignition is less likely to occur thereafter, and the occurrence of pre-ignition has not occurred. To be prevented. When the key is turned off and one trip is completed, the learning value is reset. Therefore, the next trip does not limit the throttle valve opening unless an actual pre-ignition or sign is detected. It is not done as necessary.

  On the other hand, if it is determined in step 22 that the driving condition for obtaining the learning value is outside the existing learning area in the learning value map, the process proceeds to step 27 to add a new learning area. That is, as shown in FIG. 7, in the initial state, no learning area is set in the entire learning value map, and one of the rotational speed and the load is illustrated as illustrated as “initially set learning area”. A plurality of learning areas are set in the part range (partial area on the low speed and high load side where the possibility of pre-ignition is high). This is for simplifying control and not limiting unnecessary throttle valve opening in an area where the possibility of pre-ignition is low. However, an area other than the learning area prepared in advance (for example, FIG. 7). When the pre-ignition (or a sign thereof) is detected in the area 201), the area 201 is newly added as a learning area in order to avoid the recurrence of pre-ignition thereafter.

  Therefore, in step 23 following step 27, the learning value is written in the new learning area added as described above.

  Note that the learning area information added in step 27 is stored even after the key is turned off in step 25. However, the value of the learning value written in the learning area is cleared in step 26 when the key is turned off as described above. In other words, the addition / expansion of the area (for example, addition of the area 201) executed because there is a possibility of pre-ignition is also effective for the next trip.

  In addition to the region corresponding to the operation condition at that time (for example, the region 201 in FIG. 7), in addition to the region corresponding to the operation condition at that time (for example, the region 201 in FIG. 7), the region where the possibility of pre-ignition is higher. (For example, the areas 202 and 203 in FIG. 7) may be added as a new learning area. In such a case, in step 24, for example, the learning value learned for the region 201 is reflected in the other regions 201, 203, and the like.

  Next, FIG. 4 is a time chart for explaining the action when pre-ignition occurs mainly corresponding to steps 2 to 7 in FIG. The upper part of the figure shows the ion current detection timing after the ignition timing as one parameter indicating the occurrence or sign of pre-ignition, which is earlier when pre-ignition occurs. Also, before actual occurrence of pre-ignition, it tends to be slightly earlier as a precursor. Therefore, it is possible to identify the occurrence of a pre-ignition, a sign, and normal combustion based on the ion current detection timing. In the illustrated example, for easy understanding, the ion current detection timing is simply divided into the normal combustion region, the pre-ignition sign region, and the pre-ignition occurrence region. In the example of the time chart, as the preignition avoidance control, instead of the above-described reduction correction of the throttle valve opening, the delay correction of the closing timing (IVC) of the intake valve 5 via the variable valve gear 6 is performed. The lower part of FIG. 4 shows the retardation correction amount as the control amount. Here, it is assumed that the normal intake valve closing timing is after bottom dead center.

  In the example of FIG. 4, preignition has occurred relatively suddenly from the normal combustion state, and therefore the detection of the sign is not in time, and the occurrence of preignition is detected at time t1. Along with this, a predetermined amount of control is given as preignition avoidance control, that is, the intake valve closing timing is retarded by a predetermined amount. The target intake valve closing timing at this time is retarded to a level sufficient to avoid pre-ignition. Since the speed of change of the mechanical variable valve mechanism 6 is limited, in FIG. 4, after time t1, the variable valve mechanism 6 operates at the maximum speed toward the target intake valve closing timing with the retarded angle corrected. The state is shown. Since the effective compression ratio decreases as the intake valve is closed, the pre-ignition is suppressed and the normal combustion state is restored.

  Along with the detection of this normal combustion, the intake valve closing timing is thereafter gradually advanced. It should be noted that the advance speed at this time is slower than the retard speed when the pre-ignition occurs, as shown in the figure. Thus, when the intake valve closing timing is gradually advanced, a pre-ignition sign is detected. In the illustrated example, a sign is detected at time t2, and the control amount immediately before this time is stored as a learned value. Therefore, as is clear from FIG. 4, the learning value corresponds to the minimum amount of retardation correction necessary for preventing the occurrence of pre-ignition under operating conditions including the temperature condition at that time.

  Thereafter, in the same learning region, the vehicle is operated in a state in which the retardation correction based on the learning value is added regardless of the occurrence of a pre-ignition or the presence or absence of a sign. Accordingly, the margin until the occurrence of pre-ignition is increased, and the occurrence of pre-ignition is prevented more reliably and in advance. Even if pre-ignition is likely to occur again, since the effective compression ratio has been lowered in advance, the occurrence of pre-ignition becomes slow and can be avoided at the sign stage.

  Next, FIG. 5 is a time chart for explaining the operation at the time of detecting a pre-ignition sign mainly corresponding to steps 1 and 8 to 11 in FIG. In this example, since the pre-ignition is about to occur relatively slowly from the normal combustion state, a pre-ignition sign is detected at time t1. Accordingly, the intake valve closing timing is gradually retarded as preignition avoidance control. Note that the retarded angular velocity at this time is appropriately set so as not to shift from the pre-ignition sign to the actual pre-ignition occurrence.

  The pre-ignition is suppressed by correcting the delay of the intake valve closing timing, and eventually returns from the predictive state to the normal combustion state. In the illustrated example, the transition to normal combustion is detected at time t2, and the control amount at this time is stored as a learned value. Accordingly, in this case as well, the obtained learning value corresponds to the minimum amount of retardation correction necessary for preventing the occurrence of pre-ignition under the operating conditions including the temperature condition at that time.

  Next, based on FIG. 8 and FIG. 9, the case where the driving condition deviates from the predetermined learning region range (the entire learning region range described in FIG. 7) during the learning value learning process is considered. A second embodiment will be described.

  FIG. 8 is a flowchart showing the main part of the learning process in the second embodiment. This is executed during the repeated processing of Steps 5 and 6 in FIG. 2 and during the repeated processing of Steps 8 and 9, respectively. In Step 31, the operating condition is the range of the entire learning region. If NO, the process of step 5, 6 or step 8, 9 is continued. In step 31, when the driving condition deviates from the entire learning area, the process proceeds to step 32, and the control amount value just before that and the corresponding learning area (individual learning areas such as area 101) are obtained. Memorize temporarily. Then, in step 33, it is repeatedly determined whether or not the operating condition has returned to the learning area in the middle of the learning process (stored learning area). If YES, the process proceeds to step 34 and the control amount stored in step 32 is determined. Set. Thereby, the process of step 5, 6 or step 8, 9 is restarted using the value before interruption.

  FIG. 9 is a time chart for explaining such processing. Like the example of FIG. 4, the occurrence of pre-ignition is detected, and pre-ignition avoidance control (here, the correction of the retarded timing of the intake valve closing timing) is performed. ) Is executed and the normal combustion state is restored, and the intake valve closing timing is gradually advanced as a learning process, but the process is interrupted due to a sudden change in operating conditions at time t3. Such a change in operating conditions occurs, for example, when the accelerator pedal opening is turned off or when a lockup mechanism (not shown) in the torque converter 25 is unlocked. When the same learning area is entered again at time t4, the learning process is resumed using the value of the control amount at time t3. Note that between the time t3 and the time t4, both the upper-stage ion current detection timing and the lower-stage intake valve closing timing are indefinite, and are not shown in the figure.

  According to the second embodiment, when pre-ignition or a sign thereof is actually detected, learning can be completed even if there is a sudden change in driving conditions, contributing to avoidance of subsequent pre-ignition. To do.

DESCRIPTION OF SYMBOLS 1 ... Internal combustion engine 2 ... Fuel injection valve 4 ... Ion current sensor 6 ... Variable valve operating apparatus 10 ... Engine controller 18 ... Intake temperature sensor 19 ... Water temperature sensor 12 ... Throttle valve 21 ... Accelerator opening sensor

Claims (9)

Pre-ignition in the combustion chamber can be identified in at least three states: pre-ignition occurrence, pre-ignition signs occurring in the combustion chamber, and normal combustion in which these pre-ignition occurrences or signs are not detected. In the control device for an internal combustion engine that has ignition detection means and executes predetermined pre-ignition avoidance control when occurrence of pre-ignition is detected,
Multiple learning areas are set up corresponding to at least driving conditions that may cause pre-ignition,
After execution of the pre-ignition avoidance control, when normal combustion is restored, the control amount of the pre-ignition avoidance control is gradually decreased, and the control amount immediately before the pre-ignition sign is detected is Learning as a learning value corresponding to the learning area of time,
Thereafter, in the learning area, a control amount based on the learning value is given in advance ,
A control apparatus for an internal combustion engine, characterized in that the same learning value is reflected in one or a plurality of learning areas determined in advance as a learning area having a higher possibility of occurrence of preignition than the learning area . In a control apparatus for an internal combustion engine, having a pre-ignition detection means for detecting a pre-ignition sign generated in the combustion chamber, and executing a predetermined pre-ignition avoidance control when detecting the pre-ignition sign,
Multiple learning areas are set up corresponding to at least driving conditions that may cause pre-ignition,
The control amount of the preignition avoidance control is gradually increased with respect to the detection of the sign, and the control amount when returning to normal combustion is learned as a learning value corresponding to the learning region at that time,
Thereafter, in the learning area, a control amount based on the learning value is given in advance ,
A control apparatus for an internal combustion engine, characterized in that the same learning value is reflected in one or a plurality of learning areas determined in advance as a learning area having a higher possibility of occurrence of preignition than the learning area .   The control device for an internal combustion engine according to claim 1 or 2, wherein the learning region is set with parameters of the rotational speed and load of the internal combustion engine and one or more temperature conditions. The control device for an internal combustion engine according to any one of claims 1 to 3 , wherein the learning value is held during a trip and reset at the end of the trip. When pre-ignition occurs or a sign is detected outside the range of a plurality of initially set learning areas, a learning area corresponding to the driving condition at that time is newly added to learn a learning value. The control device for an internal combustion engine according to any one of claims 1 to 4 . 6. The control apparatus for an internal combustion engine according to claim 5 , wherein information on the newly added learning region is stored even after the trip ends, and a learning value in the learning region is reset at the end of the trip. During the learning process, when the driving conditions deviate from the entire learning area, the control amount and the corresponding learning area are temporarily stored, and then the learning is performed from the stored control amount when returning to the same learning area. The control device for an internal combustion engine according to any one of claims 1 to 6 , wherein the processing is resumed. Regarding pre-ignition in the combustion chamber, at least three states are identified: pre-ignition occurrence, pre-ignition signs occurring in the combustion chamber, and normal combustion in which these pre-ignition occurrences or signs are not detected, In a control method for an internal combustion engine that executes predetermined pre-ignition avoidance control when occurrence of pre-ignition is detected,
A plurality of learning areas are set in advance corresponding to at least driving conditions that may cause pre-ignition,
After the pre-ignition avoidance control is executed, when normal combustion is restored, the control amount of the pre-ignition avoidance control is gradually decreased, and the control amount when the pre-ignition sign is detected is Learning as a learning value corresponding to the learning area,
Thereafter, in the learning area, a control amount based on the learning value is given in advance ,
A control method for an internal combustion engine, characterized in that the same learning value is reflected in one or a plurality of learning areas determined in advance as a learning area having a higher possibility of occurrence of pre-ignition than the learning area . In a control method for an internal combustion engine that monitors a pre-ignition sign generated in a combustion chamber and executes predetermined pre-ignition avoidance control when the pre-ignition sign is detected,
A plurality of learning areas are set in advance corresponding to at least driving conditions that may cause pre-ignition,
The control amount of the preignition avoidance control is gradually increased with respect to the detection of the sign, and the control amount when returning to normal combustion is learned as a learning value corresponding to the learning region at that time,
Thereafter, in the learning area, a control amount based on the learning value is given in advance ,
A control method for an internal combustion engine, characterized in that the same learning value is reflected in one or a plurality of learning areas determined in advance as a learning area having a higher possibility of occurrence of pre-ignition than the learning area .

Images