JP2894624B2 - Engine ignition timing control device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ignition timing control device for an engine that prevents knock from occurring in an operating region where knock detection is difficult.

[Related Art] In recent years, a technology for detecting an initial knock due to abnormal combustion of an engine and optimally controlling an ignition timing has been widely adopted.

In this case, the engine knock often detects pressure vibration due to abnormal combustion of the air-fuel mixture as mechanical vibration using an acceleration sensor or the like attached to a cylinder head or a cylinder block. It is difficult to distinguish the vibration component from the vibration component itself.

Therefore, knock detection is stopped in the high engine speed region, and the ignition timing is corrected based on a learning value learned in an operation region where knock detection is possible.For example, Japanese Patent Application Laid-Open No. 61-164076 discloses that A technique is disclosed in which a correction value corresponding to a deviation between an actual ignition timing and a basic ignition timing is stored, and when the knock feedback control is stopped in a predetermined operation range, the basic ignition timing is corrected by the stored correction value. ing.

[Problems to be Solved by the Invention] However, the occurrence of knock has a close correlation with the octane number of the fuel. For example, new learning is performed while the octane number of the entire fuel is reduced by supplying low octane fuel by refueling or the like. If high-speed operation is performed before execution, the ignition result is corrected based on the previous learning result because the previous learning result is stored and retained, and the ignition timing becomes excessively advanced and knocks frequently occur May damage the engine.

To cope with this, for example, the learning value is cleared each time the ignition switch is turned off, and the learning is executed each time from a blank state, or a sensor that detects the octane number of the fuel, or detects the fuel supply. Techniques for adding means and the like are conceivable, but in the former, there is a problem that the learning effect is reduced by half, so that improvement in engine output performance and fuel efficiency cannot be expected, and in the latter, the cost increases.

[Object of the Invention] The present invention has been made in view of the above circumstances, and even in an operation region where knock detection is difficult, it is possible to prevent knock generation due to a change in octane number of fuel without adding a special detection means. It is an object of the present invention to provide an ignition timing control device for an engine that can perform the following.

Means for Solving the Problems To achieve the above object, an engine ignition timing control device according to the present invention sets an ignition timing based on an operating state parameter, and detects the occurrence of knocking of the ignition timing to perform feedback correction. In addition, in the ignition timing control device for the engine that corrects with the learning value, as shown in FIG. 1, learning completion determination means M1 for determining whether or not learning of the ignition timing in the knock feedback control region has been completed.
Knock feedback control stopping means M2 for stopping knock feedback control of the ignition timing in a predetermined operation region.
When the knock feedback control of the ignition timing is stopped by the knock feedback control stopping means M2, and when the learning result of the learning completion determining means M1 indicates that learning is completed, knocking is performed based on the ignition timing learning value in the knock feedback control area. The ignition timing outside the feedback control region is corrected, and when the result of the learning by the learning completion determination unit M1 is not learned, the ignition timing outside the knock feedback control region is corrected to be more retarded than when the learning is completed. And an ignition timing correction means M3 outside the knock feedback control region.

Further, the learning completion determination means M1 determines that learning has been completed when the change amount of the ignition timing learning value in the knock feedback control region is within a preset range, and determines that the ignition timing learning value has changed in the knock feedback control region. When the amount is not within a preset range, it is determined that learning is not completed. When the knock occurrence frequency in the knock feedback control area is equal to or less than a preset frequency, it is determined that learning is completed, and the knock feedback control area is determined. When the knocking occurrence frequency exceeds a preset frequency, it is determined that learning is not completed.

[Operation] In the engine ignition timing control device having the above configuration, the learning completion determination means M1 determines whether learning of the ignition timing in the knock feedback control region has been completed,
When the knock feedback control stopping means M2 stops the knock feedback control of the ignition timing in a predetermined operation region, the ignition timing outside the knock feedback control area is changed to knock feedback in accordance with the result of the learning completion determination by the learning completion determining means M1. Out of control area ignition timing correction means M3
Is corrected by

That is, in the operation region where the knock feedback control is stopped, when the learning is completed in the knock feedback control region, the ignition timing is corrected based on the ignition timing learning value in the knock feedback control region, and the knock feedback control is performed. When learning is not completed in the region, the correction is made to be more retarded than when learning is completed.

Further, in the learning completion determination means M1, it is determined whether learning has been completed based on the amount of change in the ignition timing learning value in the knock feedback control region, and the amount of change in the ignition timing learning value falls within a preset range. It is determined that the learning is completed when there is a certain time, and that the learning is not completed when the amount of change in the ignition timing learning value is not within a preset range.

Alternatively, the learning completion determination means M1 determines whether or not learning has been completed based on a knock occurrence frequency in the knock feedback control region.When the knock occurrence frequency is equal to or less than a preset frequency, learning completion and knock occurrence frequency are determined in advance. When the set frequency is exceeded, it is determined that learning is not completed.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

2 to 6 show a first embodiment of the present invention, FIG. 2 is a schematic diagram of an engine control system, FIG. 3 is a flowchart showing an ignition timing control procedure, and FIG. FIG. 5 is a flowchart showing a knock feedback correction value calculating procedure, and FIG. 6 is a flowchart showing a learning completion determining procedure.

(Configuration of Engine Control System) In FIG. 2, reference numeral 1 denotes an engine body, and an intake pipe 3 and an exhaust pipe 4 are provided at an intake port 2a and an exhaust port 2b formed in a cylinder head 2 of the engine body 1. The cylinder head 2 is further provided with an ignition plug 5 that exposes its ignition portion to the combustion chamber 1a.

An air cleaner 6 communicates with the upstream side of the intake pipe 3, and an intake air amount sensor (hot wire air flow meter in the figure) 7 is provided immediately downstream of the air cleaner 6.
Is interposed.

A throttle valve 8 is interposed in the middle of the intake pipe 3, a throttle opening sensor 9 is connected to the throttle valve 8, and a pressure sensor is provided downstream of the throttle valve 8 of the intake pipe 3. Ten are coming.

An injector 11 is located immediately upstream of the intake port 2a communicating with the combustion chamber 1A, and an O2 sensor 13 is located immediately upstream of a catalytic converter 12 provided in the middle of the exhaust pipe 4.

Further, a cooling water temperature sensor 14 faces a cooling water passage 1c formed in the cylinder block 1b of the engine body 1,
Further, a knock sensor 15 is provided on the side wall of the cylinder block 1b.
Is installed.

A crank rotor 16 is fixed to a crankshaft 1d of the engine body 1, and a crank angle sensor 17 is provided on an outer periphery of the crank rotor 16 in a pair.

(Circuit Configuration of Control Means) On the other hand, reference numeral 20 denotes a control device (ECU) including a microcomputer or the like, and a CPU (central processing unit) 21, a ROM 22, a RAM 23, a backup RAM 23a, and an I / O
The O interface 24 is connected to each other via a bus line 25, sets the ignition timing based on the operating state parameter, detects the occurrence of knock, performs feedback correction, and corrects the ignition timing based on a learned value. Is realized, and other functions such as fuel injection control are realized.

The input ports of the I / O interface 24 are connected to the sensors 7, 9, 10, 13, 14, 15, and 17, respectively.
The injector 11 is connected to an output port of the I / O interface 24 via a drive circuit 26, and an igniter 27 is connected to the output port.

An ignition coil 28 is connected to the igniter 27, and the ignition plug 5 is connected to the ignition coil 28 by a distributor.
Connected via 29.

The ROM 22 stores a control program and fixed data such as a basic ignition timing map MPBSE and an optimum ignition timing map MPMBT, which will be described later.
Is formed with a learning map MPIGLRN, and data is stored and held even after an ignition switch (not shown) is turned off. Further, the RAM 23 stores data from the sensors, data processed by the CPU 21, various flags, and the like.

In the ECU 20, when an ignition switch (not shown) is turned on, first, the CPU 21 executes initialization, and clears various flags and the like. When the engine operates, the CPU 21 stores the ROM 22
In accordance with the control program stored in the RAM 23, the fuel injection amount Ti and the ignition timing θIG are calculated based on the various data stored in the RAM 23.

Then, a fuel injection time and an ignition time are set in a counter and a timer for control timing setting, and a drive pulse width signal of the injector 11 corresponding to the calculated fuel injection amount Ti,
An ignition signal to the igniter 27 is output at an optimal timing.

(Operation) Next, the operation of the embodiment having the above configuration will be described based on the flowcharts of FIGS.

(Ignition Timing Control Procedure) The flowchart shown in FIG. 3 is a program for ignition timing control that is repeated at predetermined time intervals.
In step S101, the engine speed N is calculated from the output signal interval of the crank angle sensor 17, and in step S102, the intake pipe pressure PM is calculated from the output signal of the pressure sensor 10, and the process proceeds to step S103.

In step S103, the coolant temperature TW is calculated from the output signal of the coolant temperature sensor 14, and in step S104, the knock sensor 15 is calculated.
The presence / absence of knock occurrence is determined based on the output signal of (1). If knock is present, a knock occurrence determination flag FLAG1 is set to identify knock occurrence.

Next, using the engine speed N calculated in step S101 and the intake pipe pressure PM calculated in step S102 as parameters, in step S105 the optimal ignition timing map MP
The MBT is searched to calculate the optimum ignition timing MBT that gives the maximum allowable advance of the knock limit at the octane number of the fuel set to be used for the engine, for example, premium gasoline, and the basic ignition timing map MPBS in step S106
Search for E and calculate the basic ignition timing IGBSE in the current operating state.

Next, in steps S107, S108, and S109, it is checked whether or not the current operating state satisfies the conditions for performing learning. That is, in step S107, it is determined whether or not the engine speed N is within a range between a learning lower limit speed NLOW (for example, 500 rpm) and a learning upper limit speed NUPPER (for example, 5000 rpm). In step S108, the intake pipe pressure PM is determined. And set value PSET
(For example, 900 mmHg when the engine is supercharged) to check whether the condition of PM> PSET is satisfied.

In step S109, the cooling water temperature TW and the set value TWSET
(For example, 70 ° C.) to check whether or not the condition of TW> TWSET is satisfied.
If all the conditions of S8 and S109 are satisfied, a subroutine for updating the learning value is called in step S110, the learning value is updated, and the process proceeds to step S111.
If there is even one step in which the conditions are not satisfied in S107, S108, S109, the above step S110 is jumped,
The process proceeds to step S111 without performing the learning value update.

In step S111, the engine speed N is compared with the learning upper limit speed NUPPER, and if N ≦ NUPPER, step S1
Proceeding to 12, the value obtained by adding the region-specific learning value APRT for correcting the ignition timing in a partial region to the overall learning value ATOT for correcting the ignition timing in the entire region, and the above-described step S105 Is compared with the optimum ignition timing MBT calculated in the above.

If MBT ≦ ATOT + APRT in step S112, the process proceeds from step S112 to step S113,
The optimal ignition timing MBT calculated in step S105 is set as the learned ignition correction value IGLRN (IGLRN ← MBT) so that the actual ignition timing does not advance beyond the optimal ignition timing MBT.

On the other hand, if MBT> ATOT + APRT in step S112, the process proceeds from step S112 to step S114, and the value obtained by adding the overall learning value ATOT and the learning value APRT for each area is set as the learning ignition correction value IGLRN (IGLRN ← ATOT + APRT). ), And proceed to step S122.

When the process proceeds from step S113 or step S114 to step S122, a knock feedback correction value calculation subroutine is called to obtain a knock feedback correction value AKNK, and in step S123, the above-described basic ignition timing IGBSE calculated in step S106 is calculated. The learned ignition correction value IGLRN set in step S113 or step S114.
Is added and the knock feedback correction value AKNK obtained in step S122 is added to calculate a final ignition timing IGT (IGT ← IGBSE + IGLRN + AKNK), and the routine exits.

On the other hand, if N> NUPPER in the above step S111,
Proceeding from step S111 to step S115, the minimum value of the learning value APRT for each region, that is, the minimum learning value APRTmi of all regions
In step S116, the value of a learning completion flag FLAG3 set by a learning completion determination subroutine described later is checked to determine whether learning is completed.

If it is determined in step S116 that the learning has been completed, the process proceeds to step S117, where the minimum learning value APRTmin obtained in step S115 is added to the overall learning value ATOT, and then the first set value ART1 is subtracted. Then, a learning value ALRN at the time of high engine rotation which is outside the knock feedback control region is set (ALRN ← ATOT + APRTmin−ART1). If the learning is not completed, the process proceeds to step S118, where the overall learning value A is set.
After adding the minimum learning value APRTmin obtained in step S115 to TOT, a second set value ART2 (ART1> ART2) larger than the first set value ART1 is subtracted to perform learning at high engine speed. Set the value ALRN (ALRN ← ATOT +
APRTmin-ART2).

That is, when setting the learning value ALRN at the time of high engine rotation outside the knock feedback control region, the ignition timing correction amount is switched depending on whether learning in the knock feedback control region has been completed, and learning is completed when learning is not completed. The amount of advance is made smaller by making the correction amount smaller than at the time, and knocking is prevented from occurring.

In this case, the minimum learning value APRTmi is added to the overall learning value ATOT.
The value obtained by adding n may be switched between the first coefficient K1 and the second coefficient K2 (K2 <K1 <1) according to the learning completion determination result and multiplied.

Then, in step S119, the learning value ALRN set in step S117 or step S118 is compared with the optimal ignition timing MBT. If ALRN ≧ MBT, the optimal ignition timing MBT is set in the learning ignition correction value IGLRN in step S120. (IGLRN ← MBT) and proceed to step S122, while ALRN <MBT
In step S121, the learning value ALRN is set to the learning ignition correction value IGLRN in step S121 (IGLRN ← ALRN), and similarly, the process proceeds to step S122.

Then, similarly, a subroutine for calculating a knock feedback correction value is called, and a final ignition timing IGT is calculated in step S123, and the program exits.

(Learning Value Update Procedure) Next, a learning value updating program will be described based on the flowchart of FIG.

In the above-described ignition timing control program, step
When all the conditions of S107, S108, and S109 are satisfied and the learning condition is satisfied, a learning value update subroutine is executed.

In this learning value update subroutine, the backup RA
For the learning map MPIGLRN of M23a, the entire learning for correcting the ignition timing values in all the regions is executed to complete the learning for adapting to the operating state at an early stage, and then the partial region-based learning is executed. By doing so, the knocking frequency is reduced early.

That is, it is determined whether or not the overall learning has been completed in step S201. If the overall learning has not been completed, the address of the overall learning value ATOT is saved in the register X in step S202, and the overall learning is completed. If the process has shifted to the learning by region, the process proceeds to step S203, and the address of the learning value APRT by region is saved in the register X.

Next, at step S204, knock occurrence determination flag FLAG1
In step S205, the process proceeds to step S205, where the value obtained by adding the region-specific learning value APRT to the overall learning value ATOT is compared with the optimum ignition timing MBT.

In step S205, when MBT ≦ ATOT + APRT, the process jumps to step S213 without updating the learning value, and
If T> ATOT + APRT, the process proceeds from step S205 to step S206 to read the time T1 from the timer TM1 and check the duration of the knock-free state.

The time T1 measured by the timer TM1 is equal to a predetermined time Ts.
When the time is equal to or shorter than ET1 (for example, 1 sec), the process similarly jumps to step S213 without updating the learning value, and sets the timer TM
When the time T1 of 1 is longer than the fixed time TSET1, that is, when knock does not occur for a fixed time, the process proceeds from step S206 to step S207 to check the value of the knock feedback correction value AKNK.

In step S207, the process jumps to step S213 when AKNK ≠ 0, and when AKNK = 0, the content of the memory designated by the register X in step S208;
That is, the learning value (X) stored in the memory at the address saved in step S202 or step S203.
And a constant advance amount AADV is added to ((X) ← (X) +
(AADV) The advance is updated to the advance side, and in step S209, the timer TM1 for counting the state without knock is cleared, and the process proceeds to step S213.

As a result, when the learning condition is satisfied and there is no knock, all of the conditions in steps 205 to S207 are YES.
At this time, the constant advance amount AADV is added at regular intervals, and the learning value is updated in the advance direction.

On the other hand, if there is a knock in upper step S204, the process proceeds from step S204 to step S210 to subtract a constant retard amount ARET from the content (X) of the memory specified by register X ((X) ← (X)). -ARET), the learning value is updated to the retard side by a fixed amount for each knock.

Next, in step S211, the timer TM1 and a timer TM2 described later are cleared, and in step S212, the counter CT1 is incremented to count the number of knocks, and the process proceeds to step S213.

Next, when proceeding to step S213 and thereafter, the end condition of the overall learning is checked, and first, in step S213, it is checked whether or not the overall learning is being performed. If the overall learning has been completed, the process returns to step S 213, and if the overall learning has not been completed, the process returns from step S 213.
Proceeding to S214, it is determined whether or not the overall learning value ATOT has reached a predetermined constant AMAX.

If the overall learning value ATOT does not reach the predetermined maximum advance angle AMAX in step S214 and if ATOT <AMAX, then step S21
Proceeding to 5, the timer TM2 is cleared to measure the accumulation time when the overall learning value ATOT is equal to or greater than the maximum advance angle AMAX, and in step S217, the count value C1 of the counter CT1, that is, the number of knock occurrences, is read out. The number of occurrences C1 is compared with a predetermined set number of times CSET1 (for example, five times).

In step S217, when C1 <CSET1, that is, when the number of knock occurrences C1 has not reached the set number of times CSET1, the routine exits from the routine and returns. Is determined. Then, in step S218, the overall learning end flag FLAG2 is set, the process exits the routine, returns, and shifts to learning by region.

On the other hand, if the overall learning value ATOT has reached the maximum advance angle AMAX in step S214, and if ATOT ≧ AMAX, the process proceeds from step S214 to step S216, and the overall learning value ATOT
Reads the value of the timer TM2 that measures the accumulated time during which the current is greater than or equal to the maximum advance angle AMAX, and accumulates the accumulated time T2 and the set time TS.
ET2 (for example, 3SEC), and when T2 <TSET2, the process proceeds to step S217 to check the knock generation circuit from the count value of the counter CT. When T2 ≧ TSET, the process proceeds to step S218 to perform overall learning. The process returns by setting the end flag FLAF2, and shifts to learning by region.

(Knock Feedback Correction Value Calculation Means) FIG. 5 is a flowchart showing a knock feedback correction value calculation procedure. When this subroutine is called in step S122 of the above-described ignition timing control program, first, in step S301, the engine rotation is started. It is determined whether or not the number N is higher than the learning upper limit rotational speed NUPPER.

If N> NUPPER in step S301, the process proceeds to step S302 to clear the knock feedback correction value AKNK to 0 (AKNK ← 0) to stop the knock feedback control and return. If N ≦ NUPPER, the process returns to step S302. Proceed to S303 to check for occurrence of knock.

If there is no knock in step S303, step
Proceed to S304 and read time T3 from timer TM3 which counts the duration of no knock state as in timer TM1 described above, and read T3
When ≤TSET1, the routine exits the routine and returns. When T3> TSET1, that is, when knock does not occur for a certain period of time, steps S304 to S305 are performed.
Then, the knock feedback correction value AKNK is added with the knock feedback advance amount AADV1 and updated (AKNK ←
(AKNK + AADV1), the timer TM3 is cleared in step S307, and the routine returns.

On the other hand, if knock has occurred in step S303, the process proceeds from step S303 to step S306, in which the knock feedback delay amount ARET1 is subtracted from the knock feedback correction value AKNK and updated (AKNK ← AKNK−ARET).
1) In step S307, the timer TM3 is cleared and the routine returns.

The knock feedback correction value AKNK at this time is 0
When the knock occurs, the learned ignition correction value IGLRN and the knock feedback correction value AKNK are simultaneously updated in the retard direction by the learning value update routine and the knock feedback correction value calculation routine.

Then, when knocking is eliminated, first, the feedback correction value AKNK is updated in the advance direction. When the knock feedback correction value AKNK is advanced to 0, no further advance is made. IGLRN is updated in the advance direction.

(Learning Completion Determining Procedure) On the other hand, learning completion is determined each time the accumulated time reaches the set time for the accumulated time in which the operating state of the engine is within the learning completion determination area.
The learning completion flag FLAG3 is set.

The learning completion flag FLAG3 is stored at a predetermined address in the RAM 23 and held during the operation of the engine, but is initialized and cleared when the engine is started.

Hereinafter, the procedure of the learning completion determination will be described based on the flowchart shown in FIG.

First, in steps S401 and S402, it is checked from the intake pipe pressure PM and the engine speed N whether or not the current operating state is within the learning completion determination region. If the condition of PM> PSET is satisfied in step S401, Proceeding to step S402, if the condition is not satisfied, exit the program and return.

In step S402, it is determined whether or not the engine speed N is within a set range between a lower limit speed N1 (for example, 2000 rpm) and an upper limit speed N2 (for example, 4500 rpm) for learning completion determination. If it is within the range, the process proceeds to step S403, and if it is outside the set range, the program similarly exits and returns.

Then, when it is determined that the current driving state is within the learning completion determination region and the process proceeds to step S403, the timer TM4 that counts the accumulated time within this determination region is counted up, and the timer TM4 is counted up at step S404. The clock 4 is read and compared with the set time TSET4 (for example, 10 sec).

When T4 ≦ TSET4 in the above step S404, the program exits the program as it is. When T4> TSET4, that is, when the accumulated time in the determination area exceeds the set time, the process proceeds from the above step S404 to step S405. It is determined whether or not the learning value change amount ΔL is within the set range (L1 <ΔL <L2; for example, −2 範 囲 <ΔL <+3 ＜).

In step S405, when the learning value change amount ΔL is not within the set range, the process jumps to step S407. On the other hand, when the learning value change amount ΔL is in the set range, the process proceeds from step S405 to step S406 to set the learning completion flag. FL
AG3 is set, the timer TM4 is cleared in step S407, the program ends, and the routine returns.

The learning completion determination routine may be executed each time the learning value is updated a predetermined number of times.

That is, as a result of the learning completion determination, when it is determined that the learning is completed, the learning result in the knock feedback control area reflects the octane value of the fuel. Knock can be prevented from being caused by a change in the fuel oil's oil number.

In addition, when stopping the engine and refueling, for example, accidentally refilling regular gasoline where premium gasoline is to be replenished, or mixing poor gasoline in poor fuel conditions, etc. Even if high-speed operation is performed suddenly with the octane number of the fuel reduced, learning is not completed because the learning completion flag FLAG3 is cleared, and the learning result in a state where the octane number of the fuel stored in the backup RAM 23a is high. Since the ignition timing is not corrected and the ignition timing is corrected with a preset value, knocking can be prevented.

Second Embodiment Next, a second embodiment of the present invention will be described. The second embodiment differs from the above-described first embodiment only in the procedure of the learning completion determination, and determines the learning completion based on the knocking occurrence frequency.

(Learning completion judgment procedure) The ignition switch (not shown) of the engine is turned on.
Then, learning completion is determined based on the number of knock pulses within each set time in the learning completion determination area, and the procedure will be described below with reference to the flowchart shown in FIG.

First, in steps S501 and S502, it is checked whether or not the current operating state is within the learning completion determination region from the intake pipe pressure PM and the engine speed N. When the condition of PM> PSET is satisfied in step S501. The process proceeds to step S502, and if the condition is not satisfied, the process exits the program and returns.

In step S502, it is determined whether or not the engine speed N is within a set range of a lower limit speed N1 (for example, 2000 rpm) and an upper limit speed N2 (for example, 4500 rpm) for learning completion determination. If it is within the range, the process proceeds to step S503, and if it is outside the set range, the program similarly exits and returns.

Then, when it is determined that the current driving state is within the learning completion determination region and the process proceeds to step S503, the timer TM4 that counts the accumulated time within this determination region is counted up, and in step S504, the presence or absence of knock is determined. Is determined.

If there is no knock in step S504, step S5
Jump to 06, if there is knock, step S5 above
The process proceeds from step 04 to step S505, in which knock pulses are counted by the counter CT2, the count value C2 is counted up, and the process proceeds to step S506.

In step S506, the time T4 of the timer TM4 is read, and compared with a set time TSET4 (for example, 10 seconds), T4 ≦ TS
In the case of ET4, exit the program as is, and T4> TSET4
In other words, when the accumulated time within the determination area exceeds the set time, the above-described steps S506 to S507
Then, the process proceeds to step S2, where the count value C2 of the counter CT2 is read and compared with a set number CSET2 (for example, four times).

Then, the count value C2 of the counter CT2 is C2 ≦ CSET2
Jumps to step S509, while C2> CSE
At T2, that is, when the number of knock occurrences is smaller than the set number, it is determined that learning is completed, the process proceeds to step S508, the learning completion flag FLAG3 is set, the counter CT2 and the timer TM4 are cleared at step S509, and the program ends. And return.

In the second embodiment, similarly to the first embodiment, it is determined that the learning is completed only when it is determined that the learning value reflects the octane value of the fuel. It can be prevented beforehand.

[Effects of the Invention] As described above, according to the present invention, the learning completion determination means determines whether or not the learning of the ignition timing in the knock feedback control area has been completed, and the knock feedback control stopping means determines the predetermined operating area. When the knock feedback control of the ignition timing is stopped in the above, when the result of the learning completion determining means by the ignition timing correction means outside the knock feedback control area is the learning completion, the knocking based on the ignition timing learning value in the knock feedback control area is performed. In order to correct the ignition timing outside the feedback control region and to correct the ignition timing outside the knock feedback control region to a more retarded angle than when the learning is completed when the result of the learning by the learning completion determination means is not completed. , Even in the driving range where knock detection is difficult Thus, knock can be prevented from occurring due to a change in octane number of the fuel, and the fuel cell can be configured at low cost without adding any special detecting means.

Further, the learning completion determination means determines whether or not learning has been completed based on the amount of change in the ignition timing learning value in the knock feedback control region, or has completed learning based on the knock occurrence frequency in the knock feedback control region. When it is determined that learning is completed, the learning value reflects the octane value of the fuel, and the learning result can be effectively incorporated into the ignition timing outside the knock feedback control region. Excellent effects are achieved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of the present invention, and FIGS.
FIG. 6 shows a first embodiment of the present invention, FIG. 2 is a schematic diagram of an engine control system, FIG. 3 is a flowchart showing an ignition timing control procedure, FIG. FIG. 5 is a flowchart showing a knock feedback correction value calculation procedure, FIG. 6 is a flowchart showing a learning completion determination procedure, FIG. 7 shows a second embodiment of the present invention,
It is a flowchart which shows a learning completion determination procedure. M1: learning completion determination means M2: knock feedback control stopping means M3: ignition timing correction means outside knock detection area

────────────────────────────────────────────────── (5) References JP-A-59-138774 (JP, A) JP-A-63-80042 (JP, A) JP-A-63-219875 (JP, A) JP-A Sho 56-138 113026 (JP, A) JP-A-63-120838 (JP, A) JP-B-6-46023 (JP, B2) (58) Fields investigated (Int. Cl. ⁶ , DB name) F02P 5 / 145-5 / 155 F02D 43/00-45/00

Claims (3)

(57) [Claims] An ignition timing control apparatus for an engine, wherein an ignition timing is set based on an operation state parameter, the ignition timing is detected by detecting knocking, and is feedback-corrected and corrected by a learning value. Learning completion determining means for determining whether or not learning of the timing has been completed; knock feedback control stopping means for stopping knock feedback control of the ignition timing in a predetermined operation region; and knocking of the ignition timing by the knock feedback control stopping means. When the feedback control is stopped, when the result of the learning by the learning completion determining means is learning completion, the ignition timing outside the knock feedback control area is corrected based on the ignition timing learning value in the knock feedback control area, and the learning completion is completed. Judgment by judgment means An ignition timing correction means outside the knock feedback control area for correcting the ignition timing outside the knock feedback control area to a more retarded side than when the learning is completed, when the result is not completed. Ignition timing control device. 2. The learning completion determining means determines that learning has been completed when the amount of change in the ignition timing learning value in the knock feedback control area is within a preset range, and determines the ignition timing learning value in the knock feedback control area. 2. The ignition timing control device for an engine according to claim 1, wherein it is determined that learning is not completed when the change amount of the engine is not within a preset range. 3. The learning completion determining means determines that learning is completed when the knock occurrence frequency in the knock feedback control area is equal to or less than a predetermined frequency, and determines that the knock occurrence frequency in the knock feedback control area is a predetermined frequency. 2. The ignition timing control device for an engine according to claim 1, wherein it is determined that learning is not completed when the engine speed exceeds the threshold.