JPS62195464A - Control method for internal combustion engine when it is transient - Google Patents

Abstract

PURPOSE:To improve a transience response characteristic and drivability, by deciding an engine, if an arithmetic value of in-cylinder pressure change rate is out of a predetermined range, to be in a transient condition and determining an ignition timing control value in accordance with the above decision. CONSTITUTION:A peak holding circuit 28 holds a peak of outputs from pressure sensors 22, which are arranged by cylinders of an internal combustion engine 20 detecting pressures in said cylinders, and a control unit 42 calculates the maximum value of pressure in the cylinder. Next, a pressure change rate is calculated from ratio of a pressure increasing rate before ignition to that after the ignition, and the engine, if the calculated value is out of a predetermined range, is decided to be in a transient condition. While an in-cylinder pressure maximum angle is calculated on the basis of outputs from the pressure sensors 22 and a crank angle sensor 36. And when the transient condition is detected, the control unit corrects a pressure maximum angle target value by a predetermined amount as compared with when the engine is in a steady operation condition while corrects the ignition timing of the cylinder, to be ignited in the next time, by a value smaller than the predetermined amount.

Description

Detailed Description of the Invention (Field of Industrial Application) The present invention relates to a transient control method for an internal combustion engine, and more specifically, to detect a transient state of engine operation by detecting fluctuations in cylinder pressure of an internal combustion engine. The present invention relates to a transient control method for an internal combustion engine that corrects ignition timing and improves transient response.

(Prior art) In the conventional transient control method for an internal combustion engine, the ignition timing main control value is determined based on the engine speed and load condition, and the ignition timing main control value is determined based on the engine cooling water temperature, throttle valve opening, manifold negative pressure, etc. The main control value was corrected by detecting the transient state of engine operation. As an example, Special Publication No. 58-40027
The technology described in this issue can be mentioned.

Furthermore, in recent years, an example has been proposed in which the ignition timing is controlled by detecting the cylinder pressure of an internal combustion engine, and the transient state is detected from the engine cooling water temperature, etc., as one example.
The technique described in No. 6-21913 can be mentioned.

(Problems to be Solved by the Invention) However, in the first conventional example, a large number of detection means and their processing circuits are required to detect a transient state, which has the disadvantage of complicating the device configuration. there were. Furthermore, when detecting transient conditions, the detection accuracy is not sufficient because it is detected indirectly through changes in engine cooling water temperature, etc., and since it is not feed-hack control, much is expected from its correction effect. It was something I couldn't get. Therefore, the control values had to be precisely determined, and corrections were only used as a secondary function, so the main control value map inevitably increased and a large capacity memory was required to store it. The disadvantage was that it had to be prepared. Furthermore, since the detection of transient states was also indirect, it was difficult to say that the transient response was sufficient, and it was difficult to say that sufficient consideration was given to drivability.

Further, the second conventional example also suffers from similar disadvantages in that a large number of detection means and their processing circuits are required to detect a transient state. In addition, feedback control is used to directly detect the combustion state of the engine, find the deviation between the target angle and the actual maximum pressure angle, and determine the control value in the direction of eliminating the deviation. There was no equivalent suggestion as to how to eliminate the deviation in a transient state without impairing drivability. Furthermore, no equivalent countermeasures were provided for non-king, which tends to occur frequently during transient conditions.

Therefore, an object of the present invention is to eliminate the above-mentioned drawbacks of the conventional example, to detect the internal cylinder pressure of the engine, directly detect the combustion state of the engine, and perform feedback control, and in particular, to measure the pressure fluctuation. This alone makes it possible to detect transient states of engine operation, and therefore, when implementing this method, the number of detection means and their processing circuits can be reduced, and by detecting from the combustion state, transient states can be detected more accurately. Therefore, it is an object of the present invention to provide a method for controlling an internal combustion engine during a transient period, which can improve transient response and drivability.

Furthermore, the maximum pressure angle θpmax can be determined by directly detecting the combustion state.
By focusing on the target value, it is possible to ignite closer to so-called M, B, and T, thereby improving the output, and also in a transient state and M.

It is an object of the present invention to provide a transient control method for an internal combustion engine, which also makes it possible to accurately detect and control knocking, which tends to occur frequently when ignition is near B, T, and so on.

(Means for Solving the Problems) To achieve the above object, the present invention provides an internal combustion engine control method using cylinder pressure as at least one parameter for determining control values, as shown in FIG. Calculate the rate of pressure increase before and after the ignition of the first cylinder and find the ratio of the two (
Step 10), then find similar ratios for the same or different cylinders (step 12), then find the fluctuation rate of those ratios, and if the fluctuation rate exceeds a predetermined range, the engine operation is in a transient state. If it is determined that there is, the maximum pressure angle target value is changed to a predetermined amount of retardation compared to the steady state of operation, and the ignition timing of the cylinder to be ignited next time is retarded by an amount smaller than the predetermined amount (step 14.16). ).

(Example) Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. still,
For convenience, the device for realizing the method according to the present invention will be described in the second section.
This will be explained with reference to the block diagram in Figure 3 and the waveform diagram in Figure 3.

In FIG. 2, reference numeral 20 indicates an internal combustion engine, which in this embodiment has four cylinders. A piezoelectric sensor 22 for detecting the cylinder pressure is disposed in each cylinder at a position facing the combustion chamber thereof.

The sensor output is passed through a low-pass filter 2 via a charge-to-voltage converter or high impedance circuit (both not shown).
4 is input. The cutoff frequency of the filter is set higher than the knocking frequency component, so that high frequencies during non-king can also be detected. A multiplexer 26 is connected to the next stage of the low-pass filter 24, and the multiplexer selectively transmits the filter output to the next stage in the order of cylinder explosion according to a command from a control unit 42, which will be described later. A peak hold circuit 28 is connected to the next stage of the multiplexer, and holds its output at a peak and outputs it as shown in FIG.
A D conversion circuit 30 is connected. The conversion circuit inputs the output from the peak hold circuit, converts it into digital data every predetermined unit time or angle, and the maximum value of the data indicates the maximum pressure value P max. Alternatively, it is possible to perform digital conversion at a predetermined crank angle before peak hold reset, and use the digital value as the maximum pressure value P may (FIG. 3).

A comparison circuit 32 is connected in parallel with the A/D conversion circuit 30 at the next stage of the peak hold circuit 28, and a pulse down edge detection circuit 34 is connected at the subsequent stage. The comparator circuit 32 has an inverting input terminal to which the peak hold circuit output is input, and a non-inverting input terminal to which the multiplexer output is directly input, so that a slight difference cannot be given to the input values of the two. Therefore, it is configured to output a pulse signal at the position where the maximum pressure value occurs (Fig. 3). As shown in Figure 3, the pulse is, in principle, one pulse at the maximum value generation position when knocking does not occur (Figure 3(a)), and when knocking occurs and high frequency components are superimposed. In such a case, a pulse is output not only at that position but also every time the sensor (multiplexer) output exceeds the peak hold output, and as a result, a plurality of pulses are output (as shown in FIG. 2(b)). Further, the pulse down edge detection circuit 34 detects the falling edge timing of the output pulse of the comparator circuit and outputs a pulse of a predetermined time width so that it can be easily processed by a circuit to be described later. Therefore, by measuring the time from the reference position, for example, the TDC position, to the pulse generation position and converting the measured value Tpmax into an angle, the maximum pressure angle θpmax can be calculated, and by counting the number of generated pulses, non-king can occur. The presence or absence of can be detected. Note that, as shown in FIG. 3C, if the sensor fails, no pulse is generated even after time measurement is completed.

A crank angle sensor 36 is provided near the rotating part of the internal combustion engine 20, and detects a predetermined crank angle, e.g., 720, when the explosion of four cylinders goes around in the order of the first, third, fourth, and second cylinders.
cylinder discrimination signal every 180 degrees, and TDC of each cylinder every 180 degrees.
A TDC signal is output at the fifth position, and angle measurement signals are output for each predetermined angle obtained by subdividing the TDC signal. Therefore, by counting the TDC signal after the cylinder discrimination signal is generated, the cylinder can be discriminated. Incidentally, the engine rotation speed can also be calculated from the angle measurement signal.

Note that the cylinder discrimination signal may be configured to obtain a required detection signal based on a predetermined amplitude value obtained from a pressure sensor.

Further, the internal combustion engine 20 is provided with a negative pressure sensor 38 at an appropriate position between the throttle valve 40 of the internal combustion engine and an intake manifold (not shown) as a sensor for detecting the load state of the engine. This sensor, together with the crank angle sensor, detects the operating state of the engine and performs bank-up when the pressure sensor is abnormal, and can also be used for basic ignition timing calculation as described later, if desired.

A control unit 42 is connected to the next stage of the sensor groups 36 and 38, the A/D conversion circuit 30, and the pulse down edge detection circuit 34, and inputs their outputs. In the case of the embodiment, the control unit has an input/output interface 42a.
, a CPU 42b, a memory 42C and a clock 42d. In addition, the CPU
includes a timer counter that counts the pulses of the pulse counter and the clock 42d to measure time, a cycle counter that counts ignition cycles for knocking control, and an advance angle counter that counts the number of ignitions after the knocking ends. (not shown). An ignition device 44 is connected to the next stage of the control unit, and receives its output to ignite the air-fuel mixture in the engine combustion chamber via a spark plug (not shown). Further, upon receiving the reset operation of the peak hold circuit 28 and the input of the crank angle sensor 36, the unit identifies the cylinder state and instructs the multiplexer 26 to select a gate.

Next, an embodiment of the control method according to the present invention will be described with reference to the waveform diagram in FIG. 3 and the flow charts in FIGS. 4 and 5.

First, in step 50, the pressure sensor output is converted to BTDC.
Capturing starts from a predetermined angle, for example 150 degrees.

Subsequently, in step 52, the cylinder is determined and a cylinder address C/A=n is assigned to specify the cylinder. This is the cylinder discrimination signal of the above-mentioned crank angle sensor signal and T
This is done by the arrival of a DC signal.

This is because this control is characterized by being performed for each cylinder.

Simultaneously with the arrival of the T and DC signals, the timer counter (TC) and pulse counter (PC) are started in step 54, and time measurement and pulse counting shown in FIG. 3 are started.

Subsequently, in step 56, after a predetermined ATDC angle, for example 30 degrees, has elapsed, the count values of both counters are referred to. Third
As shown in the figure, if the pulse counter value is “0” and no pulse is generated even though the timer counter has sufficiently exceeded the maximum pressure value generation position and time measurement has finished, the pressure sensor is It can be judged as abnormal (Figure 3 (C)
).

If not, it is determined in step 58 whether the pulse counter value exceeds a predetermined value and whether or not knocking has occurred. Note that the predetermined value is normally set to "1", but may be set to a value of "2" or more in consideration of the possibility that a plurality of pulses may be generated even during normal combustion due to noise or the like.

If the pulse count value is smaller than a predetermined value, it is determined that knocking has not occurred, and the maximum pressure angle θρmax is determined in step 60. This can be done by multiplying the elapsed time Tρmax to the position where the maximum pressure value occurs by the converted value. The converted value is obtained by ((rotation speed rpm x 360 degrees) 760 seconds).

Subsequently, in step 62, the explosion compression ratio ΔPEXP
/COMP is calculated, which is shown in FIG. 5 as a subroutine. Below, while also referring to Figure 6, this sub-
Explain the routine. First, to explain roughly, as shown in FIG. 6, the pressure value p1 at an arbitrary point before ignition
and the pressure value p2 at a point a minute angle section θ1 away from there.
, and a pressure value p3 at an arbitrary point after ignition, and a pressure value p4 at a point a minute angle section θ2 away from the ignition, and after determining the respective pressure increase rates, calculate the ratio between the two. Then calculate the same ratio for another (or the same) ignition cylinder,
Calculate the rate of change of the two ratios. When the rate of variation is not within the "predetermined dead zone region", it is determined that the system is in a transient state, and the angle is retarded. That is, the inventor discovered a causal relationship between the above fluctuation rate and the transient state of engine operation, and accomplished the present invention. The following will explain: First, in step 62a, the pressure value pl-p4 is read out and measured.

Subsequently, in step 62b, the compression pressure change rate Δpc, which is the pressure increase rate before ignition, that is, in the compressed state.

MP, the rate of change in explosion pressure ΔP EXP, which is the rate of pressure increase in an explosive state after ignition, and the ratio of the two, PEXP /COMP, are calculated as follows.

d (P2-PL) ΔP COMP =□ dθ1 d (P4-P3) ΔPEXP=□ dθ2 ΔP EXP PEXP /COMP= - ΔPCOMP Note that the point before ignition is at the time of compression as the piston rises after the intake valve is fully closed. You can choose any point. Also, even after ignition, the TDC position and maximum pressure angle θpIIlaX
Anywhere within the range is fine. However, it is convenient to select these points for the TDC position and maximum angle θpmax since the TDC signal and the maximum pressure value generation position signal (pulse) can be obtained, respectively. It goes without saying that once the point before ignition is selected, the point is searched for using the crank angle sensor output or the like. Since this control is performed for each cylinder, the calculated value is specified as PEXP/COMPn and stored. This operation is performed one after another for each ignition cylinder, and the same type of ratio is calculated and stored.

Subsequently, in step 62c, the calculated values of another explosion cylinder before one ignition and the same cylinder four times before ignition, that is, in the previous cycle, are read out. Note that initial settings are made as appropriate for the first time.

Subsequently, in step 62d, the fluctuation rate with respect to the ratio of the current cylinder is determined. In addition, the fluctuation rate from the ratio before one ignition is PE
XP/COMPA, PEX the fluctuation rate with the ratio before 4 ignition
Let P/COMPB.

PEXP /COMPn PEXP /COMPA - PEXP /COMPn-1 PEXP /COMPn PEXP /COMPB = PEXP /COMPn-4 Subsequently, in step 62e, the fluctuation rate PEXP /
It is determined whether COMPA is larger than "1+predetermined dead band area", and if it is larger, it is determined in step 62f whether the other fluctuation rate PEXP/COMPB is also larger than "1+predetermined dead band area". The dead zone area is “0.1
If it is determined to be large in step 62f, it can be determined that the fluctuation rate between cylinders is relatively large, and it can be determined that the engine operation is in a transient state, thereby avoiding knocking and deterioration of the exhaust gas composition. In this way, when looking at fluctuations between cylinders, we compare not only the cylinder before one ignition, but also the cylinder before four ignitions, and only when they are in the same phase does it become a transient state. Since the engine speed is determined, there will be no false detection of temporary fluctuations.Furthermore, since the rate of change before and after ignition is also taken into consideration, errors in determination are unlikely to occur even if the engine speed changes suddenly.

If it is determined that the state is a transient state, then in step 62g, the pressure maximum angle target value θpO is set to the first predetermined angle retard side to prepare for knocking or the like as described above. Therefore, this first predetermined angle is appropriately determined to an extent that prevents knocking and the like.

Subsequently, in step 62h, the correction amount θt for the cylinder to be ignited next time is set to the retard side by a second predetermined angle. This second predetermined angle is also set appropriately, but as a feature of the present invention, the second predetermined angle is set to a value smaller than the first predetermined angle. Therefore, if the ignition correction angle θt is larger than the first predetermined angle in step 621, it is set to the first predetermined angle (step 62m). That is, as will be described later, in the present invention, when a transient state is detected, it is necessary to perform correction in order to eliminate the deviation that occurs as a result of changing the target angle to the retarded side.

If the deviation is corrected in the relevant cylinder in the next cycle at once, the drivability will deteriorate due to sudden changes in the ignition timing, so to prevent this, partial correction is made in the next ignition cylinder, and the transient state is also maintained in the next ignition cylinder. Since this will be detected, partial correction is applied in the same way to the next ignition cylinder, and by adding partial corrections one after another in this way, the same cylinder after 4 ignitions finally almost matches the target angle. It was done as if to make it happen. Therefore, the second predetermined angle is appropriately set to a value such that the target angle is reached after four ignitions. Note that in the flow chart, "-" means a retard angle, and "+" means an advance angle.

If the answer in step 62e is negative, step 62i indicates that PEXP/COMPA is “1-
It is determined whether the fluctuation rate PEXP/COM is smaller than a predetermined dead zone area.
It is also determined whether PB is smaller than "1-dead zone area", and if it is smaller, it means that the fluctuation rate is relatively large, so the retarding angle is corrected in the same way.

If the results of steps 62f, 62i, and 62j are negative, the pressure maximum angle target value θ is assumed to have a small fluctuation rate.
po is set to the initial setting value (62p), and the ignition correction angle θ
If t does not have a retard angle correction value, no correction shall be added,
If there is a predetermined retard correction value, the ignition angle is gradually advanced by a second predetermined angle for each ignition cylinder to avoid deterioration of drivability due to sudden changes in ignition timing (steps 62p, 62n,
620.62k).

Returning to FIG. 4 again, after the subroutine of step 62, in step 64, the maximum pressure angle target value θpo and the actual maximum pressure angle θpmax calculated in step 60 are compared to determine the deviation Δθpmax. Note that if the target value is corrected in the subroutine, the corrected target value is set to θpo.

Subsequently, in step 66, nosoking correction MKN
It is determined whether R is not "0", that is, the remaining amount of non-king correction amount is determined by referring to the memory 42c, and if the remaining amount is "0", in the next step 68, it is determined whether the deviation Δθpmax is behind the target value θpo. Decide whether to proceed.

If it is a delay, the deviation correction amount θpc is a value obtained by adding a third predetermined angle appropriately set to the previous θpc (initial setting "0") to advance the angle (step 70); The value is subtracted and retarded (step 72), and if there is no deviation, the previous value is left unchanged. (Step 74
). Incidentally, if this third predetermined angle is set to a relatively small value, the deviation will be gradually eliminated, which has the advantage of improving drivability as in the subroutine.

As mentioned above, subtraction means retard angle correction, and addition means advance angle correction.

Incidentally, if knocking is detected in step 58, the knocking correction angle KNR (initial setting "O
”) by subtracting an appropriately set fourth predetermined angle (step 76), and retard the angle until the retard amount reaches a suitably set fifth predetermined angle larger than the fourth predetermined angle (steps 78 and 80). ), the deviation correction amount θpc is set to the value at the current ignition (step 82).In addition, if there is a non-king correction amount remaining in step 6G, the fourth predetermined angle is Step 84.86) to return to the advance angle side. When 80pmax is ahead of the target value, there is no need to return it to the advance side, so θpc is retarded by a third predetermined angle (Step 88.72). ), in the case of a delay, the correction amount θpc is left unchanged (step 82).The predetermined time (number of ignitions) after the end of knocking is measured using the cycle counter and advance angle counter of the CPU in the control unit. do.

Subsequently, in step 90, the sum of the correction amount θpc and the knocking correction amount KNR is set as the feedback correction amount θf. Incidentally, if it is determined that the sensor is abnormal (step 56), the value obtained by retarding the sixth predetermined angle, which is set as appropriate, is set as the feedback correction amount (step 92).
).

Next, in step 94, the feedback correction amount θf thus obtained is temporarily stored (or rewritten if already stored) so as to be used as the correction value for the next cycle for the same cylinder (cylinder address=n). Therefore, all the correction amounts including the non-king correction amount obtained in the above-described procedure are reflected only in the relevant cylinder, so that control can be performed according to the particular combustion state of each cylinder.

Subsequently, in step 96, the stored feedback correction amount θf of the cylinder to be ignited next (cylinder address=n+l) is read out (initial setting "0"), and the ignition of the cylinder is commanded. At that time, the ignition timing is the basic ignition timing + θ
It is commanded by f+θL (step 98). Furthermore, this “θ
f" and "KNR" in step 90 themselves mean quality values, and as a result, they mean subtraction, that is, retardation. This θf is the feedback correction for the cylinder in question read in the previous step. θf is the sub-value in Figure 5.
This is the amount of correction during a transient state determined by a routine. Therefore, only the correction amount θL is used as a correction amount beyond the cylinder, and by reflecting a correction amount smaller than the pressure target value correction amount in the next cylinder in this way, even during transient drivability is improved. Even with such a stepwise retardation, the transient state is detected relatively quickly by detecting the cylinder pressure and directly monitoring the combustion state, so there is no problem. Moreover, as a result of making it possible to control knocking, sufficient countermeasures against knocking, which is likely to occur during transient periods, are taken.

Note that this basic ignition timing calculation may be performed only from the cylinder pressure to calculate the maximum pressure angle target angle θpo, or may be performed by detecting the operating state of the engine from the crank angle sensor and the negative pressure sensor. Also good. Even when calculating basic ignition timing from engine speed and load, the deviation between the actual maximum pressure angle and the target angle is detected after ignition, and the speed-to-load map value is changed using the corrected angle. In preparation for the next ignition, this is repeated sequentially to perform feed-back control to the target angle, so the rotation speed-load map value only needs to be small (therefore, the device implementing this method needs to use a small memory capacity). It has sufficient advantages.

(Effects of the Invention) The present invention is configured to detect the cylinder pressure, detect the transient state of engine operation from the change in pressure, detect the maximum angle of the pressure, and correct the deviation from the target value. As a result of monitoring and controlling the combustion state, the ignition timing can be brought closer to M, B, and T, and the output can be improved. Further, even when this method is implemented, there is an advantage that the number of sensors and processing circuits for detecting a transient state can be reduced. In addition, the basic ignition timing can be calculated from only the cylinder pressure or from the engine speed and load as before, and even if it is detected from the engine speed - load, the cylinder combustion state can also be detected. As a result of performing feedbank control to focus on the target angle, an extremely rough main control value map is sufficient, and this method has the advantage that the device can be configured with a small memory capacity. Furthermore, it has the advantage that knocking control is also possible at the same time, and since the amount of correction during the transient state is made stepwise, it also has the advantage of excellent drivability.

[Brief explanation of drawings]

Fig. 1 is a claim correspondence diagram of the present invention, Fig. 2 is a block diagram of a device used to realize the present invention, Fig. 3 is its output waveform diagram, and Figs. 4 and 5 are examples of the present invention. Flow showing
6 is an explanatory diagram illustrating the contents of the flow chart of FIG. 5.

Claims (1)

[Claims] A control method for an internal combustion engine that uses cylinder pressure as at least one parameter for determining control values, which includes: a. Calculating the rate of pressure increase before and after the ignition of one cylinder and determining the ratio of the two; ,b.Next, find similar ratios for the same or different cylinders,c.Next, find the fluctuation rates of those ratios, and when the fluctuation ratio exceeds a predetermined range, engine operation is in a transient state. determining that the maximum pressure angle is present, and changing the target value of the maximum pressure angle to retard by a predetermined amount from that in the steady operating state, and retarding the ignition timing of the cylinder to be ignited next by an amount smaller than the predetermined amount. Engine ignition timing control method.