JPH02245477A - Ignition device for internal combustion engine - Google Patents

Abstract

PURPOSE:To make optimum combustion possible at any time by detecting a parameter related to combustion, determining the appropriateness of the parameter, changing the control amount related to ignition on the basis of the determination, and storing the control amount for later use in control. CONSTITUTION:Using a microcomputer 2, a ratio of an engine rpm according to an engine rpm sensor 5 to an intake air amount according to an air flow sensor and others is obtained as a load. An ignition pulse interval commensurate with the load is searched to control an ignition circuit 3. As a sensor for detecting a parameter related to combustion, a combustion light sensor 4 for example is provided for detecting combustion light in a position facing a combustion chamber. Obtaining an average (average of ignition delay time) of a plurality of signals outputted from the combustion light sensor 4 and checking if the average is within an appropriate range, and if the average is out of the appropriate range, a compensation program is started to compensate the ignition pulse intervals by increasing or decreasing it according to the ignition delay time average value.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an ignition device for an internal combustion engine, and more particularly to an ignition device for optimizing the control amount related to ignition.

[Conventional technology]

The conventional ignition system for internal combustion engines is disclosed in Japanese Patent Application Laid-Open No. 60-20496.
As described in Publication No. 8, etc., the control amount related to ignition (
The number of ignitions) was memorized depending on the operating condition.

[Problem M to be solved by the invention] The above-mentioned conventional technology does not take into consideration the optimization of the ignition control amount in response to engine differences and changes over time, and the combustion efficiency tends to decrease over time. .

An object of the present invention is to always be able to perform optimal ignition efficiently and with minimum energy under all operating conditions.

[Means to solve the problem]

The above object is achieved by detecting combustion-related parameters (changes over time) with a sensor and optimizing the control amount related to ignition based on the detected values.

[Effect]

Examples of combustion-related parameters include, in the case of a sensor that detects combustion light, the ignition delay detected from the combustion light. The control amount for one ignition, for example, the ignition interval at the time of multiple ignitions, is changed so that the ignition delay is minimized, and the interval at that time is stored and used for subsequent control. In this way, optimal ignition is always possible.

〔Example〕

An embodiment of the present invention will be explained with reference to FIG.

Reference numeral 1 denotes a spark plug arranged in each cylinder of the engine, and when an ignition signal from a microcomputer (ignition control unit) 2 is sent to an ignition circuit 3, discharge for ignition occurs.

The spark plug 1 has a built-in sensor that detects combustion light, and the detected value is input to the microcomputer 2 via the detection circuit 4. Further, a signal from the engine speed sensor 5 is also input to the microcomputer 2 . In this embodiment, the control amount related to ignition is optimized based on the detected value of the combustion light sensor.

FIG. 2 shows a specific example of the ignition circuit 3. The ignition circuit 3 is a capacitor discharge type circuit. The potential from the DC-DC converter 10 is charged to the capacitor 11. Then, when the ignition timing comes, the microcomputer 2
A pulse serving as an ignition control signal is sent to the gate circuit 12. The gate circuit 12 determines a discharge time, sends a pulse to the thyristor 13, makes the thyristor 13 conductive, and discharges the potential of the capacitor 11. With this discharge, a current flows through the primary coil 14 and a high voltage is generated in the secondary coil 15. When this high voltage exceeds the breakdown voltage of the spark plug 1, a spark discharge force of 1 is generated between the plug electrodes.In the zero-wire circuit, the control variables for ignition include the number of ignitions per stroke, and the ignition interval (from 71 icons). (controlled by the number and interval of pulses), secondary current which is discharge energy, etc., and by changing any of these, the control amount can be changed.

For example, when changing the secondary current, the value from the microcomputer 2 is sent to the DC-DC via the D/A converter 16.
It can be controlled by applying the voltage to the C converter 10 and changing the generated potential to change the amount of charge to the capacitor 11.

The ignition operation by the ignition circuit 3 will be explained with reference to FIG. (b) in FIG. 3(a) is the microcomputer 2
(b) is the pressure waveform inside the engine cylinder.

In this embodiment, the thyristor 13 is turned on and off at high speed so that multiple ignitions can be performed in one stroke. Figure 3(b) shows the multiple ignitions. (a) in Figure 3(b) is a signal from the microcomputer 2, and corresponds to (a) in Figure 3(a). do. (b) (b) is a pulse signal from the gate circuit 12.

(c) shows the potential of the capacitor 11.

When the pulse (b) is applied to the thyristor 13, the potential of the capacitor 11 decreases due to discharge.

At this time, a secondary current flows through the secondary coil 15 as shown in (d) and discharge occurs.

The number and interval of this spark discharge are determined by the microcomputer 2.
It is controlled by changing the number of times and intervals of the signal (a). Also, the value of the secondary current (that is, the ignition energy) is
As shown by the dotted line in (c), by changing the charging potential of the capacitor 11, it can be changed as shown in the dotted line in (d).

FIG. 4 shows the configuration of the combustion parameter detection means used in the above embodiment, that is, the sensor for detecting combustion light. A quartz fiber 21 is provided in the center of the center electrode 20 of the spark plug 1, and combustion light is transmitted to a photoelectric conversion circuit (
(detection circuit) 4. Further, the high voltage from the ignition circuit 3 is sent to the center electrode 20 via the lead portion 22.

FIG. 5(a) shows different ignition signals (a) and (b) and the detection value of the combustion light sensor (c).

The ignition signals (pulses) in (a) and (b) have a pulse interval of 1
1 is different. When the ignition pulse shown in (a) is applied, the waveform of the combustion light becomes as shown in (a)' in Fig. 5 (c). Further, when an ignition pulse as shown in (b) is applied, the waveform becomes as shown in (b)' in FIG. 5(c). Here, if we compare the waveforms of (a)′ and (b)′, we find that (b)
The waveform ' has a smaller ignition delay time t, and - is the period from the first ignition pulse to the ignition timing when the output of the combustion optical sensor rises. It is generally known that the smaller the td, the better the combustion. In other words, in this case (a)′
It can be said that (b)' has better combustion than . This is because the flame kernel formed by the discharge grows well. That is, by controlling the ignition pulse interval t1, the ignition delay time td can be controlled, and in turn, the combustion state can be controlled.

FIG. 5(b) shows the relationship between tl and ta. If the number of ignitions is the same, there are 11 where t is the minimum. Let this value be t. Set to 1. (F) and (E) in FIG. 5(b) are tz-td characteristics for different operating conditions. (e) t
Mu apt and (to) t topt have different values. Therefore, it is necessary to store this tto□ according to the operating state.

Figure 6 shows how this memory is stored. In the example of FIG. 6(a), tt is ttopt for the engine load.
It is a memorization of characteristics. As the load increases in this way, the value of t decreases.

The example shown in FIG. 6(b) is an example in which 1+, which is tiopt, is stored on a map corresponding to engine speed and load. The ignition pulse is generated by reading tl from the memory as shown in FIGS. 6(a) and 6(b).

FIG. 7 shows a flowchart 1 of the ignition control of the embodiment.
In the control example shown in FIG. 7, the rotational speed N is read, and the ratio QJL/N to the air amount Q4 is read as a load. and,
From a map such as that shown in FIG. 6(b), ti that matches the engine load is searched and the ignition pulse interval ti is output to the ignition circuit.

Ignition timing is determined by another flow.

Next, a program for optimizing the ignition pulse interval t is shown.

FIG. 8 is a flowchart for determining the average value of 10 times of the time td until ignition, and the value T, which is added 10 times, is divided by 1o to determine the average value ad. Since the ignition delay timing t4 has some cycle fluctuations, an average value is calculated. Next, the 1+ correction program is activated according to the flowchart of FIG. Read the engine speed N and load Qa/N, read the actual md at that time, and check whether md is within the appropriate range m1<md<mz. If it is outside the appropriate range, start the 11 correction program. FIG. 10(a) shows a tt correction program.

First, 1+ is made somewhat larger to 11 +α. The average ignition delay time md at that time is read, and the previous md
If 1IId is smaller than old, t, an increase program is activated, and if md is large, an 11 decrease program is activated.

FIG. 10(b) shows a 1+ increment program.

11 is slightly increased to 11 + α, and the previous md ol
Compare with d. When ld is small, even if t is increased, t will still become smaller, so it is further increased to tl + α. Eventually, when md becomes larger than ad old, 1+ cannot be increased any further. Therefore, t
The previous value of l, that is, the value obtained by subtracting α from the current tl, is determined and stored as the optimum t1 for the engine state N and Qa/N determined in FIG.

FIG. 10(c) shows an 11 reduction program.

In this case, since ad has become larger by increasing 11 in Figure 10(a), here we will change the current 11 to t, -α
I'll make it smaller. As a result of making it smaller, md is nod o
If it is smaller than ld, even if you reduce 11, td
is smaller, so α is further reduced. md is o
If t becomes larger than dd old, t cannot be made smaller any further, so the previous tl is determined by 11+α, and this 11 is stored as the optimal t. In this way, 11, which always minimizes td, can be found.

In the example of FIGS. 9 and 10, first, ti as shown in FIG.
I used a program to determine whether or not to perform correction, but
The correction program t in FIG. 10 may be executed directly by starting the timer without using this.

Since the combustion light sensor is installed in each cylinder of the engine, the above control can be performed for each cylinder.

FIG. 11 shows a control method for each cylinder. Figure 11 (a
) is the reference signal ref of the crank angle sensor 5, which outputs a pulse with a different length for each cylinder. This is shown in the figure below, and the pulse width is set to R using the free run counter.
1 = Convert to a digital value such as Ra and use it for cylinder discrimination. The values of R1-R4 are automatically input into the memory at address I10 in the microcomputer 2 in two time series, as shown in FIG. 11(b). This R1
t of the cylinder in the address area corresponding to the value of -R4.
One map is memorized.

FIG. 12 shows the ti detection method for each cylinder.

First, the contents of the memory in the I10 address (R1 to Rz
) Read parameters such as N and Qa/N. Next, a map of addresses corresponding to the R value is searched, and tc corresponding to N and Qa/N is output to the ignition circuit. In this way, the optimum t for each cylinder
c is required. This tc for each cylinder is learned for each cylinder according to the flow shown in FIGS. 7 to 10.

Next, a method for controlling the one ignition pulse interval t so that the peak position ΔθP of the detection value of the combustion optical sensor shown in FIGS. 5(a) to 5(c) becomes optimal will be described. ΔθP is
There is a crank angle from TDC, and normally ΔθP is 15
The best combustion condition is around 50°.

FIG. 13 shows a flowchart for determining the optimum 11 based on ΔθP. In (a), with ref interrupt,
Reads the output value of the combustion light sensor.

Next, the peak value is detected to determine the peak position ΔθP.

This ΔθP is temporarily stored. FIG. 13(b) shows a flow for starting the 1+ correction program using a timer interrupt.

When ΔθP is within a certain range, for example, 12 × ΔθP × 16, tr is not corrected. When it is outside the range, a ti correction program is started. In the 11 correction program shown in FIG. 13(c), 11 is increased by 1 degree and it is determined whether ΔθP approaches the appropriate value. If yes,
Since we have increased 11 and gotten close to the appropriate value, we start the 1+ increase program. On the other hand, in the case of no, t
Start the 1-decrease program.

FIG. 14(a) shows a program for increasing t in this case. Here, 1+ is increased, it is determined whether ΔθP is an appropriate value, and if it is not an appropriate value, ti is further increased. When ΔθP becomes an appropriate value, the tl at that time
remember.

FIG. 14(b) shows t, a reduction program, t,
, and when ΔθP is not an appropriate value, further t
Decrease z. When ΔθP reaches a proper value, ti at that time is memorized.

In the manner described above, the value of tl for which ΔθP is optimal is determined for each operating state and cylinder.

The peak position of the combustion light as described above is the same as the peak position of the output value of the cylinder pressure sensor. Therefore, the detected value by the cylinder pressure sensor may be used as a combustion-related parameter. In this case, the parameter is also the peak position ΔθP of the cylinder pressure, and the control flow is as described above depending on the peak position of the combustion light. The method is the same. Further, a method of learning t may be adopted so that the rate of increase in cylinder pressure becomes an appropriate value.

Based on the detected values of the rotation speed sensor 5 in FIGS. 15 to 17,
An example of learning tl was shown. In Figure 15(a),
The ref signal and the detection value of the rotation speed sensor are shown. If the signal from the rotational speed sensor is processed for every 1 degree of crank angle and the rotational speed is calculated, a detailed output as shown in (a) can be obtained. Looking at this detected value, the value in the portion corresponding to the cylinder with poor combustion is decreasing. In this embodiment, combustion in this cylinder is improved and learned by ignition control.

In FIG. 15(b), the cylinder discrimination signal R is detected by the ref interrupt.
is stored, and the average value N of the engine rotational speed N between ref is determined and stored together with R.

Next, as shown in FIG. 16(a), by a timer interrupt,
The cylinder whose N is lower than that of other cylinders is determined, the corresponding cylinder number is memorized, and the correction program is started.

In the t1 correction program shown in FIG. 16(b), the ti of the corresponding cylinder is increased by -degree, and it is determined whether the N1 increases. If N increases, a program to increase t is activated, and if ti decreases, a program to decrease t is activated.

In the t1 increase program shown in FIG. 17(a), t continues to be increased until N becomes approximately the same value as other cylinders.

When they become the same, remember 11 at that time and use it for the following control.

In the t1 reduction program shown in FIG. 17(b), t is decreased until N becomes the same as the values of other cylinders, and when they become the same, the 1+ at that time is stored.

As described above, 11 learning controls can be performed even based on the signal from the rotation speed sensor 5.

FIG. 18 and FIG. 19 show an example in which the combustion state is detected based on the generated torque of each cylinder and ignition control is performed. FIG. 18(a) shows a torque detection method. A rotation detector 30 made of gears is provided at the front of the engine crankshaft 29, and a similar detector 31 is provided at the rear.

At this time, the rotation detector output θ1. Torque T is determined from the phase difference Δθ of θ2.

FIG. 18(b) shows the ref signal and the detected value of torque T. As in the previous example, the torque of the cylinder with poor combustion is lower than that of the other cylinders. Identify this cylinder,
Control ignition parameters.

Here, tr is controlled. The control method in this case is the same as that for the rotation speeds shown in Figures 15 to 17, with N being T
Just change it to .

Another control method using torque is shown below.

The relationship between tl and torque T is as shown in FIG. There is a point where T is maximum due to ts. This tl is t
+opt. In control, a timer sometimes
The value t top at which this torque becomes maximum can be found by the mountain climbing method and then learned.

In the example shown above, the ignition interval ti of the capacitor discharge type multiple ignition device is controlled by learning using combustion light, cylinder pressure 2 rotation speed, torque, etc. as parameters caused by combustion, but other than t. The following control variables may also be used.

One example is the number of ignitions per stroke during multiple ignitions, n. That is, t4 is kept constant depending on the operating state, and n is controlled by learning based on combustion parameters.

Also, as shown in (c) and (d) of Figure 3(b),
The ignition energy may be made variable by controlling the amount of charge to the capacitor, that is, the amount of secondary current. In other words, the amount of charge to the capacitor is controlled by learning based on the combustion parameters. This is made possible by controlling the amount of charge via the D/A converter shown in FIG.

The control flow is similar to the control flow when using each parameter.

FIG. 20 shows an example of controlling the duration of spark discharge. FIG. 20(a) shows its hardware configuration. A transistor 43 is connected to the primary coil 41,
The transistor 43 is turned on and off according to instructions from the microcomputer 2.

When it is on, current flows through the primary coil and high voltage is generated in the secondary coil. If a voltage breakdown occurs in the power plug 1, spark discharge will continue even if the voltage from the DC-DC converter 40 is low. If the DC-DC converter 4o is operated while the transistor 43 is on, the spark discharge will continue for the on time.

This situation is shown in FIG. 20(b).

(A) in FIG. 20 is an ignition timing command pulse, and based on this signal, a pulse (B) which is ON only during ti is created and applied to the transistor 43. (c) shows the secondary current, and after the capacitive discharge (1), the inductive discharge (If)
Occurs only for a while.

The relationship between this ts and the ignition delay time 11 is shown in FIG. When ti is increased, t4 becomes smaller and converges to a constant value. ts at which ta begins to converge is tsopt with minimum energy and maximum efficiency.

Here, learning control is performed based on td so that ti always becomes tsOPt.

FIG. 22 shows the correction flow for ts.

As can be seen from the characteristics shown in FIG. 21, in order to determine whether ts is tsopt, it is sufficient to reduce ts once as shown in FIG. 22(a). At this time, if the ts increases, it is determined that the ts is in an area such as A in FIG. 21, and the ts increase program is activated. If t4 does not increase, it is determined that it is in the region B in FIG. 31, and the ts reduction program is activated.

In the ts increase program shown in FIG. 22(b), ts is increased while td continues to decrease.

If td does not decrease, the ts at that time is stored as the optimal ts.

In the ts reduction program shown in FIG. 22(c), ts is decreased until t- increases, and ts when t increases is stored for subsequent control.

In this example, ts was controlled based on combustion-related parameters as shown in FIG. 20(b).

The ignition start timing t1g may also be controlled. In other words, txg
tlg that minimizes t by advancing or delaying
All you have to do is ask for it.

FIG. 23 shows a recovery method when soot adheres to the tip of the combustion light sensor and the output decreases.

In FIG. 23(a), (a) shows the ignition pulse signal, and (b) shows the output of the combustion light sensor. (i) in FIGS. 23(a)-(b) is the case where there is no change over time due to soot, and (ii) is the case where soot is attached. In (n), the output is decreasing. Here, this soot is removed by multiple discharges.

FIG. 23(b) shows the method. Normal ignition is
The ignition ends in the range A (second half of the compression stroke) in (a), but if a change over time is detected, the ignition is continued for a longer period of time, for example, over the explosion stroke. The method shown in FIGS. 23(b)-(b) outputs multiple pulses for soot removal in a stroke other than the compression stroke (exhaust stroke). , (b), the influence of soot removal nokurusu ignition on combustion is smaller.

FIG. 23(c) shows how the spark plug discharges. Discharge sparks 51 from the cathode 50 to the center electrode 2°
occurs. At this time, the quartz fiber 21 is connected to the center electrode 2
Since it is centered at 0, the spark passes through the fiber surface and burns away the soot that has adhered to the surface. This removes the change over time and returns to the original output state.

FIG. 24 shows a flowchart for performing pulse ignition for soot removal.

In FIG. 24(a), it is determined whether or not the engine has started.

After the rich state of warm-up operation is completed, the program for soot cleaning is started every time.

The cleaning program is shown in Fig. 24(b). Here, it is judged whether or not it is the exhaust stroke, and if yes,
A multi-pulse like B in FIG. 23(b)-(b) is output to ignite.

FIG. 24(C) shows an example of another cleaning program. Determine whether it is the ignition timing, and if yes,
Multi-pulses are output as shown in FIGS. 23(b)-(a). Soot is removed through the flow described above.

FIG. 25 shows another method of starting a cleaning program. It is determined whether the operation is idling or not, and when the answer is yes, the ignition light intensity Vp is determined as shown in A in Fig. 23 (a)-(b).
Detect. If this Vp is smaller than the reference value Vrex, it is determined that soot has adhered, and a cleaning program is started. The cleaning program is shown in Figure 24 (
Use b) or (C).

FIG. 26 shows an example of a cleaning program for the ignition circuit shown in FIG. 20.

Once the cleaning program has started, the ignition timing
The discharge duration ts is set to the maximum allowable value, Max-ts, and this is applied to the transistor 43 in FIG. 20(a) to perform a longer discharge than usual. This long discharge removes soot.

〔Effect of the invention〕

According to the present invention, it is possible to always perform ignition with the optimum and minimum energy in response to changes over time.

It is possible to meet the demands for both lower fuel costs and improved drivability.

[Brief explanation of drawings]

FIG. 1 is a system diagram of engine ignition control to which the present invention is applied, FIG. 2 is a configuration diagram of an ignition circuit used in the engine ignition control system, FIG. 3 is an explanatory diagram showing the ignition operation, and FIG.
FIG. 5 is a diagram showing a combination of a spark plug and a combustion light sensor used in an embodiment of the present invention, FIG. Fig. 6 is an explanatory diagram of a map showing the relationship between engine load and optimum 1+, and the method of storing the same. Fig. 7 is a flowchart of ignition control according to an embodiment of the present invention. Fig. 8 is an average of ignition delay. FIG. 9 is a flowchart showing how to calculate the value, FIG. 9 is a flowchart showing the activation of the tl correction program, FIG. 10 is a flowchart showing the t1 correction program, FIG. 11 is an explanatory diagram showing the 1+ control method for each cylinder, and FIG. 12 is a flowchart showing a t1 search method for each cylinder, FIG. 13 is a flowchart for performing correction based on the ignition peak position ΔθP, FIG. 14 is a flowchart showing an 11 increase program for correcting the t amount in FIG. 13, and FIG. Figure 17 shows 1 based on the detected value of engine speed.
Explanatory diagram showing an example of learning +, Fig. 18 and Fig. 1
FIG. 9 is an explanatory diagram showing an example of the operation when ignition control is performed by detecting the combustion state based on the generated torque of each cylinder, and FIG. 20 is a circuit configuration showing another example of the ignition circuit to which the present invention is applied. and an operation waveform diagram, FIG. 21 is a characteristic diagram showing the relationship between the induced discharge time Ts of the ignition coil and the ignition delay time, FIG. 22 is a flowchart showing a correction program for ts, and FIG.
The figure is an explanatory diagram showing a cleaning method when soot adheres to the combustion light sensor, and FIGS. 24 to 26 are flowcharts for carrying out the above-mentioned cleaning. 1... Spark plug, 2... Microcomputer,
3... Ignition circuit, 4... Combustion light sensor, 5... Crank angle sensor (engine speed sensor), 10...
DC-DC converter, 11... capacitor, 12.
...Gate circuit, 13...Thyristor, 14...1
Secondary coil, 15... Secondary coil, 16... D/A converter, 3o. 31... Torque detection sensor, 40... DC-DC converter, 41... Primary coil, 42... Secondary coil, 43... Transistor. Figure 1 Figure 2 Figure 3...Ignition circuit Engine speed detection sensor Figure 3 (b) Figure 5 = H (a) Figure 4 (a) Load (b) Rotation speed diagram r, f interrupt Fig. 10 (a) t1 correction program Fig. 10 (a) t1 increase program Fig. 13 (al ref interrupt Fig. 13 (a) (b) ) Ref interrupt diagram (a) Timer interrupt (b) ti correction program Figure 18 (a) (b) Figure 17 (al ti increase program (b) t1 decrease program Figure 19 Figure 20 (a) (b) ) Figure 22 (C) ts reduction program Figure 21 Figure S Figure 22 (a) (b) Timer interrupt tS increase program Figure 23 (a) (A) (a) (b) Cleaning program Figure 24 ( C) Cleaning program Figure 25

Claims (1)

[Claims] 1. An ignition system for an internal combustion engine, comprising means for detecting a parameter caused by combustion in the engine, a means for determining whether the detected value of the parameter is appropriate, and an ignition system based on this determination. An ignition device for an internal combustion engine, comprising: means for changing a controlled amount regarding ignition; and means for storing the controlled amount regarding ignition. 2. The ignition device for an internal combustion engine according to claim 1, wherein the parameters caused by combustion are determined from the output value of a sensor that detects combustion light. 3. An ignition system for an internal combustion engine according to claim 2, wherein the combustion parameter is an ignition delay time. 4. The ignition system for an internal combustion engine according to claim 2, wherein the combustion parameter is the peak position of the detected value. 5. The ignition system for an internal combustion engine according to claim 1, wherein the parameters caused by combustion are determined from the output value of a sensor that detects the internal cylinder pressure of the engine. 6. The ignition system for an internal combustion engine according to claim 5, wherein the combustion parameter is a pressure peak position or a pressure increase rate. 7. The ignition device for an internal combustion engine according to claim 1, wherein the parameters caused by combustion are determined from the output value of a sensor that detects the engine speed. 8. An ignition device for an internal combustion engine according to claim 7, which changes the control amount related to ignition when the detected rotational speed value decreases. 9. The ignition system for an internal combustion engine according to claim 1, wherein the parameters caused by combustion are determined from the output value of a sensor that detects the torque generated by the engine. 10. The ignition device for an internal combustion engine according to claim 9, which changes the control amount related to ignition when the detected torque value decreases. 11. An ignition system for an internal combustion engine according to any one of claims 1 to 10, in which the control amount related to ignition is the ignition timing. 12. An ignition system for an internal combustion engine according to any one of claims 1 to 10, wherein the control amount regarding ignition is the number of ignitions per one explosion stroke. 13. An ignition system for an internal combustion engine according to any one of claims 1 to 10, wherein the control amount related to ignition is ignition energy such as a secondary current value of an ignition coil. 14. An ignition system for an internal combustion engine according to any one of claims 1 to 10, wherein the control amount related to ignition is the discharge duration of the spark plug. 15. In any one of claims 1 to 14, parameters related to combustion are detected for each cylinder,
An ignition device for an internal combustion engine that is configured to change the control amount related to ignition for each cylinder. 16. An ignition system for an internal combustion engine according to any one of claims 1 to 15, which uses a capacitor discharge type ignition system as the ignition system. 17. An ignition device for an internal combustion engine, which detects changes over time in a sensor that detects parameters caused by combustion, and when changes occur over time, temporarily changes the control amount related to ignition to eliminate the factors that cause changes over time. Ignition system for internal combustion engines. 18. An ignition system for an internal combustion engine according to claim 17, wherein the sensor for detecting combustion parameters is a sensor for detecting combustion light. 19. An ignition system for an internal combustion engine according to claim 17, wherein the ignition control amount is determined by any one of the number of ignitions per stroke, ignition energy, and discharge duration. 20. In claim 17, the change in the ignition control amount is
An ignition system for internal combustion engines that is activated after the engine has started. 21. In claim 17, the change in the ignition control amount is
An internal combustion engine ignition device that activates when the detected value of discharge spark light under specific operating conditions decreases. 22. An ignition system for an internal combustion engine according to claim 17, wherein the temporary ignition control amount is changed in a stroke other than the compression stroke.