JPS5943942A - Control of idle revolution speed of internal-combustion engine - Google Patents

Abstract

PURPOSE:To permit control with the proper transient response by setting the upper and the lower limit values in the estimated value of the state variable and the integration value of the deviation of revolution which are the main information of the control input value of a dynamic model for controlling a number of variables and allowing said limit values to be held. CONSTITUTION:An integration operation gain block 14 calculates two control inputs, width PA of the control solenoid driving pulse of VCM for controlling the bypath air amount in idling, and ignition timing IT, from the integration value of the deviation SA between the aimed value Nr of the engine revolution speed and an actual value N, and the state variable x calculated in a state supervisor 13, and controls the idle revolution speed N of an internal-combustion engine. Said state variable x is obtained through estimation of the dynamic internal state of a control subject 12 from the transmission function matrix which is linearly approximated. The upper and the lower limit values are set in the integration value of the deviation of revolution and in the estimated value of state variable, and when the limit value is reached, holding is performed. Thus, overflow and underflow in calculation can be prevented, and the idle revolution speed can be controlled with the proper transient response.

Description

Detailed Description of the Invention (Technical Field) This invention considers the internal state of an internal combustion engine, treats the engine as a dynamic system, and analyzes the dynamic behavior of the engine using state variables that define the internal state. The present invention relates to a method of controlling idle rotation speed using a state variable control technique that determines engine input variables while estimating engine input variables, and in particular, preventing automatic flow and underflow of calculations within a computer used for control. The present invention also relates to a method for correctly controlling control.

(Prior Art) As a conventional idle rotation speed sub-side method for an internal combustion engine, there is, for example, the one shown in FIG. The AAC knob 1 for idle rotation speed control determines the drive knob width PA of the control solenoid 3 of the VCM valve 2.
Due to the VC control, the amount of air flowing through the throttle valve 4 and the amount of bypass air passing through the Hs 5 changes, causing the idle rotation speed to change to 11J1111I.
-1j 0 control unit 6 receives idle (IDLE) signal from throttle valve switch 7 and neutral (NEUT) signal from 211 Trurus Inochi 8
When it is detected that the engine is in an idling state by a signal, a vehicle speed (vsp) signal from the vehicle speed sensor 9, etc., the idle rotation speed is determined by a one-dimensional table lookup corresponding to the cooling water temperature (7"W) by the water temperature sensor 10. The basic target value of the air conditioner switch 11 is calculated.
Air conditioner (A/C) (Stupid, Neutral (N)
EUT) signal, battery voltage (V#) signal, etc., and the final idle rotation speed target value NrK%J is calculated, and the idle rotation speed NrK%J at the end of the engine is determined.
The pulse width PA of the control valve 3 is controlled by proportional and integral (PI) duty so that the deviation SA between the rotation speed and its target value Nr becomes small, and the target rotation speed Nr is fed (
control.

FIG. 2 shows a flowchart of the above control method.

However, in the conventional idle rotation open side method 1 of an internal combustion engine, PI control that effectively uses the dynamic characteristics of the engine, actuator, and sensor is not performed, and the equations PI control as a control method has become unsuitable for controlling multi-input/output systems, so when the engine enters or exits the idle state from another operating state, and when various load disturbances occur, When the engine exhibits dynamic behavior, such as immediately after a load disturbance is applied or immediately after the end of a load disturbance, what happens when the engine exhibits dynamic behavior? l
T! There was a problem in that the I control followability, that is, the transient response was poor. In addition, 11 attempts to increase the degree of freedom of control by adjusting other control inputs to improve controllability had 9 problems in that it was difficult to apply with the IPI control method.

(Objective of the Invention) The present invention was developed by focusing on this more conventional problem. ,Kira optimizes control followability, that is, transient response when the engine exhibits dynamic behavior, such as after 1σ when a load disturbance is applied, and also increases the degree of control freedom by learning a large number of control input variables. The purpose of this invention is to facilitate improved controllability, thereby achieving more stable idle rotation resistance.

In particular, in order to achieve the above objectives, when controlling the idle rotation speed by effectively using models of the dynamic characteristics of the engine, actuators, and sensors, overflow and underflow of calculations within the computer used for control must be avoided. The purpose is to prevent the flow and make it more accurate to continue calculations.

(Structure and operation of yJi Akira) Therefore, this invention Q'1, a model of the four characteristics of internal combustion engine, actuator, and sensor 11 is stored in a controller consisting of a microcomputer, and the air amount is (or equivalent amount), ignition timing, 1'? ,
Input any one or any combination of two or more of the supply amount (or equivalent amount) and the exhaust gas recirculation (GR) amount (or equivalent amount -1 t) to the fljlJ input, and The rotation speed is used as the control output, and from the control input and output, a state variable representative of the internal state of an internal combustion engine, which is a dynamic model, is estimated, and the deviation between the estimated value, the target value of the idle rotation speed, and the actual clearance area is calculated. In the method of feedback controlling the idle rotation speed of the internal combustion engine to the target value by determining the control input value using the integral value of the rotation deviation and the estimated state variable quantity, It is characterized by setting.

The present invention will be explained below based on the drawings.

FIG. 3 is a block diagram of an apparatus for implementing an embodiment of the method for controlling the idle rotation speed of an internal combustion engine according to the present invention.

In the same figure, 12 is the control target, Nai'moiso Seki,
In addition to idle rotation speed control, it also performs fuel injection control including air-fuel ratio feedback control. Controlled object 1
When the control output of 2 is the idle rotation speed, the control inputs are the air amount (or equivalent amount), ignition timing, fuel supply amount (or phase drawing amount), and Jul-air flow rate (ox or equivalent amount). Any one of them or a combination of two or more of them may be used. In real MM 1+l, 2'Ni
lf# large input to adjust the 'flilJ control lunoid (11th valve 1) of 10M valve 2 for adjusting the bypass air volume of the idler 11, bβ pulse width PA (i.e. bypass air volume) The zero control output that takes the amount corresponding to ) and the ignition timing JT is the idle rotation speed N, and is the output.

I3 stores a dynamic model of the engine 12, which is the object of your control, and uses a state observer (observer) to estimate the dynamic internal state of the engine from the control input/output information PA, JT, and N of the upper H pi. ), inside: 5
State variable quantities X (for example, four quantities "I" representing the tl state)
+ ``2 + ``3 + ``4 vector representation)) is the quasi-regular territory X in total.

The state observer 13 simulates the target engine, and records the dynamic internal state as a state variable
It is represented by (71, the next vector X, 〜〜). Specifically, the state f number representing the five internal oil states of the engine 12, which is the object of IT control, is the absolute pressure of the intake manifold, the suction negative pressure, the total amount of air actually taken into the cylinder, and the combustion dynamics. mechanical behavior, engine torque, etc.These f
If it can be detected by the it sensor, the detection 1i-U'
By using c, dynamic behavior can be grasped and used for control, allowing more precise control. However, at present, there are not many practical sensors that can detect these values. Therefore, the internal state of the engine is represented by a state variable X, but the state variable X needs to correspond to various physical quantities that represent the actual internal state.
:J: 7Z < , it simulates the engine as a whole. Degree of state variable X? As for L, the larger n is, the more accurate the simulation becomes, but on the other hand, the world becomes more complex. Therefore, a low-dimensional approximation is used as a model, and approximation errors or errors due to Matsuseki individual differences are absorbed by integral operation. Appropriate. In Figure 3, 14 is the '4% operation and gain block, and the specified target value IVr and actual value N of the engine rotation speed are
The amount obtained by integrating the deviation SA from and the state divided by the state observer 13! From the FA variable amount X, two 1.1111 skin j
11 of the input lJA table JT is calculated (see Figure 5).

Next, the effect will be explained.

The engine 12 that is the target of control 1j is a 2-man power 1-output system, and the rotation synchronization sample 1 between the input and output is
++It is possible to determine the dynamic internal state 't411f of the controlled object 12 from the linearly approximated function matrix 'I' (z) obtained near the set value.Part 1 Shape H74fj411 as two hands θ mi
[Vessel] There are 3. Under operating conditions near idle speed,
The transfer function matrix 'f' (Z) of the controlled object 12 is found experimentally and is expressed as '1'(z)-[7'1(z) T2(2
):I (]). 2 indicates the 2-transformation of the sample value of the input/output signal, and 7'1 (Z) and T 2
(Z) is a quadratic transfer function of ITl>2.

Input, output and transfer function 7+, (Z), 7t2
Show the relationship between (Z)? The model structure of the target (engine) 12 on the Ijj side is shown in FIG. However, input and output use deviations δPA, δIT, and δN from the reference setting values, respectively.

From this transfer function matrix T (Z), the state view (i
ll device 13 can be configured.

First, from 'r (z), a state variable model that describes the dynamic behavior of the engine x (n) = Ax (n-1) + f3u (n-1)
(2) y(u-1) -Cx(n-1)
0 which leads to (3) Here, (n) in parentheses for fragrance amount represents the current time, and (?t-1) represents the previous sample time. ij]j is a control input vector, with the pulse width δP,1 (nl) of the control lunoid 3 and the ignition timing δIT representing the perturbation within a range where linear approximation from a certain reference setting value holds as elements. In other words, y()+,-1) is '1ljlJ output,! Similar to the IjJ side j person force vector, δ#(n-1) representing a perturbation from a certain reference rotation speed Na (for example, 650 rPL) is used as an element.

That is, y (n-1) = δ#(tt-1)
(5) X(・) is a state variable vector, matrix 8, B, and C are constant matrices determined from the coefficients of transfer function matrix 'I' (2).Here, we have the following algorithm. State observation device
' to accomplish.

x(n)=(A-GC)x(n-1)+Bu(n-1)
+Gy(n, -1) (6) Here, G is an arbitrarily given matrix, and x(ri is the internal state variable of the engine 12!
・) is the estimated value. (2)(3) Transforming from equation (6), [X(?Z)-'H(n)]-(A-GC)[x(n-
i)-x(n-1)] (7), and the matrix (A-C;
If G is chosen so that the eigenvalue of C) is within the unit circle, then x(n)−+ x(n) at n→large
(8), and the internal state variable quantity x(zz) can be estimated from the input U(·) and the output y(·). It is also possible to appropriately select the matrix G and make all the eigenvalues of the matrix (A□GC) zero, in which case the state observer 13 becomes a finitely stable state observer.

When it is determined to start idle rotation speed control, state observation is started at the same time as the start judgment, but as can be seen from the algorithm of equation (6), the initial 1st shift x (
o) must be given. For example, if the engine rotation speed is 90 Or when it is determined to start idle rotation speed control #
pm, if the initial value X(0) of the estimated state is set in a range close to that state, subsequent estimation can be performed quickly and reliably, and the target rotation speed (for example, 650 rpm) can be determined from the corsting. It also has good control over transient response when controlling the engine speed, and can prevent coasting engine stall.

However, even in the same state at 900 rpm, when the throttle is fully closed is 200 rpm, and from there the rotation speed decreases to 900 gm, and when it is 4000 rpm.
The throttle was fully closed and the rotation speed decreased to 900.
rpm, the internal state variable values of the engine are different, and the initial value of the estimated state, x(o), is the engine rotation speed when the throttle is fully closed, and the engine rotation speed when idle rotation speed control starts. By giving the information according to the engine rotational speed at the time of judgment, it becomes possible to estimate the correct state.

Generally, control inputs of physical systems necessarily have an input range. In the system of this invention, for example, when looking at the valve opening setting of VCM valve 2, ? l+11
11-II input values are limited to a range of 0% to 100%. When the control input value reaches such upper and lower limits during the control during idling, the control input value is held at the upper and lower limits 11σ. (6) The state variable amount 9 (.) estimated from the input changes depending on the idle rotation speed, which is the control output, but the byte of the state variable 1111 allocated to the RAM of the microconbeater etc. Since the number is obscene, unless we set a 1-lower limit for a certain shift,
This will cause an overflow, and there is a possibility that subsequent state estimation will be incorrect. Therefore, by setting upper and lower limits immediately before overflow and immediately before underflow, correct state estimation can be continued.

Using information on the state variable X (and the deviation SA = (Nr - N(-)) between the target rotational speed Nr and the actual rotational speed N(-)) estimated in this way,
There are 4 people! ! The standard setting for the driving pulse width of the i-controlled lenoid 3 is the increase δPA (・) within the range where a linear approximation from CL holds true, and the standard setting for the ignition timing within the range where a linear approximation from the 1st shift holds. The increase amount δIT(·) is determined, and the optimal regulator control of the engine idle speed N is performed. Regulator control () means return 11N control that controls the idle rotation speed N to match the target rotation speed Nr, which is a constant range.

In the present invention, since the experimentally obtained model is a low-dimensional approximate model as described above, an I (integral) operation is added to absorb the approximation error.
Performs optimal regulator control including operation.

As mentioned above, the institutions that are subject to this control are 2
It is a system with only one human power output, which is optimally controlled by a regulator, but the optimal regulator control algorithm for a general multivariable system is, for example, the one written by Katsuhisa Furuta,
Linear system? Since it is explained in "Li Control Theory" (1972) by Shokodo et al., detailed explanation will be omitted here. If we describe only the result, then δu(7+,)-= u(n)-u(?L-])
(9) δe (7t)−
= Nr - Mt, , ) (10),
Let the evaluation function J be 0, where the hill is the weight parameter 1], and t indicates the transposition. k is the 11 number of times when the control start point is 0, and the second JM on the right side of equation (11) is the square of equation (9) (R
Let f be a diagonal matrix) f: 1 table. Also, the second term of equation (11) is converted to fft! as shown in equation (9). The differential input of the I input is in a quadratic format, as shown in Figure 5.
This is because ′[ was added. (]1) Evaluation of formula (i1fi
The optimal control input value that minimizes the function J is given by equation (12).
3), then K is the optimal gain matrix 0 or (12
), and P is, (16) 0, which is the solution of the Riccati equation.
The meaning of the evaluation function J in equation (11) is to minimize the control output y (the deviation SAC rotation fluctuation of the idle rotation speed N from the target value Nr) while constraining the movement of the control human power U(・). The weighting of the constraints can be varied by the weight parameter matrix R. Therefore, by selecting an appropriate value H, and using the dynamometer model (state variable model) of the engine at idle, the optimal gain matrix of equation (13) calculated using P obtained by solving equation (16) is ' t Stored in the microcomputer and set the target idle rotation speed f
ljNγ and actually f17i. From the integral value of the 'flrii difference SA of N and the estimated state variable x(k), (12)
The optimal control input value u*(k) can be easily determined by the formula.

Here, the integral of the deviation SA when a disturbance occurs in a general short-term 100-hour load cycle will not be a very large 1 shift, but it will be a fairly large load disturbance (if the brake is partially engaged, the The integral value of the deviation SA is quite large when the air disturbance (for example, the state where the driver presses the accelerator pedal lightly and the idle switch is not turned on and the throttle is slightly opened) continues for a certain period of time (for example, 10-odd seconds). If the upper and lower limits 1e are not set, there is a possibility of underflow overflow, and accurate control cannot be continued.

Therefore, the integral value of the deviation that can be considered to some extent is
1+T is set, and when that value is exceeded, the upper and lower limits are held.

The values of C and G are stored in a microcomputer, and the state initial value x(o ) and calculate υ using equation (6). The area of this state initial value
For example, the engine rotation speed at 1 o'clock when the throttle is fully closed and the engine rotation speed at the start of idle control are stored in the microcomputer as a two-dimensional table.

Also, when it is determined to start idle rotation speed control (
At this time, if the engine rotation speed is 90 rpm), (12
) formula, Σ(Nr-N(j)] is given by the original j
==0, if Nr is 65Orpm, ΣCNr−N(j
)Jj=.

-250r7ym, but in this case, the 1-separate control input value becomes small, and as shown in Fig. 7 (74), an undercut with respect to the target idle rotation speed (650 rpm) occurs. , causing coasting engine stall.

Therefore, the first shift of Σ(Nr - #(j)) is set to a pseudo-large J20 region, that is, the actual rotation #(o) appears to be higher than the target idle rotation speed v1=J, or the target rotation is If it is set lower, '+lflJ Oirinotsuki Nao will become larger, and as shown in Fig. 7 ■, although some coordination with the target rotation speed will be used, undershoot will be almost eliminated and it will be extremely low. It can be controlled by women.

The above initial values are given according to the engine rotation speed when the throttle is fully closed and when it is determined that idle rotation speed ili control is to be started (for example, in the form of a two-dimensional table). 8] Figures I(4) and (5) show the control performance when idling occurs during idling. Figure 8 (A) shows that the idle switch is turned on again after idling, and at this time the predetermined (B) rotation is performed. When it is determined that idle rotation suppression is to be started at a speed below (for example, 1100 rpm) (the engine rotation speed at this time is A'
(j) ) --30Orpm, and control is performed using that 11α. Figure 8 (Jl) shows the controllability when a pseudo 500rpm is given, and a safety margin to avoid stalling is set. In order to obtain a large value, a large initial value is given to h, fruits and vegetables. It is clear that the sword obtained by the method of the present invention shown in FIG. 8(B) can prevent coasting engine stalling.
4) (Both C and C are set to the same value.

The above procedure of idle rotation speed control is shown in the sixth section.
To explain the 0 procedure shown in the figure, in step “30,”
First, it is determined whether or not to enter idle rotation speed control.If it is determined that 0 control is to be performed, in step 31, the integral value of the control deviation, the state variable quantity sense, the value of the water temperature TW, etc. are determined so that coasting engine stall does not occur. Determine the target value Nr of the idle rotation speed. In step 33, the deviation SA between the target value Nt- of the idle rotational speed and the actual value 1[N is calculated. In step 34, the rotation speed deviation SA from the start of control to the previous cycle is added, and the result is DU
Move to register N. In step 35, the reference setting value 1'Ja of the actual value N of the rotational speed (for example, 650 r
Calculate the deviation δN from pm ).

Step 36 is the state variable quantity X, ~z3 (calculated value # times) estimated at the previous control cape according to the algorithm for estimating the dynamic internal state of the engine, and the divided value ↑1j1.
Each state variable amount x1 is calculated by weighting δP4 and δIT of the first side j people's month 1 and δN, which is the control month 1. However, the matrix (A
-GC) is an example of forming a finitely settled observer in the form.

Note that (A, B, C) use observable canonical forms.

In step 37, it is determined whether the calculated DUN

However, the upper and lower limits of DUNX''+ to x4 are set so that the calculated Ri control input PAIT can be practically actuated by physical force.

In step 38, the estimated internal state variables x1 to x4 that are important to the dynamics of the engine and DUN-Σ[A#-
j = O N (, 7)] is multiplied by the element k i j of the optimal gain to calculate 7Il'l, and standard setting one time (PA)a and ITa
Calculate how much the control input value should be increased.

(Effects of the Invention) As explained above, according to the present invention, the control input (any one or a combination of any two or more of the nitrogen amount, ignition timing, fuel supply amount, and exhaust gas recirculation amount) and ffj
In the multivariable control method based on the Guinaminok model between control outputs (idle giant rotation speed), the integral value of the rotational deviation and the estimated value of the state variable quantity are the main information in determining the control input value. By setting an upper limit and a lower limit 1σ for 1σ, and holding the hold when the limit value is reached, overflow and underflow of calculations within the converter are prevented, resulting in more accurate control). The effect is that idle rotation speed control with optimal transient response can be achieved.

[Brief Description of the Drawings] Fig. 1 is a block diagram of a conventional idle rotation speed control device for an internal combustion engine, Fig. 2 is a flowchart showing a conventional idle rotation speed control method, and Fig. 3 is an internal combustion engine according to the present invention. Implement the idle rotation speed control method of jJi1. do li
The configuration diagram of the control device, Figure 4 is similar to Figure 3? Figure 5 is a block diagram showing the relationship between input and output on the ill side and the engine; Figure 5 is a diagram showing details of the top gain block; Figure 6 is a flowchart 1 explaining the control method according to the present invention. FIG. 7 (4) is a diagram showing the transient response during coasting by the conventional method, FIG. FIG. 8 is a diagram showing a transient response during dry blowing according to the method of the present invention. 1... AAC pulp, 2... VCM pulp,
3...Control solenoid, 4...Throttle valve, 5...Bypass, 7...Throttle valve switch, 8...Neutral Suinochi, 10...Water temperature sensor, 11...Air conditioner switch, I2 ・・Internal combustion engine (side] target), 13... Condition observation device, 14... Integral operation + gain block, N
r...Target value of idle rotation speed, N...Actual value of idle rotation speed, Na...Reference setting value of idle rotation speed, SA...
・Difference between target fil and actual value of idle rotation speed, PA
...control solenoid/noid drive pulse width that defines the amount of bypass air, IT...ignition timing 1 Patent applicant: Nissan Motor Co., Ltd. Patent application agent Megumi Yamamoto - No. 37 #i Diagram procedure amendment (Spontaneous) October 5, 1980 Kazuo Wakasugi, Commissioner of the Japan Patent Office 1. Indication of the case 1988 Patent Application No. 153994 2. Name of invention Method for controlling idle rotational speed of internal combustion engine 3. Amendment. Relationship with the case involving the patent applicant (399) Eli Jidosha Co., Ltd. 1. Idle rotation speed control method” and supplement 1-ro. (2) Correct 1-I (・le rotation speed control) in lines 6 to 7 on page 5 of the specification to be “idle rotation speed control.” 3) Correct “( (See Figure 5).
(See Figure 5). And the above state view 11111 hot and cold]
:3, integral action, and K/block]・1 constitute a controller. Correct it as j. -H

Claims (1)

[Claims] Based on the internal combustion engine dynamic model stored in the co/troller, Ai is the internal combustion engine control input value, which is the amount of air to be supplied to the internal combustion engine, or the corresponding amount of air. Any one dimension selected from the ignition timing of the internal combustion engine, the amount of fuel supplied to the internal combustion engine, or the amount equivalent to the amount of fuel supplied, and the amount of exhaust gas recirculation, or the amount equivalent to the 1J1 air recirculation amount, or From an arbitrary combination of two or more and the idle rotation speed which is the open side 1 output value of the internal combustion engine, a state variable quantity Z of an appropriate order representing the dynamic internal state of the internal combustion engine 1 is determined; (?:
: ], 2. ... n), and from the estimated state variable quantity xi (Z = 1, 2...n) and the integrated value of the deviation SA between the target value Nr and the actual value N of the idle rotation speed. In the method of determining the control input lunar surface and feedback controlling the idle rotation speed of the internal combustion engine, MiJ
An internal combustion engine idle rotation speed control method 0 characterized in that an upper limit value and a lower limit value are set for the integrated value of the deviation SA and the determined state variable x2.