JPH0320762B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a digital control method for controlling fuel injection timing, etc. of an automobile.

If the response speed of a hydraulic actuator used in a digital control system with a feedback loop is slow, hunting may occur even with simple on/off control where the application time of the instruction signal is constant. Never. However, when increasing the response speed,
Hunting phenomenon is likely to occur.

This is almost always true within the control system.
This is because there are components of response delay and hysteresis. Furthermore, in digital control, in addition to the response delay and hysteresis mentioned above, generally speaking, regardless of the magnitude of the control deviation value, the command signal sent to the hydraulic actuator is Hunting is likely to occur because the gain change at the rise or end must be abrupt.

To solve these problems, conventionally, the signal to the hydraulic actuator is a periodic pulse signal that vibrates in both directions, and the duty ratio of the two directions is controlled, so that the average position of the hydraulic actuator is controlled vibrationally. However, the control is essentially similar to analog control, thereby preventing the hydraulic actuator from moving on and off excessively and causing hunting in its control. (pulse width modulation)
There are control methods such as

However, these methods using duty ratio control constantly use the solenoid valve used in the hydraulic actuator in a vibrating manner, so the solenoid valve repeatedly turns on and off a large number of times, which corresponds to the number of vibrations. As a result, the durability of the solenoid valve deteriorates, and it is not appropriate to employ such duty ratio control for controlling devices that require durability.

The purpose of the present invention is to solve the above problems.
An object of the present invention is to provide a digital control method that improves the responsiveness and control accuracy of a control system without impairing durability.

To explain the present invention based on an embodiment, FIG. 1 shows an example in which a digital control method according to an embodiment of the present invention is used to operate an injection timing adjustment device 2 of a fuel injection pump 4 in a diesel engine 1. The hydraulic actuator according to the present invention is built into the injection timing adjustment device 2.

The gear 1c fitted on the crankshaft 1a of the diesel engine 1 and the gear 2c fitted on the input shaft 2a of the injection timing adjustment device 2 form a gear connection, and the output shaft 2b of the injection timing adjustment device 2
drives the fuel injection pump 4, and the injection timing adjustment device 2 changes the rotational phase angle between the input shaft 2a and the output shaft 2b according to the operating state of the diesel engine 1. This is to appropriately set the injection timing in the fuel injection pump 4, and the injection timing adjustment device 2 uses a known device.

The hydraulic pump device 3 is driven via gears 2c and 3a, and the hydraulic pump device 3 is connected to a pipe 3b.
It serves as a constant oil pressure source for the solenoid valve 5F via the solenoid valve 5F.

A disc 1b for measuring rotational angle and rotation speed is fitted onto the crankshaft 1a, and a similar disc 4a is fitted onto the output shaft 2b, and the circumferences of each of the discs 1b and 4a are fitted. On the top, protrusions are arranged at equally divided positions in the circumferential direction, and the disc 1b
Electromagnetic pickups 5H and 5G for detecting the rotational position of the protrusion are fixedly installed at positions on the circumferences of the protrusions and 4a, respectively. Note that 5a and 5b are electric wires.

The electric wire 5c is an input interface 5 for inputting the displacement of the accelerator pedal that operates the diesel engine 1.
The electric wire 5d is for transmitting the water temperature in the water jacket of the diesel engine 1 to the input interface 5A, and the electric wire 5e is for transmitting the atmospheric temperature to the input interface 5A. .

Input interface 5A, central processing circuit 5
B, memory 5E and output interface 5D
The electric wire 5k consists of one electric wire _5k1 in FIG. 3 and the other electric wire _5k2 not shown, and the output in the microcomputer 5 is electromagnetic through the electric wire 5m. It controls the valve 5F, and the solenoid valve 5
F is connected to the injection timing adjustment device 2 via the hydraulic pipe 5n.
It operates the hydraulic actuator inside.

As shown in FIG. 2, the input interface 5A is composed of a Schmitt circuit 5AB, an analog multiplexer 5AD and an A-D converter 5AE, and 5p, 5q, 5r, 5s and 5t are electric wires, respectively.

FIG. 3 shows one output interface 5D 1 in the circuit diagram of the output interface _{5D, and one electric wire 5k 1 ,} which is the output of the central processing circuit 5B, is connected to the resistor R ₁ . The electric wire 5u is connected to a constant voltage power supply, the electric wire 5m is connected to one solenoid 5F1 of the solenoid valve 5F in Fig. ₁ , D is a flywheel diode, and _R2 and _R3 are resistors. , TR ₁ and TR ₂
is a transistor.

The other output interface to which the other electric wire _5k2 is input is omitted because it is the same as in FIG. 3.

In the configuration shown in FIGS. 1, 2, and 3, the operation thereof will be explained. The diesel engine 1 is driven by the operation of an accelerator pedal (not shown), and the fuel injection pump 4 is driven by the crankshaft 1.
a, gears 1c and 2c, input shaft 2a, injection timing adjustment device 2 and output shaft 2b,
The injection timing adjustment device 2 is controlled by the rotational speed of the diesel engine 1, the displacement of the accelerator pedal, the water temperature of the water jacket (hereinafter simply referred to as water temperature), and the atmospheric condition. Hydraulic signal (control power) from piping 5n (multiple)
to operate the hydraulic actuator in the injection timing adjustment device 2 in one direction or the other,
As a result, the injection timing adjustment device 2 changes the relative rotational phase angle between the input shaft 2a (that is, also corresponds to the crankshaft 1a) and the output shaft 2b, and as a result, the injection timing of the fuel injection pump 4 is appropriately set. ing.

The injection timing is set by turning the solenoid valve 5F on or off to one side or the other according to the output instruction signal of the microcomputer 5, and the hydraulic signal from the solenoid valve 5F amplified by this operation indicates the injection timing. Although the adjustment device 2 is operated and set, the operation of the microcomputer 5 that controls the solenoid valve 5F will be explained below.

The displacement of the accelerator pedal, the water temperature, and the atmospheric temperature are detected in analog form by detectors (not shown), and as shown in FIG. 2, command signals are sent from the central processing circuit 5B via the electric wire 5r at any time. , the respective detection values are alternately selected in the analog multiplexer 5AD, and the alternately selected detection signals are A-
In the D converter 5AE, the analog values are converted into digital values and each is input to the central processing circuit 5B.

The respective rotation angles of the crankshaft 1a and the output shaft 2b are determined by the electromagnetic pickups 5H and 5.
G, and the respective detected signals are the signals 5as and 5bs shown in FIG.
The detection signal is input to the Schmitt circuit 5AB (FIG. 2).

Note that the horizontal axis in FIG. 6 indicates elapsed time.

Pulse signals sa and sb in FIG.
It shows the voltage change value that occurs when the voltage exceeds (passes).

Each of these pulse signals sa and sb is
In the Schmitt circuit 5AB (FIG. 2), the signals _pa1 and pb are converted into rectangular waveform pulse signals and input to the central processing circuit 5B.

As the rotational speed of the crankshaft 1a increases, the waveforms of the pulse signals sa and sb change as shown by sa', so the pulse signals _pa1 and pb whose waveforms are shaped with reference to the reference voltage level d are
pa changes as ₁ ′, but the pulse signals sa and sb
Since the positions of points a and b of zero voltage in are not affected by the rotational speed, the pulse signal pa ₁
and pb have their reference points at elapsed times corresponding to points a and b.

t ₁ and t ₂ each indicate the measurement time between the pulse signals pa ₁ and pb and the pulse signals pa ₁ and pa ₂ , respectively, and due to their properties, the input shaft 2a and the output shaft 2b The rotational phase angle pe between and the rotational speed n of the crankshaft 1a are as follows.

pe = (c ₁ × t ₁ ) / t ₂ ... (1) n = c ₂ / t ₂ ... (2) However, c ₁ and c ₂ are constants determined by the number of protrusions on the discs 1b and 4a. be.

In the memory 5E in FIG. 1, various data necessary for calculation in the central processing circuit 5B are stored.

To explain an example of this stored data, for example, the fuel injection timing of the diesel engine 1 is ideally determined depending on the conditions of the water temperature, the atmospheric temperature, the displacement θ of the accelerator pedal, and the rotational speed n of the crankshaft 1a. When set, it is possible to optimize the fuel consumption rate or exhaust gas components of the diesel engine 1. For this purpose, the memory 5E stores information such as the displacement θ of the accelerator pedal among these relationships. The target rotational phase angle peo to be set by the injection timing adjusting device 2 in relation to the rotational speed n of the crankshaft 1a is stored in a map as shown in FIG.

In addition, each + mark in FIG. 5 is p ₃₃ , p ₃₄ ,
As shown in p ₄₃ and p ₄₄ , each target rotational phase angle corresponding to the relationship between the displacement θ of each accelerator pedal and the rotational speed n of the crankshaft 1a is stored at equal intervals, and these are Values that change according to changes in temperature are stored.

In addition, in FIG. 5, the electromagnetic pick-up 5H
Actual operating point p determined by the rotational speed n found from and the displacement θ of the accelerator pedal at that time.
The target rotational phase angle peo is determined by the target rotational phase angles p ₃₃ , p ₃₄ , p ₄₃ and p ₄₄ of the state in the vicinity e of point p, and the vicinity e is enlarged and shown in FIG.

In FIG. 4, Ld and dL are the displacement differences of the accelerator pedal, nd and dn each indicate the rotation difference, and the target rotational phase angle peo at point p is calculated in the central processing circuit 5B using the following formula. ing.

peo={(A-B)dn/nd}+B ……(3) However, A={(p ₄₄ −p ₃₄ )dL/Ld}+p ₃₄ ……(4) B={(p ₄₃ −p ₃₃ ) dL/Ld}+p ₃₃ (5) Note that all of the above equations (1) to (5) are stored in the memory 5E.

The target rotational phase angle peo obtained from the above equation (3) is compared with the actual rotational phase angle pe obtained from the equation (1), and based on the comparison, the central processing circuit 5B performs a calculation. will proceed.

Using the calculation results of equations (1) to (5) above, the digital control method in this embodiment performs the following basic control.

That is, for each deviation Er between the target rotational phase angle peo resulting from the above calculation and the actual rotational phase angle pe, the electric wire 5k in FIG.
A pulse-shaped rectangular wave of ts is output, and the relationship between the application time ts and the deviation Er is as shown in FIG.
The characteristics on the right side of the ts axis in _{the figure indicate the application time of the signal applied to the electric wire 5k 1 in} FIG. shows the application time. Furthermore, the slopes of the characteristics on both sides of the ts axis in FIG. 7 are different. This is because the load resistance force when operating the injection timing adjustment device 2 has different values when operating it in one direction and when operating it in the other direction.

Below, the calculation steps in the central processing circuit 5B will be explained using FIG. 7 of the embodiment.

Step 1) Find peo−pe=Er.

Step 2) Store the absolute value of the deviation Er in memory 5E.
Compare with the respective absolute values of the minimum standard deviation Ero and the maximum standard deviation Er ₁ stored in .

Step 3) When the absolute value of the deviation Er is smaller than the absolute value of the minimum standard deviation Ero, the voltage value of the instruction signal is set to zero.

Step 4) The absolute value of the deviation Er is the minimum standard deviation
Greater than the absolute value of Ero and the maximum standard deviation
When Er is smaller than ₁ , the central processing circuit 5B determines the sign of the deviation Er, and when the deviation Er is positive, it is given as one instruction signal to the electric wire 5k ₁ in Fig. 3, and the deviation Er is negative. If so, this is given to the aforementioned electric wire _5k2 as the other instruction signal. At this time, the application time ts of the pulse signal in one of the above instruction signals or the other instruction signal
are each applied for an application time proportional to the deviation Er,
It has a wave-like pulse width with the addition of the minimum application time tso in FIG.

Step 5) When the absolute value of the deviation Er is larger than the absolute value of the maximum standard deviation Er ₁ , the electric wire 5k ₁ or 5k ₂ is continues to output instruction signals.

The instruction signal characteristics resulting from the above calculations are as shown in FIG.

In addition, in Fig. 8, the upper pulse is connected to the electric wire 5k ₁ ,
The lower side is given to the wire 5k ₂ .

In addition, in the explanation in FIG. 3 above,
It has been explained that the electric wire 5k in Fig. 1 consists of the electric wires 5k ₁ and 5k ₂ , but this configuration
k as a single wire, one indicating signal as a positive voltage and the other indicating signal as a negative voltage, which is applied to a single wire 5k, which wire 5k is connected on one side to the output interface 5D ₁ . , the other may be connected to the output interface 5D ₂ (the same circuit as 5D ₁ ) on the other side described above via a code converter. In this case, the effect is similar to that described above.

In the above calculation steps 1) to 5), the first instruction signal pk ₁ is applied to the electric wire 5k in FIG. 1, as shown in FIG. 9, and the application time ts of the instruction signal pk ₁ ends; Within a predetermined time (10 to 15 mmsec) after the end of the process, the deviation still remains.
Even if Er does not fall within the minimum standard deviation -Ero to +Ero shown in FIG. 7, the next instruction signal _pk2 for further minimizing the deviation Er is not transmitted.

This is due to the following reasons.

That is, as shown in FIG. 9, the first instruction signal pk ₁ is applied to the electric wire 5k at t=ts ₁ , and if, for example, the pk ₁ is the instruction signal on one side, then the pk ₁ is applied to the wire 5k ₁ in FIG. 3, and the voltage of this signal is applied to the base of the transistor TR ₁ through the resistor R ₁ , resulting in
The current I from the constant voltage source flows from the electric wires 5u and 5m to the ground through one solenoid _5F1 ,
In this case, the current characteristics with respect to the elapsed time t are as shown in f, g, h, i and j shown in FIG.

This means that even if a rectangular voltage (instruction signal pk ₁ ) is applied to the electric wire 5k ₁ , the current I gradually increases between ts ₁ and f due to the inductance of the solenoid 5F ₁ . When point f is reached, the spool valve in the solenoid valve begins to move. As a result, due to this movement, an electromotive force is generated in the coil of the solenoid _5F1 , and the current I decreases between f and g, and the spool valve is completely opened at point g. When it becomes a state and stops its movement,
The decrease in the current I also ends.

Furthermore, as described above, even after the solenoid valve 5F is completely opened, if the voltage of the instruction signal continues to be applied to the electric wire _5k1 , the current increases again as shown in the characteristics between g and h. Eventually, its value will become saturated.

After the above action, when the voltage of the instruction signal on the electric wire _5k1 becomes zero at point h where the elapsed time t becomes t=ts, the flywheel diode D
and resistor R ₃ , the current I becomes h~i
It decreases while exhibiting the characteristics shown in between.
Since the spool valve starts to close when it reaches point i, the movement of the spool valve generates an electromotive force in the coil of solenoid _5F1 , until it reaches point j, when the spool valve finishes closing completely. , the current increases again and decreases again after point j.
It should be noted that when an instruction signal is applied to the aforementioned electric wire _5k2 (not shown), the other solenoid valve 5F in FIG. 1 is actuated by performing the same action on the other side.

In this way, even if the instruction signal pk ₁ disappears at t = ts, the solenoid valve 5F is activated at point t at point j in FIG.
This means that the valve is open until = ts ₃ .
Correction of the deviation Er continues even between ts and ts ₃ . Incidentally, this phenomenon also occurs in control sections of types other than solenoid valves, and this is because, as mentioned above, the control section has mechanical or electrical inertia. be.

As a result, the deviation Er is often corrected within the desired range during the response delay between ts and ts ₃ , and if the deviation Er at that point is corrected again between ts and ts ₃ , At that point, it is assumed that the deviation Er has not been corrected to within the desired range, and the next instruction signal pk ₂ is
(Fig. 9), as mentioned above, correction of the deviation Er would normally be completed between ts and ts ₃ , but the solenoid valve 5F remains open and the command continues. By transmitting the signal pk ₂ , the deviation Er is corrected to the desired value, which then leads to overshoot.
This is not desirable in terms of control.

Therefore, at the end of the previous instruction signal pk ₁ t=
The above-mentioned "predetermined time" is the time that has elapsed from ts until the last stage of the control system completes complete control due to the previous instruction signal transmission.
This is what has become the norm.

Moreover, this has the following important meaning.

By using the characteristics shown in Figure 7 as described later, this control basically allows for deviations by transmitting only one pulse signal when a deviation occurs in the control. As long as the purpose is to return the deviation within the deviation and the deviation in the control is within the allowable deviation, the above-mentioned instruction signal will not be transmitted.

However, in this control, it is desirable to detect whether a deviation has occurred at as short a time interval as possible. This is because when a new deviation occurs, the control accuracy will be improved if the deviation is corrected as soon as possible.

In other words, in the control method of the present application, after the first instruction signal (one pulse signal) for correcting the deviation is transmitted, the control method continues until the final stage of the control system is completely controlled by the instruction signal. , although the next instruction signal is not transmitted, this means that deviation detection may be performed at much shorter time intervals than the above-mentioned predetermined time intervals.

As a result, when the deviation starts to deviate from the allowable deviation from the state where the deviation has been within the allowable deviation, as mentioned above,
This is because the deviation can be corrected as soon as the deviation starts.

Furthermore, the above problem becomes even more important due to the use of the characteristics shown in FIG.

This means that in the characteristics shown in Figure 7, the deviation Er is -
If it exists in the range of Er ₁ to −Ero or +Ero to +Er ₁ , the fact that the instruction signal application time ts has a component proportional to the deviation Er means that if the resistance of the load is taken into account, it is・Even in loop control, that is, without configuring a negative feedback control circuit, conditions are met such that the deviation Er is corrected to the desired value by one instruction signal. As a result, the instruction signal pk ₁ (Fig. 9) transmitted in accordance with the characteristics shown in Fig. 7 is t = ts ₃ at point j.
There is a high possibility that the deviation Er has been sufficiently corrected to the range of -Ero to Ero to +Ero, and even if the deviation Er has not been corrected to the range of -Ero to +Ero, reaching the vicinity.

For this reason, the measured value of the deviation Er used for the next instruction signal pk ₂ is the same as that of the first instruction signal.
The value after a predetermined period of time, ts ₃ - ts, is used after the pk ₁ signal disappears, thereby reducing the probability of an overshoot phenomenon occurring in the control system. .

As is clear from the above description, the effects of the present invention are as follows.

Unlike the conventional method of continuing to output the instruction signal until the deviation becomes zero, after the detection of the deviation Er is completed, based on the detected deviation value Er, A method is adopted in which an instruction signal with an application time ts that corrects the value of the deviation Er to exactly zero is output. Therefore, the deviation Er is corrected to a value of almost zero by only one pulse of the instruction signal in that control cycle.

Therefore, the response of the control becomes very quick, and the number of signal transmissions applied to the solenoid valve is reduced, so that the durability of the solenoid valve is improved.

Furthermore, in the control of the present invention, when transmitting the second instruction signal following the first instruction signal,
Even after the first instruction signal disappears, the first control result will be sufficiently close to the allowable deviation as predicted, and after the final stage of the control system has completed control, Since the second instruction signal is transmitted, the second instruction signal only needs to be an instruction signal with an application time necessary to correct the slight deviation from the tolerance. , and in a state where the deviation is still large between the end of the first instruction signal and before the completion of control of the final stage of the control system, it is possible to transmit a second instruction signal corresponding to the large deviation. Therefore, the control is not prone to overshoot, and the deviation is rapidly converged to within the allowable deviation.

Therefore, the control based on the second instruction signal not only reduces the frequency of use of the solenoid valve as described above, but also enables control with higher control accuracy and higher responsiveness.

Furthermore, the present invention prevents the next instruction signal from being transmitted until the previous instruction signal disappears and the control at the final stage of the control system is subsequently completed due to the instruction signal. However, it is not prohibited to detect deviations at time intervals shorter than the time required from the transmission of the previous instruction signal to the transmission of the next instruction signal. Therefore, regardless of the required time, if deviation detection is performed periodically at sufficiently short time intervals, provided that the final stage of control has been completed due to the previous instruction signal, As soon as a control deviation occurs, it becomes possible to send an instruction signal for correcting the deviation, and the responsiveness of the control according to the present invention is greatly improved.

[Brief explanation of the drawing]

FIG. 1 shows a system diagram when the digital control method of the present invention is used in a fuel injection timing adjustment device 2 of a diesel engine 1, and FIG. 2 shows a block diagram of the input interface 5A in FIG. 3 is a circuit diagram showing one of the output interfaces 5D ₁ in FIG. 1, and FIG. 4 is an enlarged view of the neighborhood e in FIG. 5. 5 shows a map digitally showing the target rotational phase angle of the fuel injection pump determined by the relationship between the accelerator pedal θ and the engine rotational speed n. FIG. 6 shows that the signals 5as and 5bs detected by the electromagnetic pickups 5G and 5H in FIG. The characteristics of the waveform signals 5ap and 5bp are shown. Figure 7 shows the electric wire 5k in Figure 1.
FIG. 8 shows the characteristics of the instruction signal voltage V appearing on the electric wire 5k in FIG. 1, in accordance with the characteristics in FIG. 7. FIG. 9 shows that according to the instruction signal pk ₁ ,
It shows the characteristics of the current I flowing through the solenoid _5F1 . The symbols used in the examples are as follows:
1: Diesel engine, 1a: Crankshaft, 1
b: disk, 1c: gear. 2: Injection timing adjustment device,
2a: input shaft, 2b: output shaft, 2c: gear. 3:
Hydraulic pump device, 3a: gear, 3b: piping. 4:
Fuel injection pump, 4a: disc. 5: Microcomputer, 5A: Input interface, 5
AB: Schmitt circuit, 5AD: Analog multiplexer, 5AE: A-D converter, 5B:
Central processing circuit, 5D: Output interface, 5
D ₁ : One output interface, 5E: Memory, 5F: Solenoid valve, 5F ₁ : One solenoid,
5G and 5H: Electromagnetic pick-up. 5a, 5
b, 5c, 5d, 5e, 5f, 5g, 5h, 5
j, 5k, 5k ₁ , 5m, 5p, 5q, 5r, 5
s, 5t and 5u: electric wire, 5n: piping, 5as,
5bs, 5ap and 5bp: signal. sa, sa′, sb,
pa ₁ , pa ₁ ′, pa ₂ and pb: pulse signal, pk ₁ and pk ₂ : instruction signal, t ₁ and t ₂ : measurement time, θ:
Displacement of the accelerator pedal, n: rotation speed of the diesel engine 1, p ₃₃ , p ₃₄ , p ₄₃ and p ₄₄ : target rotational phase angle in the injection timing adjustment device 2, Ld and dL: displacement difference, dn and nd: rotation difference , p: actual operating point, d: reference voltage, a, b, c, f, g, h,
i and j: points, R ₁ , R ₂ and R ₃ : resistors,
TR ₁ and TR ₂ : Transistor, D: Diode, ts: Application time of the square wave voltage applied to the electric wire 5k, Er: Deviation between the target rotational phase angle and the actual rotational phase angle for the injection timing adjustment device 2, Ero: Minimum standard deviation, Er ₁ : Maximum standard deviation.

Claims (1)

[Scope of Claims] 1. An actuator that operates a load is actuated by application of an instruction signal, and the application of the instruction signal includes: a; a detected value of a setting value set in the load; A calculator calculates the deviation value from the target setting value to be set for that load, and b; the absolute value of the deviation value and the absolute value of the minimum standard deviation value stored in the memory in the calculator; c; when the absolute value of the deviation value is smaller than the absolute value of the minimum standard deviation value, the value of the instruction signal is set to zero; d; the absolute value of the deviation value is smaller than the absolute value of the minimum standard deviation value; When it is larger than the absolute value of the deviation value, classify the instruction signal into one instruction signal or the other instruction signal by determining the sign of the deviation value; e; Applying each instruction signal; The time is the application time of the value corresponding to each deviation value, f; When the application of the instruction signal is completed, a new deviation is detected again, and the above control is performed again based on the new deviation value. The above control is repeated, and the corresponding relationship between the deviation value and the application time is such that the application time is the application time proportional to the absolute value of the deviation value, plus the minimum application time. The final stage of the control system is completely controlled by the previous instruction signal to determine whether or not to transmit the next instruction signal following that instruction signal. A digital control method characterized by not performing any operation until completion.