CN1065586C - Fuel injection control device for IC engine - Google Patents

Abstract

In a case where a feedback correction coefficient is suitably calculated using a control algorithm having an adaptative parameter regulating mechanism employing the Landau's regulation law, even when the parameter regulating mechanism is operated for each fuel control cycle for each TDC, an input used in the parameter regulating mechanism should be a value for each combustion cycle such as TDC in a specific cylinder. In addition, an adaptative controller and the parameter regulating mechanism for calculating an adaptative parameter used in the adaptative controller are operated in synchronism with a combustion cycle, and the control cycle of the adaptative controller and the parameter regulating mechanism is changed in accordance with the number of revolutions of the engine. This makes it possible to reduce a matrix operation volume so as to reduce in turn the load of an on-vehicle computer, thereby making it possible not only to complete an operation within one TDC but also to improve the controllability by reducing idle time. Thus, it is possible to continue the adaptative control even in a driving situation such as at a high revolution when the operation time tends to be reduced, thereby making it possible to obtain a good controllability.

Description

Fuel injection control device for internal combustion engine

[ technical field ]

The present invention relates to a fuel injection control device for an internal combustion engine, and more particularly to a fuel injection control device that performs fuel injection control by adaptive control and can be realized in an actual engine.

[ background art ]

In recent years, adaptive control theory has been introduced in the field of internal combustion engines, and a technique of performing control using an optimum regulator, which is one of modern control theories, to match the amount of fuel actually sucked into a cylinder with a desired amount of fuel has been proposed, for example, as described in japanese patent laid-open No. 1-110853.

The present applicant has also proposed a fuel injection control system for an internal combustion engine using adaptive control in japanese patent application laid-open No. 6-66594 and the like. However, when the fuel injection control device using the above adaptive control is mounted on an internal combustion engine, the calculation time of the internal combustion engine increases and decreases due to a change in the engine rotation speed, and the mounted microcomputer cannot be freely selected due to a performance constraint. In addition, although the normal fuel control cycle is performed once for each TDC, since 8 to 12 TDCs are required from after the fuel injection until the control result is detected, there is a waiting time of 8 to 12 control cycles. In general, the control performance is deteriorated when the waiting time of the control target is long as compared with when the waiting time is short. This phenomenon is particularly significant in adaptive control.

Therefore, an object of the present invention is to provide a fuel injection control apparatus for an internal combustion engine that solves the above-described problems, ensures controllability, and enables an adaptive controller to be used for a real engine.

It is another object of the present invention to provide an internal combustion engine fuel injection control device that determines an operation amount by an adaptive control method, and that can continuously perform adaptive control even in an operating state in which a calculation time is reduced such as high-speed rotation, thereby achieving excellent control performance.

[ disclosure of the invention ]

In order to achieve the above object, according to the present invention, a fuel injection control device for a multi-cylinder internal combustion engine includes: a fuel injection amount control device for controlling a fuel injection amount of a multi-cylinder internal combustion engine, an adaptive controller for adaptively matching the fuel injection amount as an operation amount with a desired value, and an adaptive parameter adjusting means for calculating an adaptive parameter used by the adaptive controller; the input to the adaptive parameter adjusting means is synchronized with a fuel control cycle of the internal combustion engine, and the adaptive parameter adjusting means performs calculation of an adaptive parameter based on at least one parameter of an air/fuel ratio or an in-cylinder fuel amount based on a specific combustion cycle.

Further, the input to the adaptive parameter adjusting means is synchronized with a specific cylinder fuel control cycle of the internal combustion engine.

The adaptive controller is operated in synchronization with a fuel control cycle of the internal combustion engine.

Further, a fuel injection control device for an internal combustion engine according to the present invention includes: an air/fuel ratio detecting means for detecting an exhaust air/fuel ratio of the internal combustion engine, a fuel injection amount control means for controlling a fuel injection amount of the internal combustion engine for each fuel control period, and a recursive controller for making the fuel injection amount coincide with a desired value by using the fuel injection amount as an operation amount based on at least the detected exhaust air/fuel ratio; the controller of the above-described recursive type is operated in synchronization with a cycle longer than the fuel control cycle in a predetermined operation state.

In addition, the above-described recursive form of controller is an adaptive controller.

The adaptive controller includes adaptive parameter adjusting means for calculating an adaptive parameter to be used for the adaptive controller, and the adaptive parameter adjusting means is operated in synchronization with a cycle longer than the fuel control cycle in a predetermined operating state, while at least the detected exhaust air/fuel ratio is inputted into the adaptive parameter adjusting means.

The period longer than the fuel control period is a value corresponding to an integral multiple of the combustion period.

The detected air/fuel ratio input to the recursive controller is a value based on several values detected at a cycle shorter than the operating cycle of the recursive controller.

The detected air/fuel ratio input by the adaptive parameter adjustment means is a value based on several values detected in a cycle in which the operation cycle of the adaptive parameter adjustment means is short.

A fuel injection control device for an internal combustion engine according to the present invention includes a fuel injection amount control means for controlling a fuel injection amount of the internal combustion engine, an adaptive controller for operating the fuel injection amount as an operation amount in accordance with a desired value, and an adaptive parameter adjusting means for calculating an adaptive parameter used by the adaptive controller; and an operation state detecting means for detecting an operation state of the internal combustion engine, wherein the control cycle of at least one of the adaptive controller and the adaptive parameter adjusting means is changed in accordance with the detected operation state.

The control cycle of the adaptive parameter adjusting means is set to be equal to or greater than the control cycle of the adaptive controller.

The control cycle of the adaptive parameter adjusting means is set to be an integral multiple of the control cycle of the adaptive controller.

Further, the control cycle of at least one of the adaptive controller and the adaptive parameter adjusting means is changed by a cycle which is an integral multiple of the fuel control cycle.

The operating state is at least the rotational speed of the machine.

[ brief description of the drawings ]

Fig. 1 is a schematic diagram showing an overall fuel injection amount control apparatus for an internal combustion engine according to the present invention.

Fig. 2 is an explanatory diagram showing a detailed structure of the exhaust gas recirculation mechanism in fig. 1.

Fig. 3 is an explanatory view showing a detailed configuration of the tank type washing mechanism in fig. 1.

Fig. 4 is an explanatory diagram showing a valve speed regulation characteristic of the variable angle governor in fig. 1.

Fig. 5 is a block diagram showing a detailed configuration of the control unit in fig. 1.

Fig. 6 is a main block diagram showing the operation of the fuel injection control device of the internal combustion engine according to the present invention.

Fig. 7 is a block diagram functionally representing the operation of the program of fig. 6.

Fig. 8 is a pulse waveform diagram showing an example of the operation of an adaptive controller used in the fuel injection system of an internal combustion engine according to the present invention.

Fig. 9 is a diagram of a pulse waveform 1 showing another example of the operation of an adaptive controller used in the fuel injection system of an internal combustion engine according to the present invention.

Fig. 10 is a block diagram re-writing the block diagram structure of fig. 6 with emphasis placed on the STR regulator and adaptive parameter adjustment mechanism.

Fig. 11 is a subroutine block diagram showing an operation of calculating an average value such as a feedback correction coefficient by the adaptive control method in the subroutine block diagram of fig. 6.

Fig. 12 is a timing chart for explaining the calculation job in the flowchart of fig. 11.

Fig. 13 is a subroutine block diagram illustrating the instability determination process of the adaptive control system of the flowchart of fig. 6.

FIG. 14 is a view for explaining an instability determination process of the flowchart of FIG. 13,

fig. 15 is a view similar to fig. 14, and illustrates an instability determination process in the flowchart of fig. 13.

Fig. 16 is a pulse waveform diagram showing another example of the operation of the adaptive controller similar to that of fig. 8.

Fig. 17 is a pulse waveform diagram showing another example of the operation of the adaptive controller similar to that of fig. 8.

FIG. 18 is a block diagram showing the procedure of embodiment 2 of the apparatus of the present invention.

Fig. 19 is an explanatory diagram showing characteristics of the graph used in the flowchart of fig. 18.

Fig. 20 is an explanatory diagram showing characteristics of the graph used in the block diagram of the program of fig. 18.

Fig. 21 is an explanatory diagram showing characteristics of a graph similar to fig. 20 used in the block diagram of fig. 18.

Fig. 22 is an explanatory diagram showing characteristics of a graph similar to fig. 20 used in the block diagram of fig. 18.

Fig. 23 is an explanatory diagram showing characteristics of a graph similar to fig. 20 used in the block diagram of fig. 18.

FIG. 24 is a block diagram showing the procedure of embodiment 3 of the apparatus of the present invention.

FIG. 25 is a block diagram showing the procedure of embodiment 4 of the apparatus of the present invention.

Fig. 26 is an explanatory diagram showing dead zone characteristics used in the block diagram of fig. 25.

FIG. 27 is a block diagram showing the procedure of embodiment 5 of the apparatus of the present invention.

Fig. 28 is an explanatory diagram showing characteristics of the limiter used in the block diagram of fig. 27.

FIG. 29 is a block diagram showing the procedure of embodiment 6 of the apparatus of the present invention.

Fig. 30 is an explanatory diagram showing characteristics of the diagram used in the block diagram of fig. 29.

FIG. 31 is a block diagram showing the procedure of embodiment 7 of the apparatus of the present invention.

Fig. 32 is a diagram illustrating a job in the block diagram of fig. 31.

FIG. 33 is a block diagram showing the procedure of the 8 th embodiment of the apparatus of the present invention.

FIG. 34 is a block diagram showing the procedure of the 9 th embodiment of the apparatus of the present invention.

FIG. 35 is a block diagram showing the procedure of the 10 th embodiment of the apparatus of the present invention.

Fig. 36 is a signal flow diagram illustrating the operation of the block diagram of fig. 35.

Fig. 37 is an explanatory diagram showing the relationship between TDC and the air/fuel ratio at the merging portion of the exhaust system in the multi-cylinder internal combustion engine.

Fig. 38 is an explanatory diagram showing whether or not the selection of the sampling time for the actual air/fuel ratio is good.

Fig. 39 is a block diagram showing the air/fuel ratio sampling operation performed by Sel-V in the signal flow chart of fig. 36.

Fig. 40 is one of explanatory views of the detecting device shown in the block diagram of fig. 36, and shows an example of modeling the detecting operation of the air/fuel ratio sensor described in the above-mentioned application.

Fig. 41 is a model obtained by discretizing the model shown in fig. 40 with a period Δ T.

Fig. 42 is a signal flow diagram showing a real air/fuel ratio estimator that models the detection operation of the air/fuel ratio sensor.

Fig. 43 is a signal flow diagram showing a model showing the operation of the exhaust system of the internal combustion engine.

FIG. 44 is a data chart showing the air/fuel ratio for 3 cylinders at 14.7: 1 and 1 cylinder at 12.0: 1 for a 4 cylinder engine using the model shown in FIG. 43.

FIG. 45 is a data graph showing the air/fuel ratio at the merge of the model of FIG. 43 given the inputs shown in FIG. 44.

Fig. 46 is a graph comparing a model output value, which is data of an air-fuel ratio at the merging portion of the model of fig. 43 when the input shown in fig. 44 is given in consideration of the response delay of the LAF sensor, with an actual measurement value of the LAF sensor output in the same case.

Fig. 47 is a signal flow chart showing the configuration of a general detection device.

Fig. 48 is a signal flow diagram of the detection apparatus shown in the signal flow diagram of fig. 36, showing the components of the detection apparatus used in the prior application.

Fig. 49 is a signal flow chart showing a configuration in which the model shown in fig. 43 and the detection device shown in fig. 48 are combined.

Fig. 50 is a signal flow diagram showing the air/fuel ratio feedback control in the signal flow diagram of fig. 36.

Fig. 51 is an explanatory diagram showing characteristics of a pulse waveform chart used in the block diagram of fig. 39.

Fig. 52 is an explanatory diagram illustrating sensor output characteristics corresponding to the machine rotation speed and the machine load, explaining the characteristics of fig. 51.

Fig. 53 is a pulse waveform diagram illustrating a sampling operation in the block diagram of fig. 39.

FIG. 54 is a block diagram showing the procedure of the 11 th embodiment of the apparatus of the present invention.

Figure 55 is a block diagram illustrating the actions of figure 54 and the block diagram,

fig. 56 is a subroutine block diagram showing an instability determination operation of the adaptive control system shown in the flowchart of fig. 54.

Fig. 57 is a pulse waveform diagram illustrating the waiting time when calculating the fuel injection amount of the internal combustion engine.

[ best mode for carrying out the invention ]

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

Fig. 1 is a general view schematically showing a fuel injection control apparatus for an internal combustion engine according to the present invention.

In the figure, reference numeral 10 denotes an OHC up to a 4-cylinder internal combustion engine. Air introduced from an air cleaner 14 disposed at the front end of the intake pipe 12 is regulated in flow rate by a throttle valve 16, passes through a surge chamber 18 and an intake straight pipe 20, and flows into the 1 st to 4 th cylinders in sequence through 2 intake valves (not shown). An injection nozzle 22 for injecting fuel is provided in the vicinity of an intake valve (not shown) of each cylinder. The air-fuel mixture in which the injected fuel is integrated with the intake air is ignited and burned by an unillustrated ignition plug in the order of the 1 st, 3 rd, 4 th, and 2 nd cylinders, and drives pistons (not shown).

The burned exhaust gas is discharged to an exhaust straight pipe 24 through 2 exhaust valves (not shown), purified by a catalytic device (three-way catalyst) 28 through an exhaust pipe 26, and discharged to the outside of the machine. The throttle valve 16 is mechanically disconnected from an accelerator pedal (not shown) and controlled to an opening degree corresponding to the stepping amount and the operating state of the accelerator pedal by a pulse motor M. The intake pipe 12 is provided with a bypass passage 32 for bypassing the throttle valve 16 in the vicinity of the position where the throttle valve is disposed.

The internal combustion engine 10 is provided with an exhaust gas recirculation mechanism 100 for recirculating exhaust gas to the intake side.

This is explained with reference to fig. 2. The exhaust gas recirculation passage 121 of the exhaust gas recirculation mechanism 100 has one end 121a communicating with the upstream side of the 1 st catalyst device 28 (not shown in fig. 2) of the exhaust pipe 26 and the other end 121b communicating with the downstream side of the throttle valve 16 (not shown in fig. 2) of the intake pipe 12. An exhaust gas recirculation valve (recirculation gas control valve) 122 for adjusting the amount of exhaust gas recirculation and a volume chamber 121c are provided in the exhaust gas recirculation passage 121. The exhaust gas recirculation valve 122 is a solenoid valve having a solenoid 122a, and the solenoid 122a is connected to a control unit (ECU)34 described later, and the valve opening degree is linearly changed by an output from the control unit 34. The exhaust gas recirculation valve 122 is provided with an upward opening/closing sensor 123 for detecting the opening degree of the valve, and the output of the upward opening/closing sensor is sent to the control unit 34.

Also in communication between the intake system of the engine 10 and the fuel tank 36 is a canister filter 200.

As shown in fig. 3, the canister filter 200 is configured by a vapor supply passage 221, a canister 223 containing an adsorbent 231 therein, and a filtration passage 224 provided between an upper portion of the closed fuel tank 36 and a downstream side of the throttle valve 16 of the intake pipe 12. A two-way valve 222 is installed in the middle of the steam supply passage 221, and a filtration control valve 225, a flow meter 226, and an HC concentration sensor 227 are installed in the middle of the filtration passage 224. The flow meter 226 detects the flow rate of the mixture gas containing the fuel vapor flowing through the purge passage 224, and the HC concentration sensor 227 detects the HC concentration in the mixture gas. As will be described later, the filter control valve (solenoid valve) 225 is connected to the control unit 34 and is controlled by an output signal of the control unit 34 so that the valve opening degree changes linearly.

When the fuel vapor generated in the fuel tank 36 reaches a predetermined set amount, the canister filter opens the positive pressure valve of the two-way valve 222, flows into the canister 223, and is adsorbed and stored by the adsorbent 231. When the valve opening degree of the filter control valve 225 corresponds to the load ratio of the on/off control signal from the control unit 34, the vaporized fuel temporarily stored in the tank 223 is sucked into the intake pipe 12 through the filter control valve 225 together with the outside air sucked from the outside air intake port 232 by the negative pressure action of the intake pipe 12, and is sent to each cylinder. When the fuel tank 36 is cooled by outside air or the like and the negative pressure in the fuel tank increases, the negative pressure valve of the two-way valve 222 opens, and the vaporized fuel temporarily stored in the canister 223 returns to the fuel tank 36.

The internal combustion engine 10 is further provided with a so-called variable valve governor 300 (shown as V/T in fig. 1). As described in japanese patent application laid-open No. h 2-275043, the variable valve-type governor 300 switches the valve-type governor V/T of the machine between 2 kinds of governing characteristics Lov/T, Hiv/T shown in fig. 4 according to the operating conditions such as the machine rotation speed Ne and the intake pressure Pb. However, since they are known per se, their description will be omitted. The switching of the valve timing characteristics includes an operation of stopping one of the 2 intake valves.

In fig. 1, a distributor (not shown) of an internal combustion engine 10 is provided with a crank angle sensor 40 for detecting a crank angle position of a piston (not shown), a throttle opening sensor 42 for detecting an opening of a throttle valve 16, and an absolute pressure sensor 44 for detecting an intake pressure Pb downstream of the throttle valve 16 by an absolute pressure. An atmospheric pressure sensor 46 that detects atmospheric pressure Pa is provided at an appropriate position of the internal combustion engine 10, an intake air temperature sensor 48 that detects the temperature of intake air is provided upstream of the throttle valve 16, and a water temperature sensor 50 that detects the temperature of machine cooling water is provided at an appropriate position of the machine. A valve governor (V/T) sensor 52 (not shown in fig. 1) is provided, and the sensor 52 detects the valve governor characteristics selected by the variable valve governor 300 by oil pressure. In the mutual exhaust system, a wide-range air/fuel ratio sensor 54 is provided at an exhaust system joining portion on the downstream side of the straight exhaust pipe 24 and on the upstream side of the catalyst device 28. The outputs of these sensors are sent to the control unit 34.

Fig. 5 is a block diagram showing a detailed configuration of the control unit 34. The output of the wide-range air/fuel ratio sensor 54 is input to a detection circuit 62, where appropriate linearization is performed to output a detection signal that is composed of a linear characteristic proportional to the oxygen concentration in the exhaust gas over a wide range from lean to rich (hereinafter, this wide-range air/fuel ratio sensor is referred to as "LAF sensor").

The output of the detection circuit 62 is input into the CPU through a multiplexer 66 and an a/D conversion circuit 68. The CPU includes a CPU chip 70, a ROM72, and a RAM74, and more specifically, the output of the detection circuit 62 is a/D converted once for every predetermined crank angle (for example, 15 degrees) and is sequentially stored in one of the buffers in the RAM 74. As shown in fig. 53 described later, 12 buffers are numbered from 0 to 11. Similarly, the analog sensor output of the throttle opening sensor 42 and the like enters the CPU through the multiplexer 66 and the a/D conversion circuit 68, and is stored in the RAM 74.

After the output of the crank angle sensor 40 is shaped by the waveform shaping circuit 76, the output value is counted by the counter 78, and the counted value is input to the CPU. In the CPU, the CPU chip 70 calculates control values as described below in accordance with instructions stored in the ROM72, and drives the nozzles 22 of the respective cylinders by the drive circuit 82. The CPU chip 70 drives the solenoid valve 90 (opening/closing of the bypass passage 32 for adjusting 2 times the air flow rate), the solenoid valve 122 for controlling exhaust gas recirculation, and the solenoid valve 225 for controlling canister filtration through the drive circuits 84, 86, 88. The on-off sensor 123, the flow meter 226, and the HC concentration sensor 227 are not shown in fig. 5.

Fig. 6 is a block diagram showing a program for operating the control device of the present invention.

As shown in the figure, the detected machine rotation speed Ne and intake pressure Pb are read in S10, the process proceeds to S12, it is determined whether the crank is rotating, and if not, the process proceeds to S14, and it is determined whether the fuel supply is cut off. The fuel cut is performed in a specific operation state, for example, when the throttle valve is in a fully closed position and the machine rotation speed is equal to or higher than a predetermined value, the fuel supply is stopped at that time, and the injection amount is controlled in an open loop manner.

When it is determined at S14 that fuel cut has not occurred, the routine proceeds to S16, where a map is searched based on the detected engine speed Ne and intake pressure Pb, and a basic fuel injection amount Tim is calculated. Subsequently, the process proceeds to S18, where it is determined whether or not the activation of the LAF sensor 54 is completed. This determination is made by: the difference between the output voltage of the LAF sensor 54 and the center voltage thereof is compared with a predetermined value (for example, 0.4V), and if the difference is smaller than the predetermined value, it is judged that the startup is completed. When it is determined at S18 that the startup is completed, the routine proceeds to S20, where it is determined whether or not the control is in the feedback control region. When the operating state has changed due to a high rotational speed, a full load, a high water temperature, and the like, the injection amount is subjected to open-loop control. When it is determined at S20 that the detected value is the feedback control region, the routine proceeds to S22, where the detected value of the LAF sensor is read, and then to S24, where the detected air/fuel ratio kact (K) (K: sampling time, the same applies hereinafter) is determined based on the detected value. Subsequently, the process proceeds to S26, where the feedback correction coefficient klaf (k) calculated according to the PID control law is calculated.

The feedback correction coefficient calculated according to the PID control law is calculated as follows.

The control deviation DKAF of the desired air/fuel ratio KCMD and the detected air/fuel ratio KACT is first calculated using the formula DKAF (K) = KCMD (K-d') -KACT (K). In the above equation, KCMD (K-d ') represents the desired air/fuel ratio (where d' represents the latency before KCMD is reflected to KACT, thus meaning the desired air/fuel ratio before the latency control period); KACT (k) represents the detected air/fuel ratio (of this control cycle). In the present specification, it is desirable that both the air/fuel KCMD and the detected air/fuel ratio KACT are equivalence ratios, i.e., expressed in the form of Mst/M = 1/λ (Mst: theoretical air/fuel ratio, M = a/F (a: air consumption, F: fuel consumption, λ: air excess ratio)).

Then, the deviation value is multiplied by a predetermined coefficient to obtain a P term KLAFP (k), an I term KLAFI (k), and a D term KLAFD (k) by the following equations.

And P item: KLAFP (k) = DKAF (k) x kp

Item I: KLAFI (K) = KLAFI (K-1) + DKAF (K) XKI

Item D: KLAFD (K) = (DKAF (K) -DKAF (K-1)). times.KD

Thus, the P term is obtained by multiplying the deviation by the proportional gain Kp; the I term is obtained by multiplying the deviation by an integral gain KI and adding a previous value KLAFI (K-1) of a feedback correction coefficient; the term D is obtained by multiplying the difference between the present value DKAF (K) and the previous value DKAF (K-1) by the differential gain KD. The gains Kp, KI, and KD are obtained from the machine rotation speed and the machine load, and are specifically designed to be obtained by searching a map from the machine rotation speed Ne and the intake pressure Pb. Finally, the obtained values are added to obtain the current value KLAF (k) of the feedback correction coefficient determined by the PID control law. In the case of klaf (k) = klafp (k) + klafi (k) + klafd (k), since the feedback correction coefficient is obtained by multiplicative correction, the deviation component 1.0 is included in the I term klafi (k) (that is, the initial value of klafi (k) is 1.0).

Then, at S28 in the block diagram of fig. 6, the feedback correction factor kstr (k) is calculated by the adaptive control method. The process of calculating the feedback correction factor kstr (k) using the adaptive control law will be described in detail below.

Subsequently, the routine proceeds to S30, where the desired air/fuel ratio correction coefficient kcmdm (k) and the other correction coefficient KTOTAL (product of various correction coefficients by multiplication of water temperature correction and the like) are multiplied by the obtained basic fuel injection Tim to obtain a required fuel injection quantity T required for the internal combustion engine_cyl (k). As described above, in this control, the desired air/fuel ratio is expressed by the equivalence ratio and is also used as a correction coefficient for the fuel injection amount. Specifically, since the charging efficiency of intake air differs depending on the heat of vaporization, the desired air/fuel ratio is corrected by the appropriate characteristics to obtain the desired air/fuel ratio correction coefficient KCMDM.

Proceeding next to S32, the fuel injection quantity T will be requested_cyl (k) is multiplied by the feedback correction coefficient klaf (k) obtained at S26 or by the feedback correction coefficient kstr (k) obtained at S28, and the addition term TTOTAL is added to the product of the multiplication, thereby determining the output fuel injection amount tout (k). The addition term TTOTAL is a sum of correction coefficients by an addition value such as air pressure correction (the dead time of the injector is added when the output fuel injection amount Tout is output, and is therefore not included in TTOTAL).

Subsequently, the process proceeds to S34, where a sticking coefficient map is searched for based on the machine coolant temperature and the like, and the determined output fuel injection amount Tout (k), that is, the intake pipe wall surface sticking correction of the output fuel injection amount Tout (k) is performed using the obtained sticking coefficient (the value after sticking correction is Tout-f (k)). The correction of the wall sticking of the intake pipe itself is not directly related to the gist of the present invention, and therefore, the description thereof will be omitted. Then, the routine proceeds to S36, where the corrected output fuel injection amount Tout-f (k) is output, and the routine ends.

If the determination at S18-S20 is negative, the routine proceeds to S38, where the basic fuel injection amount tim (k) is multiplied by the desired air/fuel ratio correction coefficient kcmdm (k) and various correction coefficients KTOTAL, the product is added to the correction addition term TTOTAL to determine the output fuel injection amount Tout, and then the routine proceeds to S34, where the routine proceeds to S40 if the determination at S12 is that the crankshaft is rotating, the routine proceeds to S42 after the fuel injection amount Ticr at the time of cranking is retrieved, and the output fuel injection amount Tout is calculated as a starting model. When it is determined at S14 that the fuel cut is performed, the routine proceeds to S44, where the output fuel injection amount tout (k) is set to zero.

The following describes the process involved in calculating the feedback correction factor kstr (k) using adaptive control at S28 in the block diagram of fig. 6.

Figure 7 is a block diagram that further functionally represents its actions.

The illustrated apparatus is in the context of adaptive control techniques previously proposed by the applicant. The device is composed of an adaptive controller and an adaptive (control) parameter adjusting mechanism for adjusting adaptive (control) parameters (vectors) of the adaptive controller, wherein the adaptive controller is composed of an STR (self-adjusting regulator) controller. The STR controller receives the expected value and the controlled variable (plant output) of the fuel injection quantity control feedback system, and calculates the output by receiving the coefficient vector recognized by the adaptive parameter adjusting means.

In such adaptive control, one of the adjustment rules (mechanisms) of the adaptive control is a parameter adjustment rule proposed by i.d. landau et al. The adaptive control system is converted into an equivalent feedback system consisting of a linear section and a nonlinear section, and an integral inequality of POPOPOV related to input and output is established for the nonlinear section, so that the linear section is accurate by determining an adjustment rule, and the stability of the adaptive control system is ensured. That is, in the parameter adjustment rule proposed by Landau et al, the adjustment rule (adaptive rule) expressed in the form of a recursive formula is at least either of the above-described POPOV hyperstabilization theory and the Lyapunov direct method, and therefore, the stability thereof is ensured.

The methods are described in, for example, "computer" (Corona Co., Ltd.) No. 27, pages 28 to 41, "Automation Manual" (Ohm Co., Ltd.) pages 703 to 707, "A. D. Landau" university Time Explicit Model Reference Adaptive Techniques-teaching "(Automatica journal 10, pages 353 to 378, 1974, I.D. Landau)" unity of differentiation Time Explicit Model Reference Adaptive controls "(Automatica journal 17, pages 593 to 611, 1981, I.D. Landau et al), and" binding Reference Adaptive controls and storage of Tuning rules "(Automatica journal 18, page 84, published by Australia corporation), all of which are well-known technologies.

The adaptive control technique of the example of the figure is described below using the Landau et al regulation rule. Landau et al regulation law in which the transfer function B (Z) of a discrete system control object^-1)／A(Z^-1) When the polynomial of the denominator and the numerator are expressed by the formulas 1 and 2, the adaptive parameter theta (k) identified by the parameter adjusting mechanism (and <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) similarly, the same applies hereinafter) is expressed by a vector (transposed vector) as in equation 3. The input ξ (k) to the parameter adjustment mechanism is determined by equation 4. Where m =1, n =1, d =3, i.e. the exemplified device is in the form of a linear system and has a latency of 3 control cycles.

A(z^-1)=1+a₁z^-1+···+a_nz^-nA 1. formula

B(z^-1)=b₀+b₁z^-1+···+b_mz^-mA. formula 2

Wherein, the adaptive parameter represented by formula 3 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> By a scalar quantity determining the gain $\widehat{\mathrm{bo}}$ ^-1(K) Control element expressed by operation amount $\widehat{\mathrm{BR}}$ (Z^-1K) and control element expressed by control amount $\hat{S}$ (Z^-1And K) are respectively expressed as formulas 5 to 7.

Parameter adjusting mechanismAfter identifying and estimating the scalar quantities and the coefficients of the control factors, the coefficients are used as adaptive parameters expressed by the above equation 3 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> And sent to the STR controller. Calculating adaptive parameters by using the operation amount U (i) and the control amount y (j) of the device for the parameter adjusting mechanism, wherein (i, j includes past values) <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The deviation between the desired value and the controlled variable is made zero. In particular, adaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Calculated using equation 8. In equation 8, Γ (k) is a gain matrix ((m + n + d) -order square matrix) e) that determines the recognition and estimation speed of the adaptive parameter^*(k) The signals are signals representing recognition and estimation errors, and are represented by recursive formulas such as formula 9 and formula 10, respectively.

Lambda in factor 9₁(k)、λ₂(k) Various specific algorithms can be given according to different selection methods. For example, let λ₁(k)=1，λ₂(k) If = λ (0 < λ < 2), there is a decreasing gain algorithm, (λ =1 is the least squares method), let λ be₁(k)=λ₁(0＜λ₁＜1)、λ₂(k)=λ₂(0＜λ₂< lambda), there is a variable gain algorithm (lambda)₂The addition least squares method when = 1). When lambda is₁(k)／λ₂(k)=σ、λ₃As shown in equation 11, if λ 1(k) = λ 3, there is a fixed trajectory algorithm. In addition, when λ 1(k) =1 and λ 2(k) =0, there is a fixed gain algorithm. In this case, as is clear from equation 9, Γ (K) = Γ (K-1), and therefore Γ (K) = Γ becomes a constant value. A decreasing gain algorithm, a variable gain algorithm, a fixed gain algorithm, and a fixed trajectory algorithm, whichever is appropriate for the fuelTime varying devices of injection and air/fuel ratio, etc.

In fig. 7, the STR controller (adaptive controller) and the adaptive parameter adjusting mechanism are disposed outside the fuel injection amount calculating system, and act to adaptively match the detected air/fuel ratio KACT (k) with the desired air/fuel ratio KCMD (k-d ') (d' is the waiting time until the KCMD is reflected as KACT as described above) and calculate the feedback correction coefficient kstr (k). That is, the STR controller accepts a coefficient vector adaptively identified by an adaptive parameter adjustment mechanism <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) And is shown as a feedback compensator, which is made to coincide with the desired air/fuel ratio KCMD (k-d'). The calculated feedback correction coefficient kstr (k) is multiplied by the requested fuel injection quantity T_cyl (k), the corrected fuel injection quantity is supplied to the control apparatus (internal combustion engine) by attaching the correction compensator as the output fuel injection quantity tout (k).

Thus, the feedback correction coefficient kstr (k) and the detected air/fuel ratio kact (k) are obtained and input to the adaptive parameter adjusting mechanism, where the adaptive parameter θ (k) is calculated and input to the STR controller. The desired air/fuel ratio KCMD (k) is provided as an input to the STR controller so that the detected air/fuel ratio KACT (k) is matched with the desired air/fuel ratio KCMD (k) to calculate the feedback correction factor KSTR (k) by a recurrence formula,

the feedback correction coefficient kstr (k) is obtained by equation 12.

On the other hand, the detected air/fuel ratio KACT and the desired air/fuel ratio kcmd (k) are also inputted to a controller according to the PID control principle (the controller is described in S26 in the block diagram of fig. 6, and PID is shown in the figure), and in order to eliminate the deviation between the detected air/fuel ratio and the desired air/fuel ratio at the merging portion of the exhaust system, the 2 nd feedback correction coefficient klaf (k) is calculated according to the PID control principle. Either one of the feedback correction coefficient KSTR calculated by the adaptive control principle and the feedback correction coefficient KLAF calculated by the PID control principle is used for calculating the fuel injection amount by the switching mechanism 400 shown in fig. 7. As described later, when the operation of the adaptive control system (STR controller) is determined to be unstable or when the operation is outside the adaptive region of the adaptive control system, the feedback correction coefficient klaf (k) calculated according to the PID control rule is used instead of the feedback correction coefficient kstr (k) calculated according to the adaptive control rule.

As shown in fig. 57, when controlling the fuel injection amount of the internal combustion engine, the injection amount is calculated, and it takes time corresponding to the time when the calculated fuel is compressed, exploded, and exhausted in the cylinder. In addition, the time is longer in consideration of the time when the exhaust gas reaches the LAF sensor or the detection delay of the sensor itself, and the calculation time required to calculate the amount of fuel actually drawn into the cylinder from the detected value. Therefore, in the fuel injection amount control of the internal combustion engine, there is inevitably a waiting time. Assuming 1 cylinder as an example, if the waiting time is 3 combustion cycles as described above, the number of TDCs is 12TDC when the internal combustion engine is 4 cylinders as shown in fig. 8. The "combustion cycle" described herein is a 4-stroke cycle consisting of intake, compression, combustion, and exhaust, and corresponds to 4TDC in this embodiment.

In the adaptive controller (STR controller), as can be seen from equation 3, the number of factors of the adaptive parameter θ (k) is m + n + d, and is proportional to the waiting time d. As described above, assuming that the waiting time is 3, when the STR controller and the adaptive parameter adjusting means are operated simultaneously with the TDC so as to correspond to the operating state that changes every moment, the number of factors of the adaptive parameter θ (k) is set to m = n =1, as shown in fig. 8, d =12(3 combustion cycles × 4TDC), and m + n + d = 14. As a result, the gain matrix Γ is calculated as a 14 × 14 matrix, the amount of calculation increases, the load on the vehicle-mounted computer increases, and the general-performance vehicle-mounted computer cannot complete the calculation within 1TDC as the machine rotation speed increases. Also, as described above, an increase in the number of times of waiting time may cause deterioration in controllability.

The illustrated fuel injection control device for an internal combustion engine can maximally correspond to the operating state that changes every moment, and reduce the amount of matrix calculation to reduce the load on the vehicle-mounted computer. Specifically, as shown in fig. 9, the parameter adjusting means inputs the output of the control device in a combustion cycle, more specifically, in synchronization with only a predetermined crank angle (TDC or the like) of a specific cylinder (1 st cylinder or the like), and calculates the adaptive parameter θ described above.

As can be seen from FIG. 9, the adaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The calculation of (2) is performed at a predetermined crank angle (TDC, etc.) of all cylinders. The STR controller operates in synchronization with a predetermined crank angle (TDC, etc.) of all cylinders to calculate a feedback correction coefficient, which is similar to the configuration shown in fig. 8.

Therefore, for example, when the engine is operated in synchronization with only a predetermined crank angle of a specific cylinder in a combustion cycle (fuel control cycle), the number of factors of the adaptive parameter θ is m + n + d =5 with d =3, and the calculation of the gain matrix F is reduced from 14 × 14 to 5 × 5, so that the load on the vehicle-mounted computer is reduced and the calculation can be completed within 1 TDC. As described above, the waiting time of the control target is long, and the controllability is generally deteriorated as compared with the case where the waiting time is short, and this is particularly significant in the adaptive control. The structure can greatly reduce waiting time and improve controllability.

Specifically, the above-described effects can be achieved by using the control period K of equations 1 to 12 for each cylinder. In the case of a 4-cylinder internal combustion engine, equation 4 may be changed to equation 13, equation 8 may be changed to equation 14, equation 9 may be changed to equation 15, and equations 10 to 12 may be changed to equations 16 to 18. ζ T (k) = [ u (k) u (k-4) u (k-8) u (k-12) y (k)]9 … formula 13

Therefore, even with the configuration shown in fig. 9, the adaptive parameters can be calculated by using the control period (operation period) for each TDC of all the cylinders, that is, in synchronization with the TDCs of all the cylinders, and the order of the matrix or vector used for calculation can be reduced, as in the configuration shown in fig. 8. Of course, the control cycle is used for each individual cylinder, and the control cycle K in expressions 1 to 12 is K = the number of cylinders × K, and the same operation can be performed even if the structure is such that each cylinder has an internal variable. Here, K represents the number of combustion cycles, and K represents TDC. Fig. 10 is a diagram re-writing the structure shown in fig. 8 with emphasis placed on the STR controller and the parameter adjusting mechanism. In fig. 10, if the operation cycle m × TDC of the STR controller and the operation cycle n × TDC of the parameter adjusting means are set to m = n =1, the configuration shown in fig. 8 and 9 is obtained. Here, if the input cycle of the parameter adjustment means is synchronized with the TDC and the waiting time is d =2, the configuration of fig. 8 is obtained. On the other hand, if the input cycle of the parameter adjustment means is synchronized with the combustion cycle and the waiting time is d =3, the configuration of fig. 9 is obtained.

However, since the device output is synchronized with the combustion cycle and input to the parameter adjustment mechanism for calculation (operation) corresponds to the synchronous operation at a predetermined crank angle of the specific cylinder, the device output is strongly affected by the exhaust gas air/fuel ratio of the specific cylinder. As a result, when the exhaust gas air/fuel ratio of the specific cylinder is lean and the exhaust gas air/fuel ratio of the remaining cylinders is rich, for example, during the control of the stoichiometric air/fuel ratio, the adaptive controller (STR controller) adjusts the operation amount in the rich direction to operate in accordance with the desired value, and the air/fuel ratios of the remaining cylinders become denser.

In order to solve this problem, the illustrated apparatus reduces the number of adaptive parameter factors and the amount of matrix calculation by inputting the plant output operation and the combustion cycle in synchronization with each other to the parameter adjustment mechanism, as will be described later. And is not strongly influenced by the exhaust gas air/fuel ratio of the specific cylinder. In order to achieve the above object, the following operation is performed. The internal variables can also be acted upon in the same way. Here, K represents the number of combustion cycles, and K represents TDC. Fig. 10 is a diagram re-writing the structure shown in fig. 8 with emphasis placed on the STR controller and the parameter adjusting mechanism. In fig. 10, if the operation cycle m × TDC of the STR controller and the operation cycle n × TDC of the parameter adjusting means are set to m = n =1, the configuration shown in fig. 8 and 9 is obtained. Here, if the input cycle of the parameter adjustment means is synchronized with the TDC and the waiting time is d =2, the configuration of fig. 8 is obtained. On the other hand, if the input cycle of the parameter adjustment means is synchronized with the combustion cycle and the waiting time is d =3, the configuration of fig. 9 is obtained.

However, since the device output is synchronized with the combustion cycle and input to the parameter adjustment mechanism for calculation (operation) corresponds to the synchronous operation at a predetermined crank angle of the specific cylinder, the device output is strongly affected by the exhaust gas air/fuel ratio of the specific cylinder. As a result, when the exhaust gas air/fuel ratio of the specific cylinder is lean and the exhaust gas air/fuel ratio of the remaining cylinders is rich, for example, during the control of the stoichiometric air/fuel ratio, the adaptive controller (STR controller) adjusts the operation amount in the rich direction to operate in accordance with the desired value, and the air/fuel ratios of the remaining cylinders become denser.

To solve this problem, as will be described later, the illustrated apparatus diagram inputs the plant output operation and the combustion cycle in synchronization with each other to the parameter adjustment mechanism, thereby reducing the number of adaptive parameter factors and reducing the amount of matrix calculation. And is not strongly influenced by the exhaust gas air/fuel ratio of the specific cylinder. In order to achieve the above object, the following operation is performed.

The parameter adjusting means operates in synchronization with the combustion cycle, that is, in synchronization with the crank angle specified for a specific cylinder among the 4 cylinders, and the control amount y (k) is obtained as an average (e.g., a simple average) of the detected air/fuel ratios kact (k) at every TDC specified crank angle for each cylinder during the combustion cycle, and is input to the parameter adjusting means, so that the control amount is not greatly affected by the exhaust gas air/fuel ratio of the specific cylinder.

Further, the adaptive parameter calculated by the parameter adjusting means is adjusted for each predetermined crank angle <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The feedback correction factor kstr (k) calculated by the STR controller is also averaged, thus making it more immune to the exhaust gas air/fuel ratio of the particular cylinder.

The parameter adjusting means operates in synchronization with the combustion cycle, that is, operates in synchronization with a crank angle specified for a specific cylinder among the 4 cylinders, and obtains the control amount y (k) as an average (e.g., a simple average) of the detected air/fuel ratios kact (k) at each TDC specified crank angle (e.g., at each TDC) for each cylinder during the combustion cycle, and inputs the control amount y (k) to the parameter adjusting means.

Further, the adaptive parameter calculated by the parameter adjusting means is adjusted for each predetermined crank angle <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The feedback correction factor kstr (k) calculated by the STR controller is also averaged so that it is more immune to the exhaust gas air/fuel ratio of the particular cylinder.

Fig. 11 is a subroutine block diagram showing the calculation job.

As shown in fig. 11, it is first determined at S100 whether or not the device is within a predetermined operation area. The "operating region" as used herein means a low-speed rotation region including idling. When it is determined at S100 that the cylinder is not in the predetermined operating region, the routine proceeds to S102, and the calculation device outputs a control amount y (K) which is an average value KACTAVE of the air/fuel ratio KACT (K) calculated for the cylinder at S24 in fig. 6 at this time, the air/fuel ratio KACT (K-1) calculated for the previous time in the previous combustion cylinder, the air/fuel ratio KACT (K-2) calculated for the previous time in the previous combustion cylinder, and the air/fuel ratio KACT (K-3) calculated for the previous time in the previous combustion cylinder. That is, the control period is set back three periods, and the average value of the air/fuel ratio calculated during one combustion period is obtained as the control amount y (k) for 4 cylinders including the cylinder. This approach can reduce the effect of the exhaust gas air/fuel ratio of a particular cylinder.

Subsequently, the process proceeds to S104, where, as shown at the end of FIG. 7, the parameter adjusting means calculates the adaptive parameter according to equation 3 from the control quantity y (k) obtained above and the like <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) And then input to the STR controller.

Then, the process proceeds to S106, where adaptive parameters including the current calculation are calculated <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Calculated for the first 3 control cycles of (1 combustion cycle period) <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k)、 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (K-1)、 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (K-2) and <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> average value of (K-3), e.g. simple average value AVE- <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) In that respect I.e. adaptive parameters not on the input side of the parameter adjustment means, but on its output side <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Finding 4 control cycles (1 combustion cycle) corresponding to 4 cylinders <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> And then input to the STR controller. In this way, even if 4 cylinders are input to the STR controller <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The object of reducing the influence of the exhaust gas air/fuel ratio of the specific cylinder can also be achieved. In addition, as shown in the formula 3, <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> since the average value is calculated in the form of a vector, the average value can be obtained by calculating the average value of the factors S0, r1, r2, r3, and b0 of the vector. It is also possible to calculate the average value of any one of the factors, and it is preferable to calculate the average value of θ by calculating the amount of change in proportion to the other factors. S106 schema the meaning of <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Formula for the average value.

Then, the process proceeds to S108, where the feedback correction coefficient KSTR (K) is calculated in the STR controller according to equation 12 based on the input value, and then to S110, where the average value of KSTR (K), KSTR (K-1), KSTR (K-2), and KSTR (K-3) during 1 combustion cycle, which is the calculated value of the first 3 control cycles including the feedback correction coefficient KSTR (K) calculated in the previous calculation, for example, is calculated as the simple average value avekstr (K). That is, the purpose of reducing the influence of the exhaust gas air/fuel ratio of a specific cylinder can also be achieved by averaging the KSTR values over 4 control cycles (1 combustion cycle) corresponding to 4 cylinders, not by the parameter adjusting mechanism, but by the STR controller that outputs a control input (i.e., a feedback correction coefficient of the fuel computing system).

When it is determined in S100 that the machine is in the predetermined operation region, the process proceeds to S112, and y (k) is calculated. That is, for this cylinder, the currently calculated equivalence ratio kact (k) obtained at S24 in fig. 6 is directly used as the control amount (plant output). The process then proceeds to S114, where adaptive parameters are calculated in the same manner as in S104 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Then, the process proceeds to S116, and the feedback correction coefficient kstr (k) is calculated in the same manner as in S108.

Since the average value of the air/fuel ratios of all the cylinders is obtained and input to the parameter adjusting mechanism as the control amount y (k), the average value is not greatly affected by the equivalence ratio of the specific cylinder (for example, the 1 st cylinder), more specifically, the exhaust gas air/fuel ratio. In addition, as for the STR controller output, the value of 4 control cycles including the latest value u (k) = kstr (k) is also used, and the signal vector ξ is obtained and input to the parameter adjustment mechanism, so that the influence of the exhaust gas air/fuel ratio of the specific cylinder is further reduced.

In addition, the adaptive parameter is not applied to the input side of the parameter adjusting mechanism but to the output value of the parameter adjusting mechanism <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Determining 4 control cycles (1 combustion cycle) for 4 cylinders <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> And input to the STR controller, the purpose of reducing the effect of the air/fuel ratio of the exhaust gas of a particular cylinder is also achieved due to its smoothing effect. Further, since the STR controller which outputs KSTR (feedback correction coefficient of fuel calculation system) is allowed to obtain the average value of KSTR for 4 control cycles (1 combustion cycle) corresponding to 4 cylinders even if it is not for the parameter adjustment mechanism, exhaust gas of a specific cylinder can be reduced similarlyThe air/fuel ratio of.

In S100, it is determined whether or not the machine is in a predetermined operation region, specifically, in a low-speed rotation region including idling, and if the determination is affirmative, no problem occurs because the average value is not calculated. That is, since the control period is long at low-speed rotation, the response delay of the LAF sensor can be ignored. On the contrary, since the phases of the detected air/fuel ratio kact (k) and the average value KACTAVE thereof are deviated as shown in fig. 12, the same phenomenon as the variation in the waiting time of the control system occurs is generated. Therefore, if adaptive control is performed using kactave (k) with a phase shift, adverse effects such as oscillation may occur. Therefore, when a low rotation state such as idling is affected by this, the smoothing effect is not continued.

In the above, the adaptive parameters calculated by S106 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Is not used to calculate the recognition error signal e shown in equation 10^*. I.e. the identification error signal e^*Is a function of evaluating the magnitude of error between the detected air/fuel ratio and the desired air/fuel ratio if the determined AVE- <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The calculation using the formula 10 sometimes results in incorrect errors, so AVE- <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) It is beneficial to use only in the calculation of equation 8, not in equation 10.

In addition, in the above, the air/fuel ratio is set in S102, S106, S110, <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The average values of kstr (k) are all used, although any one or suitably 2 may be used. In addition, the average at the time of starting the computing machine or when the computation of the STR controller is restartedWhen there is no past value, an appropriate predetermined value may of course be adopted.

In-process adaptive parameter calculation <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Or the average value of the correction factor kstr (k), it is not necessary to input these values to the parameter adjustment mechanism. This is because the feedback correction coefficient kstr (k) calculated by the STR controller with the average value of the adaptation parameter θ (k) is already a value that is not greatly affected by the specific cylinder exhaust gas air/fuel ratio. Likewise, the average value of the feedback correction coefficient kstr (k) calculated by the STR controller is itself a value that is not greatly affected by the specific cylinder exhaust gas air/fuel ratio.

The selection of the feedback correction coefficient shown in S32 of the block diagram of fig. 6 will be described below.

Fig. 13 is a subroutine block diagram showing the procedure.

As shown in fig. 13, when it is determined at S200 whether or not the machine is within the applicable region of the adaptive control system, for example, in the combustion unstable operation region such as an extremely low water temperature, since the air/fuel ratio kact (k) cannot be accurately obtained, the process proceeds to S210, and the output fuel injection amount tout (k) is calculated using the feedback correction coefficient klaf (k) obtained by the PID control principle. If the adaptive parameter is determined to be in the applicable region in S200, the process proceeds to S202, where the adaptive parameter is used <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The stability of the adaptive control system is judged by the factors.

Specifically, the transmission characteristic of the feedback correction coefficient kstr (k) calculated by the STR controller is shown in equation 19.

In the formula, the transfer characteristics of KSTR (k) and KACT (k) are shown in formula 20, assuming that the correction is correct and the fuel-bearing computing system is in a disordered state.

KACT(z^-1)=z^-3KSTR(z^-1) … formula 20

The transfer function from kcmd (k) to the correction coefficient kstr (k) is shown in equation 21.

Since bo is a scalar quantity determining the gain and cannot be 0 or a negative number, the denominator function f (z) = b of the transfer function of equation 21₀Z³+r₁Z²+r₂Z+r₃+S₀Is one of the functions shown in fig. 14. Whether the real root is within the unit circle, that is, whether f (-1) < 0 or f (1) > 0 is determined as shown in fig. 15, and if yes, the real root is within the unit circle, and whether the system is stable can be easily determined.

Then, the process proceeds to S204, whether the adaptive control system is unstable is judged, and if yes, the process proceeds to S206, and the adaptive parameter vector is adjusted <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> And returning to the initial value. Thus, the stability of the system can be recovered. Subsequently, the process proceeds to S208, where the gain matrix Γ is corrected. The gain matrix Γ is used to determine the rate of change (convergence) of the parameter adjustment mechanism, and therefore this correction is performed to slow down the rate of convergence. Here, the factors of the gain matrix Γ are replaced with small values. This also restores the stability of the system. Subsequently, the process proceeds to S210, where, as shown in the figure, since the adaptive control system is unstable, the correction coefficient klaf (k) calculated according to the PID control principle is used as the feedback correction coefficient, and is multiplied by the required fuel injection amount T_cyl (k), the product is added to the addition term TTOTAL to determine the output fuel injection quantity tout (k).

When it is determined at S04 that the adaptive control system is not unstable, the routine proceeds to S212, where the output fuel injection amount tout (k) is calculated using the correction coefficient kstrz (k) calculated according to the adaptive control method as a feedback correction coefficient, as shown in the drawing. At this time, if the average value of the feedback correction coefficient KSTR is found at S110 in the block diagram of fig. 11, it is a matter of course to calculate tout (k) using the average value.

In the block diagram of fig. 7, the output u (k) of the switching mechanism 400 is input into the STR controller and the parameter adjustment mechanism. This is to calculate the feedback correction coefficient KSTR in accordance with the adaptive control principle even when the feedback correction coefficient KLAF calculated by the PID control principle is selected.

With the above-described configuration, the parameter adjusting means reduces the number of adaptive parameter factors to 5, the calculation of the Γ matrix to 5 × 5, and the load on the vehicle computer, regardless of the operation of all the cylinders per TDC, and the load on the vehicle computer can be reduced to 1 with a general performance^#The calculation is completed between the TDCs. On the other hand, the STR controller can also calculate the feedback correction coefficient KSTR for each TDC in all cylinders and change it for each TDC in all cylinders, which greatly corresponds to the change in the operating state. In addition, the controllability can be improved because the waiting time is greatly reduced.

Further, the parameter adjusting means operates every combustion cycle as seen from the individual cylinder, and as a result, it always operates at a predetermined crank angle of a specific cylinder, for example, the 1 st cylinder, but since it is made to determine the average value of the detected air/fuel ratios (controlled amounts) of all the cylinders including the remaining cylinders during the combustion cycle, the average value is inputted into the parameter adjusting means, or the adaptive parameter is determined <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Or the average value of the feedback correction coefficient KSTR output as the STR controller is found and used, there is no inconvenience of reflecting only the combustion state of the specific cylinder strongly.

That is, when the feedback correction coefficient KSTR is determined based on the specific cylinder control amount, for example, when the air/fuel ratio of the 1 st cylinder is rich and the air/fuel ratio of the other cylinders is lean, the feedback correction coefficient KSTR should correct the air/fuel ratio direction so that the air/fuel ratio of the other cylinders is made lean, but the above problem does not occur because the average value of all the cylinders is used.

For further simplification, it is also possible to calculate the adaptive parameter not in synchronization with each TDC of all the cylinders but in synchronization with the combustion cycle of the specific cylinder, i.e., once every 4 TDCs, as shown in fig. 16 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> At the STR controller, the adaptive parameters are adjusted <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> For the same value as the number of cylinders. (this corresponds to the case where m =1 and n =4 in fig. 10).

The method is particularly effective for the case that the revolution of the machine is increased and the calculation time is reduced. At high speed of rotation, due to the adaptive parameters required by each cylinder <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Is reduced even if the adaptive parameters of the specific cylinder are adjusted <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Since the deterioration of controllability is small for all cylinders including other cylinders, the calculation time can be shortened without deteriorating the controllability.

As shown in fig. 17, if the STR controller is also operated once every 4TDC in synchronization with the combustion cycle, the configuration can be further simplified. Although the control accuracy is lowered, there is an effect of a certain program (equivalent to the case where m = n =4 in fig. 10).

Fig. 18 is a block diagram showing a procedure of embodiment 2 of the apparatus of the present invention, and is related to setting of gain matrix Γ used for calculation of feedback correction parameter KSTR.

As shown in the above formulas 1 to 12, feedback correction is performedThe gain matrix Γ (k) is used in the calculation of the coefficients KSTR. Example 2 in formula 9₁=1、λ₂=0, i.e. in the case of using a fixed gain algorithm, the off-diagonal factors of the gain matrix Γ (k) are all made 0 in order to shorten the computation time and to be easy to adjust.

For the sake of convenience of explanation, the internal variable Γ is taken_ξThe calculation of (k-d) is explained as an example. In embodiment 1 using a 5 x 5 gain matrix Γ, the calculation of Γ is performed using equation 22, using 25 multiplications and 20 additions.

When all the non-diagonal factors of the gain matrix Γ are set to 0 in this equation, equation 23 can be expressed, and the calculation can be shortened to 5 multiplications.

In addition, since all the non-diagonal factors of the gain matrix Γ are set to 0, the adaptive parameters are performed <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Is calculated as shown in equation 24.

As a result, the matrix factor g₁₁、g₂₂、g₃₃、g₄₄、g₅₅Is to make the adaptive parameter <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Is a value corresponding to only one factor of ξ (k), and is independently adjustable. If the off-diagonal factor of the gain matrix Γ is not 0, then it can be seen from equations 22 and 24 thatAdaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) As shown in equation 25, in order to determine <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The change speed of one factor of (b) needs to take into account 5 variables corresponding to all the factors of ξ (k-d), and is difficult to adjust. By setting all the off-diagonal factors of the gain matrix Γ to 0, the calculation time can be shortened and the adjustment can be facilitated.

The inventors have just made experiments to determine g in the gamma matrix₁₁～g₅₅Setting several of the 5 adjustment elements to the same value, and appropriately setting adaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The ratio of the varying speeds of the factors indicates that the controllability becomes the best. For example, set to g₁₁=g₂₂=g₃₃=g₄₄And (= g). Thus, the adjustment factor can be reduced to only g and g₅₅= Γ, the adjustment steps can be reduced, and the internal variable ξ^TThe calculation of (k-d) Γ ξ (k-d) is shown in equation 26, and the multiplication is 12 times.

In contrast, in g₁₁～g₄₄When different values are taken, the above calculation is increased to 15 times as shown in equation 27. Zeta^T(k-d)Γζ(k-d)=g₁₁u(k-d)²+…+g₅₅y(k-d)²… formula 27

From the above, g₁₁～g₅₅Several of them are set to the same value, the number of adjustment factors can be reduced, and the calculation time can be further shortened. In addition, since the adaptive parameters can be set appropriately <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The controllability is also good because of the change speed ratio of each factor. At this time, if g is assumed₁₁=g₂₂=g₃₃=g₄₄=g₅₅The effect is of course most pronounced.

In addition, taking the operation region in which the combustion is unstable and the plant output is also unstable as an example, the above-mentioned g is reduced₅₅Can inhibit S₀(k) Oscillations of (c), etc. In this way, the advantage of facilitating the adjustment of the control characteristic is made more remarkable by setting the non-diagonal factor of the gain matrix Γ to 0. In addition, the gain matrix is replaced according to different operation areas, so that the machine always has the best controllability.

In this case, g₁₁～g₅₅The RAM74 in the control unit 34 is stored in accordance with the operation state. Specifically, in addition to the operating state, information corresponding to the operating state of the machine control device such as the canister filter and the exhaust gas recirculation is stored. At this time, g₁₁～g₅₅May be all the same value, all different values, or several of them the same value. Also in this case, if the capacity of RAM74 is large enough or the computation time is sufficient, the off-diagonal factor of gain matrix Γ may be employed as well.

For the above reasons, the following describes embodiment 2 of the apparatus of the present invention with reference to the flowchart of fig. 18.

First, the operating state of the exhaust gas recirculation mechanism or the cylinder filter of the machine operation parameters such as the machine rotation speed Ne and the intake pressure Pb is read in S300, the process proceeds to S302, whether or not the engine is in the idling region is determined, and when the determination is affirmative, the process proceeds to S304, and a Γ diagram for idling is searched. If it is determined in S302 that the vehicle is not in the coasting area, the routine proceeds to S306, where it is determined whether the variable valve adjusting mechanism is operating at the Hi-valving characteristic, and if the determination is affirmative, the routine proceeds to S308, where a Γ map for Hi-valving is retrieved, and if the determination is negative, the routine proceeds to S310, where a Γ map for L0-valving is retrieved.

L shown in FIG. 19₀Characteristics of gamma diagram for speed regulationAnd (4) sex. As shown in the figure, matrix factors g11 to g55 are retrieved according to the machine rotation speed Ne and the intake pressure Pb. In addition, Γ diagrams for coasting and Hi valve speed control also have the same characteristics. In addition, since the values of the gain matrix Γ are searched for from the intake pressure Pb representing the machine load, the most appropriate values of the gain matrix can be obtained even in a deceleration state or the like in which the machine load abruptly changes.

Subsequently, the process proceeds to S312, where it is determined whether or not EGR (exhaust gas recirculation mechanism) is operating, and if the determination is affirmative, the process proceeds to S314, where the gain matrix Γ is corrected based on the fuel correction coefficient KEGRN corresponding to the exhaust gas recirculation rate. More specifically, the correction coefficient K Γ EGR is calculated by searching a map showing the characteristics of the fuel correction coefficient KEGRN corresponding to the exhaust gas recirculation ratio in fig. 20, and the obtained correction coefficient K Γ GRR is multiplied by the gain matrix Γ to correct the correction. The reason why the gain matrix Γ is corrected based on the fuel correction coefficient KEGRN corresponding to the exhaust gas recirculation rate is that the correction coefficient K Γ GRR decreases and the disturbance increases as the exhaust gas recirculation amount increases, as shown in the figure, and therefore, in order to improve the stability of the adaptive control system, the gain matrix is made smaller as the fuel correction coefficient KEGRN corresponding to the exhaust gas recirculation rate decreases.

The correction coefficient corresponding to the exhaust gas recirculation rate KEGRN is multiplied by the fuel injection amount, and is set to 0.9 or the like, for example. However, the present invention is not intended to determine the recirculation rate, and the determination of the exhaust recirculation rate is described in Japanese patent application laid-open No. 6-294014, which has been proposed by the present applicant, and therefore the description thereof will be omitted.

Subsequently, the process proceeds to S316, where it is determined whether or not the canister filter is operating, and if the determination is affirmative, the process proceeds to S318, where the gain matrix Γ is corrected in accordance with the cleaning quality. Specifically, the correction coefficient K Γ PUG is obtained by searching a map showing the characteristics of the fuel correction coefficient KPUG corresponding to the purge quality in fig. 21, and the obtained correction coefficient K Γ PUG is multiplied by the gain matrix Γ to correct the correction. As shown in the figure, the correction coefficient K Γ PUG decreases the fuel correction coefficient KPUG corresponding to the purge mass as the purge mass increases, while the disturbance increases, so that the gain matrix is made smaller as the fuel correction coefficient KPUG corresponding to the purge mass decreases. The fuel correction coefficient KPUG corresponding to the purge quality is also described in japanese patent application laid-open No. 6-101522, which was filed by the present applicant, and therefore, the description thereof is omitted.

Subsequently, the process proceeds to S320, where the gain matrix Γ is corrected based on the detected atmospheric pressure Pa. Specifically, a table showing the characteristics of the detected atmospheric pressure Pa is searched for in fig. 22 to obtain a correction coefficient K Γ Pa, and the obtained correction coefficient K Γ Pa is multiplied by a gain matrix Γ to correct the correction. The reason why the gain matrix Γ is corrected based on the detected atmospheric pressure Pa is that the filling efficiency decreases as the detected atmospheric pressure Pa decreases, that is, as the height of the machine position increases, and the data set at the atmospheric pressure is disturbed, and therefore, in order to improve the stability of the adaptive control system, the gain matrix is made smaller as the detected atmospheric pressure Pa decreases.

Subsequently, the process proceeds to S322, where the gain matrix Γ is corrected based on the detected water temperature TW. Specifically, a map showing the characteristics of the detected water temperature TW is retrieved from fig. 23, a correction coefficient K Γ TW is obtained, and the obtained correction coefficient K Γ TW is multiplied by a gain matrix Γ to correct the correction coefficient K Γ. The reason why the gain matrix Γ is corrected based on the detected water temperature TW is that the correction coefficient K Γ TW causes combustion instability when the detected water temperature TW is a low water temperature or a high water temperature, as shown in the drawing, and therefore the gain matrix Γ is reduced at the low water temperature or the high water temperature in order to improve the stability of the adaptive control system, in contrast to the disturbance of data set at the normal temperature.

As described above, in embodiment 2, the adaptive parameter is appropriately set and determined according to the operating state <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The gain matrix of the change (convergence) speed can obtain the stable change speed of the adaptive parameters and improve the controllability.

In addition, although the gain matrix Γ is determined using a fixed gain in embodiment 2, a variable gain algorithm may be used, and in this case, the initial values of the factors of the gain matrix Γ are corrected in accordance with the operating state as described above, and when the operating state changes, the initial values are set as predetermined values.

It should be noted that, in embodiment 2, the description is made by using the fixed gain algorithm, and when the gain matrix Γ (k) is calculated by using an algorithm other than the fixed gain algorithm, that is, the variable gain algorithm shown in equation 9, the calculation of the off-diagonal factor of the gain matrix Γ (k) is not performed, but is fixed to 0, which naturally achieves the reduction of the calculation amount and the simplification of the adjustment shown in embodiment 2.

FIG. 24 is a block diagram showing the procedure of embodiment 3 of the apparatus of the present invention.

In embodiments 1 and 2, the gain matrix Γ (k) is calculated with fixed gains, while embodiment 3 is calculated with algorithms other than fixed gains. In this case, when the control result (i.e., the device input, specifically, the detected air/fuel ratio KACT) using the adaptive parameters shows a good condition, if the calculated value is stored in advance according to the operating state of the machine, it is not necessary to recalculate the gain matrix Γ (k) in the region, and at the same time, the most appropriate gain matrix Γ (k) can always be used in the region, thereby improving controllability. Stored Γ (k) at this time may be a trimmed value such as an average value between 4 TDCs. In addition, when the gain matrix Γ (k) is calculated based on a fixed gain algorithm, the behavior of the device output is not good. The gain matrix Γ (k-1) at this time starts as an initial value stored in each operating region.

The above-described assumption will be explained with reference to fig. 24.

Fig. 24 shows a block diagram of the procedure of embodiment 3, and the operation is performed when searching for a map of gain matrix Γ in S308, S310, S304, or the like in fig. 18.

As shown in FIG. 24, in S400, a map of a gain matrix Γ is retrieved as in the case of the embodiment 2 based on the machine rotation speed Ne and the intake pressure Pb, the process proceeds to S402, it is judged by an appropriate method whether the condition of the detected air/fuel ratio KACT as the output of the apparatus is good, if the judgment is negative, the process proceeds to S404, the gain matrix Γ (k) is calculated, and the process proceeds to S406, and the gain matrix Γ (k) is stored in a predetermined region of the retrieved map. When the determination result at S402 is affirmative, the process proceeds directly to S406. The judgment in S402 as to whether the detected air/fuel ratio KACT is in good condition or not is as follows: for example, when the detected air/fuel ratio KACT between 10 TDCs enters the desired air/fuel ratio KCMD ± a predetermined value range, it is determined to be good.

Since embodiment 3 is configured as described above, when the detected air/fuel ratio KACT is good, the gain matrix Γ (k) can be calculated by looking up the map without using the formula shown in equation 9, and therefore, the amount of calculation can be reduced. Further, when the detected air/fuel ratio KACT is not good, the optimum gain matrix Γ (k) is recalculated and learning is performed for each operating region of the internal combustion engine, so that it is possible to cope with the aging of the internal combustion engine, and the condition of detecting the equivalence ratio KACT (k) can be made always good, thereby making it possible to improve the controllability.

FIG. 25 is a block diagram showing the procedure of the 4 th embodiment of the apparatus of the present invention.

In embodiment 4, a dead zone is provided in detecting the characteristics of the air/fuel ratio KACT in order not to destabilize the adaptive control system. That is, the STR controller operates so that the detected air/fuel ratio KACT coincides with the desired air/fuel ratio KCMD, and the adaptation parameter hardly changes when the detected air/fuel ratio KACT input to the STR controller coincides with the desired air/fuel ratio KCMD. Therefore, if the detected air/fuel ratio KACT slightly changes due to a small disturbance such as sensor noise, in order to prevent the adaptive control system from being affected by the small disturbance and performing unnecessary over-correction, as shown in fig. 26, a dead zone is provided in the vicinity of the desired air/fuel ratio KCMD in the characteristics of the detected air/fuel ratio KACT. Specifically, the detected air/fuel ratio KACT is made to be the same in the range from KCMD-beta to KCMD + alpha.

As shown in the block diagram of the routine of FIG. 25, at S500, the detected air/fuel ratio KACT is compared with a lower limit predetermined value KCMD-beta, and if it is greater than or equal to the lower limit predetermined value KCMD-beta, the flow proceeds to S502, where the detected air/fuel ratio KACT is compared with an upper limit predetermined value KCMD + alpha. If the result of the comparison at S502 is that the detected air/fuel ratio is less than or equal to the predetermined value KCMD + α, the process proceeds to S504, where the detected air/fuel ratio KACT is taken as the predetermined value, for example, the desired air/fuel ratio KCMD. When it is determined at S500 that the detected air/fuel ratio KACT is less than the lower limit predetermined value KCMD- β, and when it is determined at S502 that the detected air/fuel ratio KACT is greater than the upper limit predetermined value KCMD + α, the routine is immediately ended. Therefore, in this case, the detection value is directly taken as the detection air/fuel ratio KACT. Through the above processing, as shown in fig. 26, a dead zone can be provided in the vicinity of the desired air/fuel ratio KCMD in detecting the characteristic of the air/fuel ratio KAG.

Since the 4 th embodiment is configured as described above, when a minute change in the detected air/fuel ratio KACT occurs, the STR controller is not affected by the change, and can be operated stably, so that a good control result can be obtained. In addition, although the desired air/fuel ratio KCMD is taken as the detected air/fuel ratio in S502, it may be taken as a value in the range of KCMD- β to KCMD + α.

FIG. 27 is a block diagram showing the procedure of the embodiment 5 of the apparatus of the present invention. The reason for the embodiment 5 is to prevent the instability of the adaptive control system, as in the embodiment 4. For identification error signal e^*And setting an upper limit and a lower limit so as to obtain stable self-adaptive parameters.

That is, as can be seen from equation 8, by identifying the error signal e^*The value of (A) is limited within a certain range, and the adaptive parameter can be limited <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> The rate of change of (c). In this way, it is prevented from exceeding the adaptive parameter <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Optimum value of. As a result, the adaptive control system can be stably operated, and a favorable control result can be obtained.

As shown in the block diagram of FIG. 27, the calculated identification error signal e is first processed in S600^*(k) If it is determined that the upper limit value a is exceeded as compared with the upper limit value a (shown in fig. 28), the process proceeds to S602, and a predetermined value, for example, the upper limit value a is set as the identification error signal e^*(k) In that respect When it is judged at S600 that the signal e is discriminated from the attenuation signal e^*(k) If the value is less than or equal to the upper limit value a, the process proceeds to S604, and the calculated identification error signal e is transmitted^*(k) If the value is determined to be less than the lower limit b (shown in fig. 28), the process proceeds to S606, where the 2 nd predetermined value, for example, the lower limit b is set as the identification error signal e^*(k) In that respect When it is judged at S604 that the error signal e is recognized^*(k) And if the value is greater than or equal to the lower limit value b, immediately ending the process. Therefore, in this case, the error signal e is recognized^*(k) That is, the calculated value.

Since the 5 th embodiment is constructed as described above, the identification error signal e is generated^*(k) The value of (A) is limited within a certain range, and the adaptive parameter can be limited <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Is changed. In this way, the adaptive parameters can be prevented from being exceeded <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The optimum value of (b) enables the adaptive control system to operate stably, and a good control result is obtained.

In addition, in S602 to S606, although the error signal e is recognized^*(k) The value of (b) is defined as the upper and lower limit values, but it may be defined as an appropriate value between the upper and lower limit values or as an appropriate value in the vicinity of the upper and lower limit values.

FIG. 29 is a flowchart showing a 6 th embodiment of the apparatus of the present invention.

Embodiment 6 is an STR controller as described in embodiment 1, wherein the identification error signal e for determining the adaptive parameter θ is used^*Constant used in denominator of formula 101 is variable, so that the speed change is stable and the control performance is improved.

Embodiment 6 is premised on a technique of limiting the range of variation of intermediate variables used for calculation by the parameter adjusting mechanism and realizing adaptive control as shown in the figure by using a general-performance vehicle-mounted computer. In this respect, Japanese patent application laid-open No. 6-161511, which was proposed by the present applicant, is described, and the description thereof is omitted here.

I.e. in the theoretical formula, the identification error signal e^*(k) As in equation 10. Now, ζ (k) and y (k) are multiplied by 1/10 (hereinafter, j) and inputted to the parameter adjusting means, and the denominator thereof is shown by equation 28 (when the gain matrix Γ (k-1) is a fixed gain, it is fixed).

In the formula, the right term is the square of a multiplication coefficient of ζ (k) and y (k), and when the coefficient is a small value of 1 or less (e.g., 1/10)²= 1/100), the left term =1 is very small. Thus, the right term, however varying, identifies the error signal e^*(k) Also has a value of approximately 1, identifies the error signal e than before multiplication by the coefficient^*(k) The rate of change of (c) is changed. To solve this problem, the left term may be set to a value other than 1. For clarity, assuming the above coefficient is i, the left term is j²The same rate of change can be obtained as before multiplication by the coefficient i. Otherwise, due to the recognition error signal e^*(k) Speed of change and adaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) Is proportional to the rate of change (convergence), i.e., <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) is calculated by equation 8 by making the left term j²Other than the value, the adaptive parameter may be changed <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The rate of change of (c). Therefore, the recognition error signal e shown in equation 29^*(k) In the denominator calculation formula (b), I in the formula is set to a value other than 1, that is, a value I ≠ 1.

As shown in the block diagram of FIG. 29, in S700, it is determined that the error signal e is recognized^*(k) Determined adaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) If the determination is affirmative, the routine proceeds to S702, where i is set to a value other than 1, specifically, i is obtained by searching a map showing the characteristics of the detected machine rotation speed Ne and the detected intake pressure Pb in fig. 30. If the determination at S700 is negative, the process proceeds to S704, where i is defined as j²The same rate of change is obtained as before multiplication by the coefficient j. In the graph characteristics shown in FIG. 30, j is a constant, and the value of i takes into account j²For example, set to i = j²X 0.5 to i = j²X 2, etc.

Specifically, j is usually set to a value smaller than l, and for example, if j = 1/10 is set, if the determination at S700 is negative, i = j is set²1/100. Therefore, even if the determination at S700 is affirmative, i-map values are set in fig. 30, centering on i = 1/100, for example, between 1/50 and 1/200. At this point, the smaller i (e.g., 1/200), the adaptive parameter <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The greater the rate of change (convergence) of (i), the greater i (e.g., 1/50), the adaptive parameter <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> (k) The smaller the change (convergence) speed of (c). Therefore, the i map value in child 30 is set to be larger (e.g., 1/50) in the high-speed rotation high-load state and smaller (e.g., 1/200) in the low-speed rotation low-load state.

Since the embodiment 6 is constructed as described above,by making the identification error signal e determining the adaptive parameter theta^*The constant of (2) is variable, and the coefficient to the input can be adjusted, so that the changing speed of the adaptive parameter theta is stable, and good controllability is obtained.

In addition, in embodiment 6, the STR controller used in embodiment 1 is described as an example, but the adaptive controller is not limited to the form of embodiment 1, and may be any adaptive controller of the MRACS type as long as it operates in accordance with the Landau rule.

FIG. 31 is a block diagram showing the procedure of embodiment 7 of the apparatus of the present invention.

In embodiment 7, the control cycle of the parameter adjusting means and the STR controller shown in embodiment 1 is made variable, and the control cycle is determined according to the operating state, specifically, the machine rotation speed. That is, since the control cycle of the adaptive controller parameter adjustment means or the controller is made variable according to the operating state, the calculation load can be reduced to the maximum, and adaptive control can be performed even in an operating state with a small calculation time such as at the time of high-speed rotation, thereby achieving excellent control performance.

As shown in fig. 31, first, in S800, the detected machine rotation speed Ne is compared with a predetermined value NeP1, and when it is determined that the detected machine rotation speed Ne is smaller than a predetermined value NeP1, the process proceeds to S802, and the detected machine rotation speed Ne is compared with another predetermined value NeC 1. When it is determined at S802 that the detected machine rotation speed Ne is less than the other predetermined value NeC1, the routine proceeds to S804, and the control cycle of the parameter adjusting mechanism (shown as P in fig. 31) and the STR control (shown as C in fig. 31) is set to 1 TDC.

Fig. 32 is a diagram illustrating the operation of the block diagram of fig. 31, as shown at predetermined value NeP 1. In the case of the NC1, since there is a margin in calculation time in a relatively low rotation region, the control accuracy is high, and as shown in fig. 8 and 9, the parameter adjusting mechanism and the STR controller are operated for each TDC,

in fig. 31, when it is determined at S802 that the detected machine rotation speed Ne exceeds the predetermined value NeC1, the process proceeds to S806, the detected machine rotation speed Ne is compared with another predetermined value NeC2, and if it is determined to be smaller than the predetermined value, the process proceeds to S808, the parameter adjusting means is operated once per TDC, and the STR controller is operated once per 2 TDCs. When it is determined at S806 that the detected machine rotation number Ne is equal to or greater than the predetermined value NeC2, the routine proceeds to S810, where the parameter adjusting means is operated once per TDC and the STR controller is operated once per 4 TDCs.

When it is determined at S800 that the detected machine rotation speed Ne is equal to or greater than the predetermined value NeP1, the process proceeds to S812, the detected machine rotation speed Ne is compared with a predetermined value NeP2, if it is determined to be less than the predetermined value, the process proceeds to S814, the detected machine rotation speed Ne is compared with a predetermined value NeC3, and if it is determined to be less than the predetermined value NeC3, the process proceeds to S816, the parameter adjusting means is operated once every 2 TDCs, and the STR controller is operated every TDC.

When it is determined in S814 that the detected machine rotation speed Ne is equal to or greater than the predetermined value NeC3 pair, the routine proceeds to S818, and the detected machine rotation speed Ne is compared with the predetermined value NeC4, and if it is determined to be smaller than the predetermined value, the routine proceeds to S820, and both the parameter adjusting mechanism and the STR controller are operated once every 2 TDCs. When it is determined at S818 that the detected machine rotation speed Ne is equal to or greater than the predetermined value NeC4, the routine proceeds to S822, where the parameter adjustment mechanism is operated once every 2 TDCs, and the STR controller is operated once every 4 TDCs.

When it is determined at S812 that the detected machine rotation speed Ne is equal to or greater than the predetermined value NeP2, the process proceeds to 824, the detected machine rotation speed Ne is compared with a predetermined value NeP3, if it is determined to be smaller than the predetermined value, the process proceeds to S826, the detected machine rotation speed Ne is compared with a predetermined value NeC5, and if it is determined to be smaller than the predetermined value, the process proceeds to S828, the parameter adjusting means is operated once every 4 TDCs, and the STR controller is operated once every TDC (shown in fig. 16).

When it is determined in S826 that the detected machine rotation speed Ne is greater than or equal to NeC5, the routine proceeds to S830, the detected machine rotation speed Ne is compared with a predetermined value NeC6, and if it is determined to be smaller than the predetermined value, the routine proceeds to S820, the parameter adjusting means is operated once every 4 TDCs, and the STR controller is operated once every 2 TDCs. When it is determined at S830 that the detected machine rotation speed Ne is equal to or greater than the predetermined value NeC6, the routine proceeds to S834 where the parameter adjusting mechanism and the STR controller are both operated once every 4 TDCs (shown in fig. 17). When it is determined at S824 that the detected machine rotation speed Ne is equal to or greater than the predetermined value NeP3, the routine proceeds to S836 to stop the adaptive controller STR.

In embodiment 7, since the control cycle of the parameter adjusting means of the adaptive controller and the STR controller is determined based on the machine rotation speed as described above, the calculation load can be reduced as much as possible, and adaptive control can be performed even in an operating state with a small calculation time such as at the time of high-speed rotation, thereby achieving excellent control performance.

The operation states of the adaptive controller STR shown in fig. 32 do not necessarily have to be all states 1 to 10 (indicated by circled numbers in the figure), and may be appropriately selected according to the capabilities of the CPU of the device and the control unit. For example, 1, 3, 5, 9, 10 or 1, 3, 6, 9, 10 or 1, 7, 9, 10 or 1, 4, 7, 10, etc. may be selected.

In the present embodiment, the operation state is the machine rotation speed used, but the present invention is not limited to this, and factors of the upper machine load may be considered in addition to the rotation speed when determining the control cycle. In this case, for example in the case of high load, due to the adaptive parameters <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Is small, so the process parameter adjustment mechanism is considered every 4 TDCs.

Fig. 33 is a subroutine block diagram showing an operation of calculating the average value of the feedback correction coefficient KSTR and the like in the same manner as in fig. ll in embodiment 8 of the apparatus of the present invention.

In embodiment 1, in order to avoid the influence of the exhaust air/fuel ratio of the specific cylinder, the factors that determine the feedback correction coefficient KSTR are basically averaged, and the calculation of the average value is stopped in a predetermined operating state, that is, the idling state.

In contrast to embodiment 1, embodiment 8 does not calculate the average value in principle, and calculates the average value only in a predetermined operating state, specifically, when Exhaust Gas Recirculation (EGR) is performed.

In the above exhaust gas recirculation mechanism, when the exhaust gas is recirculated, the exhaust gas may be unevenly introduced into the 4 cylinders depending on the operation state, and for example, a cylinder close to the recirculation port 121b may suck a large amount of exhaust gas, while a cylinder far away may suck only a small amount of exhaust gas. This is likely to happen. Therefore, in this case, the air/fuel ratio kact (k) detected at each TDC is greatly affected by the particular cylinder, and if the air/fuel ratio kact (k) is employed so that only the equivalence ratio of the cylinder coincides with the desired air/fuel ratio, the control values for all cylinders deviate depending on the amount of deviation of the cylinder, thereby shifting the air/fuel ratios of the other cylinders. Therefore, to avoid this problem, it is preferable to average as shown in the figure.

As shown in fig. 33, it is judged at S900 whether or not EGR (exhaust gas recirculation control) is being executed, and if the judgment is affirmative, the routine proceeds to S902, and an average value of KACTAVE and the like is obtained as in the 1 st embodiment shown in fig. ll. If the result at S900 is negative, the routine proceeds to S9l2, and the same processing as in the l-th embodiment shown in fig. ll is performed.

The structure of embodiment 8 as described above does not greatly affect the controllability of the exhaust gas recirculation by being affected by the specific cylinder.

Fig. 34 is a subroutine block diagram showing an operation of calculating an average value of the feedback correction coefficient KSTR and the like similar to fig. 33 according to embodiment 9 of the apparatus of the present invention.

When the cylinder type cleaner is operated and gas is supplied as in the case of the exhaust gas recirculation, the same problem as in embodiment 8 occurs, that is, the gas is unevenly introduced into the cylinder depending on the operation state. The 9 th embodiment is to solve this problem.

As shown in fig. 34, it is judged at S1000 whether or not the cylinder filter is in operation, and if the judgment is affirmative, the routine proceeds to S1002, where an average value of KACTAVE and the like is obtained as in the 1 st embodiment shown in fig. 11. When it is determined as negative in S1000, the process proceeds to S1012, and the same processing as in embodiment 1 shown in fig. 9 is performed.

Since the 9 th embodiment is constructed as described above, the cylinder cleaner is not seriously affected by the specific cylinder even when it is operated, thereby improving the controllability.

Although not shown in the figure, it is desirable to perform the averaging in the same manner even in a combustion unstable state in other cases, for example, when the atmospheric pressure Pa is low, that is, when the machine is located at a high position, or when the water temperature is low, or when the combustion is in a poor operation, and the like, whereby the controllability can be improved.

Fig. 35 and 36 are a block diagram and a signal flow diagram illustrating an embodiment of the apparatus 10 of the present invention.

As shown in fig. 36, in the 10 th embodiment, the feedback loop (correction coefficient KLAF) of the equivalent ratio at the merging portion of the exhaust system according to the PID control principle in the configuration of the 1 st embodiment is removed, and the feedback loop (correction coefficient # nKLAF) of the individual cylinder according to the same PID control principle is inserted.

That is, the air/fuel ratio # nA/F (n: cylinder) of each cylinder is estimated from the output of a single air/fuel ratio sensor disposed at the exhaust merging portion by the monitor proposed by the applicant of the present invention in Japanese patent application laid-open No. 5-180040, and the feedback correction coefficient # nKLAF of each cylinder is obtained by the PID control principle from the deviation of the estimated value from a predetermined value of the air/fuel ratio F/B of each cylinder, and is multiplied by the output fuel injection amount Tout for correction.

More specifically, in order to eliminate the deviation between the expected value of the air/fuel ratio F/B for each cylinder (the value obtained by dividing the merge air/fuel ratio by the last calculated value of the average value of the feedback correction coefficients # nKLAF for each cylinder) and the air/fuel ratio # nA/F estimated by the monitor, the feedback correction coefficient # nKLAF for each cylinder is obtained by using the PID control law. The details thereof are described in Japanese patent application laid-open No. 5-251138, which is filed by the present applicant, and the description thereof will be omitted here. In addition, the drawing attached with the correction compensator is also omitted.

In the 10 th embodiment, a sample block (shown as Sel-VOBSV) is provided for synchronous sampling of the output of the LAF sensor as appropriate, and a similar sample block (shown as Sel-VSTR) is also provided in the STR controller.

These sampling blocks and monitors are explained below, and the sampling operation block is shown as Sel-VOBSV in fig. 36.

In the internal combustion engine, exhaust gas is discharged in the exhaust stroke, and therefore, the air/fuel ratio is observed at the exhaust merging portion of the multi-cylinder internal combustion engine, apparently in synchronization with TDC. Therefore, the above-described wide band sensor is provided in the exhaust system of the internal combustion engine, and when the air/fuel ratio is sampled, it is also necessary to synchronize with the TDC, and thus, the change in the air/fuel ratio may not be accurately specified due to the sampling synchronization of the control unit (ECU) that processes the detection output. That is, for example, when the air/fuel ratio at the exhaust merging portion is as shown in fig. 37 with respect to TDC, the air/fuel ratio recognized by the control unit becomes a completely different value due to sampling synchronization as shown in fig. 38. In this case, it is preferable to sample the air/fuel ratio sensor at a position where the change in the output of the actual air/fuel ratio sensor can be accurately grasped as much as possible.

In addition, the air/fuel ratio varies depending on the time at which the exhaust gas reaches the sensor and the sensor response time. The time to reach the sensor varies depending on the pressure of the exhaust gas, the volume of the exhaust gas, and the like. Further, since sampling in synchronization with TDC is performed based on the crank angle, it is inevitably affected by the rotational speed of the machine. Thus, the detection of the air-fuel ratio has a large relationship with the operating state of the machine. Therefore, in the technique described in japanese patent application laid-open No. 1-313644, although it is determined whether detection is appropriate for each predetermined crank angle, the structure is complicated, the calculation time is long, there is a possibility that the detection cannot be dealt with in a high-speed rotation region, and the change point of the output of the air/fuel ratio sensor is passed through at a timing of determining the detection.

Fig. 39 is a block diagram showing a sampling operation of the LAF sensor. Since the detection accuracy of the air/fuel ratio is closely related to the estimation accuracy of the monitor, the air/fuel ratio estimation of the monitor will be briefly described before the description of fig. 39. First, in order to separate and extract the air/fuel ratio of each cylinder from the output of 1 LAF sensor with high accuracy, it is necessary to accurately figure out the detection reaction delay of the LAF sensor. Therefore, the delay is first approximately modeled as a 1-time delay system, and the model shown in fig. 40 is created. Wherein, if LAF and LAF sensor output and A/F and input A/F are set, the state equation is shown in the formula 30.

When the period Δ T is discretized, the equation becomes 31. Fig. 41 is a signal flow chart showing expression 31.

Wherein,

therefore, by using equation 31, the true air/fuel ratio can be calculated from the sensor output. That is, since equation 31 can be rewritten to equation 32, the value at time k-1 can be inversely calculated from the value at time k as shown in equation 33.

Specifically, if the Z conversion is used to express equation 31 as a transfer function, the inverse transfer function is multiplied by the current LAF sensor output LAF to estimate the previous input air/fuel ratio in real time as shown in equation 34. Fig. 42 is a signal flow diagram showing the real-time a/F estimator.

Next, a method of separating and extracting the air/fuel ratio in each cylinder from the true air/fuel ratio obtained above will be described. As described in the above-mentioned application, the air/fuel ratio at the exhaust system merging portion is considered as a weighted average value, and the value at time K is expressed by equation 35, taking into account the degree of influence of the air/fuel ratio of each cylinder with time. Note that, although the "fuel/air ratio F/a" is used herein because F (fuel amount) is used as the control amount, in the following description, the "air/fuel ratio" is used for the convenience of understanding as long as it is not confused. The air/fuel ratio (or fuel/air ratio) is the true value obtained by correcting the reaction delay obtained by equation 34.

That is, the air/fuel ratio at the merging portion is the sum of products of the past combustion lag for each cylinder and the weighting coefficient Cn (for example, 40% for the most recently combusted cylinder, 30% for the cylinder combusted before that, and so on). When the model is represented by a signal flow diagram, the model is in the form of fig. 43.

The equation of state is shown in equation 36.When the air/fuel ratio at the merging portion is defined as y (k), the output equation is expressed by equation 37.

In the formula, C: 0.05, C₂∶0．15，C₃∶0．30，C₄∶0．50

In the above, since u (k) cannot be observed, it is found from the state equation that x (k) cannot be observed even if an observer is provided. Therefore, assuming that the air/fuel ratio before 4TDC (i.e., the same cylinder) is in a normal operating state of sharp change, let x (K +1) = x (K-3), as shown by equation 38.

Now, simulation results of the model obtained as described above are disclosed. FIG. 44 shows a 4-cylinder internal combustion engine fueled with 3 cylinders at a 14: 7: 1 air/fuel ratio and only one cylinder at a 12.0: 1 air/fuel ratio. Fig. 45 shows a case where the air/fuel ratio at the merging portion at this time is obtained by the above model. Although the stepped output can be obtained in this figure, when the response delay of the LAF sensor is further taken into consideration, the sensing output has a gentle waveform as shown by the "model output value" in fig. 46. The "actual measurement value" in the figure is the actual measurement value of the LAF sensor output in the same case, and the comparison of the two shows that the above model can favorably model the exhaust system of the multi-cylinder internal combustion engine.

Therefore, it is concluded that the problem of the conventional kalman filter of the equation of state and the output equation observation x (k) shown in equation 39 is solved. The load matrix Q, R is shown in equation 40, and when the Riccati equation is solved, the gain matrix k is shown in equation 41.

In the formula,

when A-Kc is obtained from the above, it is expressed by the formula 42

A general monitor structure is shown in fig. 47, but since no value u (k) is input to the current model, only y (k) is input as shown in fig. 48, and this is expressed by the formula shown in formula 43Shown in the figure.

In the formula, a system moment matrix of a kalman filter, which is a monitor having y (k) as an input, is shown in formula 44.

In the present model, the load distribution R of the Riccati equation is given by: when the factor of Q = 1: 1, the system matrix S of the Kalman filtering program is shown as formula 45,

fig. 49 shows a case where the above model is combined with a monitor. The simulation results are shown in the prior application, and the air/fuel ratio of each cylinder is reliably extracted from the merged air/fuel ratio in a manner not shown here.

Since the air/fuel ratio of each cylinder can be estimated from the air/fuel ratio at the merging portion by the monitor, the air/fuel ratio can be controlled individually for each cylinder by a control law such as PID. Specifically, as shown in fig. 50 (feedback section constituted by the monitor of fig. 36), the joining portion feedback correction coefficient KLAF is determined from the sensor output (joining portion air/fuel ratio) and the desired air/fuel ratio by the PID control law, and the feedback correction coefficient # nKLAF (n: cylinder) for each cylinder is determined from the monitor estimated value # nA/F.

More specifically, the feedback correction coefficient # nKLAF for each cylinder is found by the PID rule so as to eliminate the deviation of the expected value from the monitor estimated value # nA/F. The expected value is obtained by dividing the air/fuel ratio at the confluence by the previous calculation of the average value for all cylinders using the feedback correction coefficient # nKLAF for each cylinder.

As a result, the air/fuel ratio of all the cylinders converges to the desired air/fuel ratio. Here, the fuel injection amount # nTout (defined by the valve opening time of the injector) for each cylinder is expressed by the formula

#nTout=TCYl×#nKLAF×KLAF

And (4) obtaining.

Next, referring back to the block diagram of fig. 39, the sampling of the output of the LAF sensor will be described. In addition, the routine is started at the TDC position.

As shown in FIG. 39, the machine rotation speed Ne, the intake air pressure Pb, and the valve speed V/T are read in S1200, the process proceeds to S1204 and S1206, a speed map for HiV/T or LoV/T (described later) is searched, the process proceeds to S1208, and the sensor output calculated by the monitor for adjusting the speed of the Hi or Lo valve is sampled. Specifically, a pulse waveform diagram is searched based on the machine rotation speed Ne and the intake pressure Pb, any one of the buffers is selected by the number of the buffer, and the sample value stored therein is selected.

Fig. 51 is an explanatory diagram showing characteristics of a pulse waveform diagram. The illustrated characteristics are set as follows: the lower the machine rotation speed Ne or the higher the intake pressure (load) Pb, the value sampled at the earlier crank angle is selected. "earlier" as used herein refers to a value sampled at a position closer to the previous TDC position (in other words, an old value). Conversely, as the machine rotation speed Ne is higher or the intake air pressure Pb is lower, the value sampled at the later crank angle, that is, the value sampled at an angle near the latter TDC position (in other words, a new value) is selected.

That is, as shown in fig. 38, it is preferable to sample the output of the LAF sensor at an inflection point as close to the actual air/fuel ratio as possible, but if the response time of the sensor is constant, the inflection point, for example, the initial peak, is generated at an early crank angle as the machine rotation speed is lower as shown in fig. 52. Further, the higher the load, the more the exhaust gas pressure and the exhaust gas volume increase, and therefore the flow velocity of the exhaust gas increases and the time to reach the sensor becomes earlier. In this sense, the sampling timing is set as shown in fig. 51.

The valve timing is characterized in that Ne1-Lo is set for Lo side and Ne-Hi side as arbitrary values Ne1 of the machine rotation speed, Pb1-Lo is set for Lo side and Pb1-Hi is set for Hi side as arbitrary values of the intake pressure

Pb1-Lo＞Pb1-Hi

Ne1-Lo＞Ne1-Hi

That is, the characteristics of the graph are set as follows: at HiV/T, the exhaust valve opening timing is earlier than LoV/T, so when the machine speed or intake pressure values are the same, the earlier sampled value is selected.

Then, the process proceeds to S1210, where HiV/T is used to calculate the monitor matrix, and then to S1212, where LoV/T is used to calculate the same. And S1214, judging the valve speed regulation again, entering S1216 and S1218 according to the judgment result, selecting the calculation result and finishing.

That is, since the state of the air/fuel ratio merging portion changes with the switching of the valve timing, the monitor matrix needs to be changed. However, the estimation of the air/fuel ratio of each cylinder cannot be performed instantaneously, and the estimation calculation of the air/fuel ratio of each cylinder is performed several times before convergence ends, so that the calculation by the monitor matrix before the valve speed change and the calculation by the monitor matrix after the change are superimposed, and even if the change of the valve speed is performed, the selection can be performed in S1214 based on the valve speed after the change. After each cylinder is estimated, as described above, the feedback correction coefficient is obtained to determine the injection amount in order to eliminate the deviation from the expected value.

According to this structure, the detection accuracy of the air/fuel ratio can be improved. That is, as shown in fig. 53, sampling at short intervals allows the sampled value to substantially reflect the output of the sensor, and the values sampled at short intervals are sequentially stored in the buffer, and the inflection point of the output of the sensor is predicted from the machine rotation speed and the intake pressure (load), and a value corresponding to the inflection point is selected from the buffer at a predetermined crank angle. Then, the monitor is calculated to estimate the air/fuel ratio of each cylinder, and as shown in fig. 50, feedback control of the air/fuel ratio can be performed for each cylinder.

Therefore, as shown in the lower part of fig. 53, the CPU chip can correctly recognize the maximum value and the minimum value of the sensor output. Therefore, according to this configuration, when the air/fuel ratio of each cylinder is estimated by the monitor, the estimation accuracy of the monitor can be improved by using a value similar to the actual air/fuel ratio condition, and as a result, the accuracy when the air/fuel ratio feedback control is performed for each cylinder as shown in fig. 50 is also improved. The details thereof are described in Japanese patent application laid-open No. Hei 6-243277, which was filed by the present applicant, and the detailed description thereof is omitted here.

While the monitor samples the output of the LAF sensor (indicated as Sel-OBSV in fig. 36), the STR controller also samples the same (indicated as Sel-VSTR in fig. 36).

That is, the Sel-VSTR is also obtained in the same procedure as that performed in the Sel-VOBSV, that is, in the same procedure as that shown in the block diagram in fig. 39. The Sel-VOBSV detects the air/fuel ratio by using the same map as that shown in fig. 51 of the Sel-VOBSV in accordance with the detected air/fuel ratio at the optimum timing (for example, the optimum timing of the weighting coefficient C of the monitor with respect to the model) for estimating the air/fuel ratio of the cylinder by the monitor. So that Sel-VSTR thereof becomes the optimum timing (for example, the timing of detection of the air/fuel ratio most affected by the cylinder of the latest exhaust stroke) when the STR is operated.

For the above reasons, the 10 th embodiment will be described with reference to the flowchart of fig. 35. The process proceeds to S1112 through step S1100 or S1110 as in embodiment 1, and the output of the LAF sensor is sampled by Sel-VSTR, that is, the air/fuel ratio kact (k) is detected. Then, the process proceeds to S1114, where feedback correction coefficient KSTR is calculated in the same manner as in embodiment 1. Specifically, the calculation is performed using the block diagram of the routine of fig. 11 used in embodiment 1.

Subsequently, the routine proceeds to S1116 and S1118, the requested fuel injection amount tcyl (k) and the output fuel injection amount tout (k) are determined, and the routine proceeds to S1120, where the output of the LAF sensor is sampled by the Sel-VOBSV, that is, the equivalence ratio kact (k) is detected. Subsequently, the routine proceeds to S1122, the air/fuel ratio # nA/F of each cylinder is estimated by the monitor, the routine proceeds to S1124, the feedback correction coefficient # nKLAF of each cylinder is calculated, the routine proceeds to S1126, the learned value # nKLAFsty is calculated from a weighted average value with the previous value, the routine proceeds to S1128, the output fuel injection amount TOUT is multiplied by the feedback correction coefficient # nKLAF of each cylinder to obtain the output injection amount # ntot of the cylinder, the routine proceeds to S1130, the intake pipe wall adhesion correction is performed, and the routine proceeds to S1132 for output.

If the results of S1108 to S1110 are negative, the routine proceeds to S1134, where the required fuel injection amount tcyl (k) is determined as shown, proceeds to S1136, where the learned value of the feedback correction coefficient # nKLAFsty for each cylinder is read, and proceeds to S1138, where the learned value is used as the correction coefficient # nKLAF. When it is determined at S1104 that the fuel is cut, the routine proceeds to S1146 via S1144, the matrix calculation is stopped, and the routine proceeds to S1148, where the feedback correction coefficient for each cylinder is taken as the previous value. The remaining steps are the same as in example 1.

Since the 10 th embodiment is configured as described above, as in the 1 st embodiment, the input to the parameter adjusting means is synchronized with the combustion cycle while calculating the adaptive parameters, so that the calculation load of the parameter adjusting means is greatly reduced, the control performance is ensured, the adaptive controller can be used in the actual machine, and the variation between cylinders can be reduced.

Further, as in embodiment 1, the average value of the air/fuel ratio KACT and the average value of the adaptive parameter are obtained for all the cylinders during 1 combustion cycle and inputted into the parameter adjusting mechanism, and the output average value of the STR controller is also obtained, so that the influence of the combustion state of the specific cylinder is not so great.

In the 10 th embodiment, as in the 2 nd embodiment, the adaptive parameter θ or KSTR may be averaged, or the air/fuel ratio KACT and the adaptive parameter θ may be averaged together. In addition, the desired air/fuel ratio kcmd (k) may also be the same value in all the cylinders.

In addition, the descriptions in embodiments 2, 3, 4, 5, 6, 7, 8, and 9 are applicable to embodiment 10.

Fig. 54 and 55 are a block diagram and a signal flow diagram showing an embodiment 11 of the apparatus of the present invention.

Embodiment 11 as shown in fig. 55, an STR controller and a parameter adjusting mechanism are inserted in series in a fuel injection quantity calculating system. That is, in the same manner as in embodiment 1, the desired air/fuel ratio correction coefficient kcmdm (k) and the various correction coefficients KTOTAL are multiplied by the basic fuel injection amount Tim to obtain a required fuel injection amount tcyl (k), and then the corrected required fuel injection amount tcyl (k) is input to the STR controller.

On the other hand, the average value KACTAVE and KACTAVE are obtained from the detected air/fuel ratio at the merging portion of the air-fuel system, as in embodiment 1 <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> AVE dynamically corrects the requested fuel injection quantity Tcyl (k) in the STR controller to calculate a corrected fuel injection quantity Gfuel-STR (k).

At the same time, a feedback correction system KLAF of the junction is obtained from the detected air/fuel ratio at the junction of the exhaust system by PID control, and the obtained product is multiplied by a required fuel injection amount Tcyl (k) to calculate a corrected fuel injection amount Gfuel-KLAF (k).

In fig. 55, the STR controller calculates the output fuel quantity Gfuel-STR (k) so that the actual sucked fuel quantity (more precisely, the estimated sucked fuel quantity) Gfuel (k) is adaptively supplied as the output fuel injection quantity tout (k) to the internal combustion engine in accordance with the desired fuel quantity tcyl (k). Further, the wall surface attachment correction of the virtual machine has been described in detail in Japanese patent application No. 4-200331 (Japanese patent application laid-open No. 6-17681) proposed by the present applicant, and is not essential to the present invention, and therefore, the description thereof will be omitted.

The actual intake fuel quantity gfuel (k) described here may be obtained by dividing the detected air quantity by the detected air/fuel ratio, but in the embodiment, since an air quantity detector (airflow meter) is not provided, the detected air/fuel ratio is multiplied by the desired intake fuel quantity (required injection quantity) tcyl (k). The actual intake fuel amount calculated by this method is equal to the value calculated by the detected air amount. In addition, as described previously, in this control, the desired air/fuel ratio and the detected air/fuel ratio are actually expressed as the equivalence ratio.

When the desired air/fuel ratio is not the theoretical air/fuel ratio, the calculated value is further divided by the desired air/fuel ratio to find the actual amount of the intake fuel. That is, when the desired air/fuel ratio is the theoretical air/fuel ratio, the actual intake fuel amount is expressed by the equation:

the actual intake fuel quantity = required injection quantity (desired intake fuel quantity) × detection air/fuel ratio (equivalence ratio) is found; when the desired air/fuel ratio is not the theoretical air/fuel ratio, the equation for indicating the amount of fuel drawn

Actual intake fuel amount = (required injection amount (expected intake fuel amount) × detected air/fuel ratio (equivalence ratio))/expected air/fuel ratio (equivalence ratio)

And (4) obtaining.

The above is explained with reference to fig. 54. The routine proceeds to S1318 through steps S1300 to S1316 similar to those of the above-described embodiment, and calculates the average value KACTAVE of the air/fuel ratio and the adaptive parameter <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Average value of (2) <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> -AVE。

Subsequently, the process proceeds to S1324 through S1320 to S1322, and it is determined whether the adaptive control system (STR controller) is unstable in the same manner as in embodiment 1.

Fig. 56 is a subroutine block diagram showing this job.

As shown in fig. 56, the stability of the STR control system is first determined using the various factors of the adaptive parameter θ in S1400.

Specifically, the fuel injection quantity Gfuel-STR (k) calculated by the STR controller is calculated by equation 46.

Assuming that the correction is correct, the transfer function of the hypothetical device is shown as equation 47. Gfuel (z)^-1)=z^-3Gfuel-str(z^-1) A.47

From equations 46 and 47, the transfer function from tcyl (k) to injection quantity Gfuel-str (k) is shown in equation 48.

In the formula, b0 is a vector for determining the gain, and cannot be 0 or a negative number, so the denominator function f (z) = BOZ of the transfer function of formula 48³+r1Z²+ R2Z + R3+ S0 is one of the functions shown in fig. 14. Therefore, it is determined whether or not the circle is actually within the unit circle. That is, as shown in FIG. 15, it is determined whether f (-1) < 0 or f (1) > 0, and if the determination is affirmative, it indicates that the real root is within the unit circle, whereby it can be easily determined whether the system is stable.

Then, in S1402, it is determined whether the STR control system is unstable according to the above description, and if so, the process proceeds to S1404, where adaptive parameters are adjusted <math> <mover> <mi>&theta;</mi> <mo>^</mo> </mover> </math> Returning to the initial value. In this way, the stability of the system can be restored. Subsequently, the process proceeds to S1406, where the gain matrix Γ is corrected to determine the convergence rate, and therefore, the correction is to slow down the convergence rate, thereby similarly recovering the system stability. Subsequently, as shown in S1408, the correction coefficient KLAF (k) calculated by the PID control law is used as a feedback correction coefficient, and the corrected fuel injection quantity Gfuel-KLAF is added to the addition term TTOTAL to calculate the output fuel injection quantity tout (k).

When it is determined in S1402 that the STR control system is stable, the routine proceeds to S1410, where the correction coefficient kstr (k) calculated by the adaptive control law is used as a feedback correction coefficient, and the corrected fuel injection quantity Gfuel-STR (k) is added to the addition term TTOTAL to calculate the output fuel injection quantity tout (k), as shown in the figure.

Returning to the block diagram of fig. 54, the routine proceeds to S1326, and the output fuel injection amount is output and then ended. In embodiment 11, the calculation of the average value of the air/fuel ratio and the like is not limited to the predetermined crank angle of the specific cylinder, but may be performed at the predetermined crank angle of each cylinder, unlike the above-described embodiments. The rest of the constitution is the same as the above embodiment.

In embodiment 11, the adaptive controller may be configured as described above, and the input to the parameter adjusting means may be synchronized with the fuel cycle while calculating the adaptive parameters, as in embodiment 1, and in this case, the computational load of the parameter adjusting means can be significantly reduced, and the controllability can be ensured, and the adaptive controller can be actually used in the machine. In addition, the waiting time is shortened, and the control performance can be improved.

In addition, in embodiment 11, the control amounts of all the cylinders are averaged and input to the parameter adjustment mechanism, and therefore, the fuel condition of a specific cylinder is not greatly affected.

In addition, in embodiments 1 to 11, the average value is a simple average value used, but is not limited thereto, and a weighted average value, a moving average value, a weighted moving average value, or the like may be used. In addition, in the 1 st to 11 th embodiments, the average value is obtained for 1 combustion cycle period in which the input to the parameter adjusting means is performed in synchronization, but the average value may be obtained for the last 2 combustion cycles, or the average value may be obtained for, for example, 2 to 3TDC for less than 1 combustion cycle.

It is needless to say that the Sel-VOBSV and Sel-VSTR may be separately provided as described above and used to separately detect the optimum air/fuel ratio, but the Sel-VOBSVt and Sel-VSTR may show the same detected air/fuel ratio in almost most of the operating region due to the characteristics of the machine and the configuration of the exhaust system, and in this case, the air/fuel ratio may be detected by combining their sampling functions and the output may be used as the input of both the monitor and STR. For example, fig. 36 may be configured to have only the Sel-VOBSV, and its output may be utilized in the monitor and STR. The air/fuel ratio in embodiment 1 and the like is the equivalence ratio used, but it is needless to say that the air/fuel ratio and the equivalence ratio may be set separately. The feedback correction coefficients KSTR, # nKLAF, and KLAF are calculated as multiplication terms, but they may be calculated as addition terms.

In the above embodiment, the adaptive controller has been described by taking STR as an example, but MRACS (typical reference type adaptive control) may be used.

In the above embodiment, the output of the single air/fuel ratio sensor provided at the exhaust merging portion is used, but the present invention is not limited to this, and an air/fuel ratio sensor may be provided for each cylinder, and air/fuel ratio feedback control may be performed for each cylinder based on the air/fuel ratio detected by these sensors.

[ possibility of industrial utilization ]

The fuel injection control device for a multi-cylinder internal combustion engine of the present invention includes: a fuel injection amount control device for controlling the fuel injection amount of a multicylinder internal combustion engine, an adaptive controller for adaptively matching the fuel injection amount as an operation amount with a desired value, and an adaptive parameter adjusting means for calculating an adaptive parameter used by the adaptive controller, wherein the input to the adaptive parameter adjusting means is synchronized with a specific combustion cycle of the internal combustion engine, and the adaptive parameter adjusting means calculates the adaptive parameter based on at least one of an air/fuel ratio and an in-cylinder fuel amount in a fuel control cycle of the internal combustion engine, so that the amount of matrix calculation can be reduced, the load on an on-board computer can be reduced, and the calculation can be completed within 1TDC even with a general on-board computer. Meanwhile, in the case of adaptively calculating the feedback correction coefficient by using an adaptive control algorithm having a parameter adjusting mechanism using the adjustment rule of Landau et al, when the parameter adjusting mechanism is operated for each fuel control cycle of each TDC, the input used in the adaptive adjusting mechanism is used as the value for each combustion cycle, so that the control performance can be improved, the waiting time can be reduced, and the number of times of calculation of the internal variables can be reduced.

Further, since the input to the adaptive parameter adjusting means is synchronized with the fuel control cycle of the specific cylinder of the internal combustion engine, the adaptive parameter adjusting means can calculate in synchronization with the fuel control cycle of the specific cylinder in addition to the above-described operation or effect, and the calculation time can be further shortened, and the adaptive control can be continued even at a high revolution.

Further, since the adaptive controller is operated in synchronization with the fuel control cycle of the internal combustion engine, the adaptive controller that receives the adaptive parameter and calculates the feedback correction coefficient independently of the calculation cycle of the adaptive parameter is operated for each fuel control cycle such as TDC, and even when the number of calculations of the parameter means is reduced to 1 per combustion cycle, the feedback correction coefficient is calculated for each combustion control cycle, so that the air/fuel ratio can be optimally and always feedback-controlled.

Further, a fuel injection control device for an internal combustion engine according to the present invention includes: an air/fuel ratio detecting device for detecting an exhaust air/fuel ratio of an internal combustion engine, a fuel injection amount control means for controlling a fuel injection amount of the internal combustion engine for each fuel control period, and a controller of a recursive form for making the fuel injection amount match a desired value by using the controller of the recursive form as an operation amount based on at least the detected exhaust air/fuel ratio. Therefore, the amount of calculation by the recursive controller can be reduced, the load on the vehicle-mounted computer can be reduced, and the calculation can be completed within 1TDC by the general vehicle-mounted computer.

The "predetermined operation state" referred to herein means a high-speed rotation state of the internal combustion engine. That is, although the time available for 1 calculation is shortened at the time of high-speed rotation, the above configuration enables the adaptive control to be continued even at the time of high-speed rotation, and the above configuration does not deteriorate the controllability even at the time of high-speed rotation because the adaptive parameter and the detected air/fuel ratio are less in variance. Therefore, even in an operating state with a small calculation time, such as at the time of high-speed rotation, adaptive control can be continued, and good controllability of the air/fuel ratio can be ensured.

Further, since the controller of the above recursive form is an adaptive controller, when calculating the feedback correction coefficient by an adaptive control algorithm having a parameter adjusting means using the adjustment rule of Landau et al, the amount of calculation by the adaptive controller having a long calculation time in particular in the recursive form controller can be reduced, the load on the on-board computer can be reduced, and the calculation can be completed within 1TDC even by a general on-board computer.

Further, since the adaptive controller is provided with adaptive parameter adjusting means for calculating adaptive parameters to be used for the adaptive controller, and at least the detected exhaust gas air/fuel ratio is inputted to the adaptive parameter adjusting means, and the adaptive parameter adjusting means is operated in synchronization with the fuel control cycle of the specific cylinder in a predetermined operating state at a cycle longer than the fuel control cycle, the adaptive parameter adjusting means can be operated in synchronization with the fuel control cycle of the specific cylinder in addition to the above-described operation or effect, and the calculation time can be further shortened, and the adaptive control can be continued even at high-speed rotation. The period longer than the fuel control period as described herein means a value corresponding to an integral multiple of the combustion period.

Further, since the detected air/fuel ratio input to the recursive controller is based on a plurality of values detected in a cycle shorter than the operation cycle of the recursive controller, there is no problem that the combustion state of a specific cylinder is strongly reflected only when the specific cylinder is constantly operated at a predetermined crank angle, for example, by averaging the plurality of detected values in addition to the above-described operation and effect.

Further, since the detected air/fuel ratio inputted by the adaptive parameter adjusting means is based on a plurality of values detected in a period shorter than the operating period of the adaptive parameter adjusting means, similarly, for example, by averaging the plurality of detected values, there is no problem that only the combustion state of the specific cylinder is strongly reflected when the specific cylinder is constantly operated at a predetermined crank angle.

Further, the fuel injection control device for an internal combustion engine according to the present invention is constituted by a fuel injection amount control device for controlling a fuel injection amount of the internal combustion engine, an adaptive controller for operating the fuel injection amount as an operation amount in accordance with a desired value, and an adaptive parameter adjusting means for calculating an adaptive parameter used by the adaptive controller, and is provided with an operation state detecting device for detecting an operation state of the internal combustion engine, and changes a control cycle of at least one of the adaptive controller and the adaptive parameter adjusting means in accordance with the detected operation state.

Further, since the control cycle of the adaptive parameter adjusting means is an integral multiple of the control cycle of the adaptive controller, the calculation of the adaptive parameter adjusting means, which requires a particular amount of time, is performed at a rate of 1 time among several times of the control cycle of the STR controller, thereby ensuring the control performance and effectively reducing the amount of calculation, and the number of times of calculation of the STR controller, which actually performs the fuel control, is relatively increased.

Further, since the control cycle of at least one of the adaptive control adaptive parameter adjusting means is changed by a cycle which is an integral multiple of the fuel control cycle, the operation amount obtained by the adaptive controller is continuously used for a period which is an integral multiple of the fuel control cycle, and therefore, the calculation load can be further reduced, and the adaptive control can be easily continued even in an operating state in which the calculation time is reduced such as high-speed rotation, and excellent control performance can be obtained.

Further, since the above-described operating state is at least the rotational speed of the machine, it is possible to reliably detect an operating state in which the calculation time is reduced, such as high-speed rotation, and thus it is possible to reduce the calculation load, and it is possible to continue adaptive control even in such an operating state, and it is possible to obtain excellent control performance.

Further, the fuel injection control device for an internal combustion engine according to the present invention includes: an air/fuel ratio detecting device for detecting an exhaust air/fuel ratio provided in an exhaust system of an internal combustion engine, an operating state detecting device for detecting an operating state of the internal combustion engine including at least a machine rotation speed and a machine load, a fuel injection amount determining device for determining a fuel injection amount of each cylinder at a predetermined crank angle of each cylinder based on at least the detected operating state of the internal combustion engine, a fuel injection device for injecting fuel into each cylinder based on the determined fuel injection amount, and a feedback device for correcting the fuel injection amount, wherein the feedback device comprises an adaptive controller and an adaptive parameter adjusting means for estimating an adaptive parameter, the adaptive controller at least matching a control amount obtained based on an output of the air/fuel ratio detecting device with a desired value, the adaptive parameter detecting means and the adaptive controller being operated at independent operation periods, therefore, the same advantages as described above can be obtained.

The above-described embodiments 1 to 11 have the above-described effects and actions, and when several of these embodiments are combined, it is possible to obtain better control performance in the fuel control device for the internal combustion engine, in other words, it is possible to more accurately perform control of the exhaust gas air/fuel ratio. The most effective action and effect can of course be obtained if all embodiments are applied to the operating conditions of the machine.

The above-described embodiments 1 to 11 can be classified into the following categories according to the action and effect.

Embodiment 1 is an example in which an adaptive controller is used in a fuel control apparatus for an internal combustion engine, and has the effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability. It also has the effect of eliminating the air/fuel ratio deviation of the individual cylinders caused by the specific operating state of the machine. In addition, the operation and effect of preventing the deterioration of the control performance caused when the adaptive controller and the PID controller are switched with the change of the operation state of the machine are provided. Embodiment 7 corresponds to an example of practical use of embodiment 1. Embodiment 7 has an operation and an effect of ensuring excellent control performance of the adaptive controller in all operating states.

Embodiments 2 and 3 relate to a calculation method of an adaptive controller. In embodiment 2, the gain matrix Γ of the adaptive controller is appropriately set according to the operating state of the machine. The method has the effects of improving the control performance (calculation precision) of the adaptive controller, expanding the calculation processing capacity and facilitating the setting of control characteristics. Embodiment 3 is an embodiment in which the gain matrix Γ of an adaptive controller is set in accordance with the output of a device, and has the effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability.

Embodiment 4 relates to processing of a signal input to an adaptive controller. In embodiment 4, since the dead zone is provided in the detected air/fuel ratio input to the adaptive controller, there is an effect that deterioration of the control performance (calculation accuracy) of the adaptive controller due to a slight change in the detected air/fuel ratio can be prevented.

The 5 th and 6 th embodiments relate to a calculation method of an adaptive controller, and particularly to a change speed of an adaptive parameter. In embodiment 5, a limit is set on the change speed of the adaptive parameter used by the adaptive controller, and the control stability of the adaptive controller is improved. Embodiment 5 has an action and an effect of stably calculating the change speed of the adaptive parameter used by the adaptive controller and improving the control performance (calculation accuracy) of the adaptive controller.

Embodiments 8 and 9 relate to a method of calculating an adaptive controller, and particularly to a method of calculating an adaptive controller in a specific operating state. In embodiments 8 and 9, the calculation method of the adaptive controller is changed according to the specific operating state, and there are effects and effects of eliminating the air/fuel ratio deviation of the individual cylinder due to the specific operating state of the adaptive controller.

The 10 th embodiment relates to the calculation of the fuel injection amount determined by the adaptive controller and the individual cylinder air/fuel ratio control means. In the 10 th embodiment, an air/fuel ratio control device controlled by an adaptive controller is added to the device for eliminating the air/fuel ratio deviation of individual cylinders. The control method has the effects of eliminating the air/fuel ratio deviation of individual cylinders and improving the control performance (calculation accuracy) of the adaptive controller. Further, since the timing of detecting the air/fuel ratio is appropriately determined according to the operating state of the machine, there is an effect of improving the accuracy of detecting (calculating) the air/fuel ratio of each cylinder and the control performance (calculation accuracy) of the adaptive controller.

Embodiment 11 relates to a method of connecting an adaptive controller to a device, and corresponds to the modifications of embodiments 1 and 3. In embodiment 11, the direct calculation of the fuel injection amount has an effect of improving the control performance (calculation accuracy) of the adaptive controller. In addition, the stability of the adaptive controller is judged according to the adaptive parameters used by the adaptive controller, and the method has the effects of improving the control stability of the adaptive controller and expanding the calculation processing capacity.

The above-described embodiments are configured by combining the embodiments belonging to the same kind, respectively, and therefore, the operations and effects described in the respective embodiments can be improved. If several kinds of embodiments are combined, the effects multiplied by the effects of the respective embodiments can be obtained, and good controllability, in other words, the air/fuel ratio of exhaust gas can be controlled more accurately as described above can be obtained in the fuel control device for an internal combustion engine.

If the 1 st and 7 th embodiments are used in combination with the 2 nd and 3 rd embodiments, there are effects and effects of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability.

The combination of embodiments 1, 7 and 4 has the effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability.

If the 1 st and 7 th embodiments are used in combination with the 5 th and 6 th embodiments, there are effects of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability.

When the 1 st and 7 th embodiments are used in combination with the 8 th and 9 th embodiments, there are effects that it is possible to eliminate the deviation of the air/fuel ratio of the individual cylinder due to the specific operating state of the adaptive controller, to improve the control performance (calculation accuracy) of the adaptive controller, and to expand the calculation processing capability.

The combined use of the embodiments 1, 7 and 10 can eliminate the deviation of the air/fuel ratio of each individual cylinder, improve the control performance (calculation accuracy) of the adaptive controller, and expand the calculation processing capability.

The combined use of embodiments 1, 7 and 11 has the effects of improving the control stability of the adaptive controller, improving the control performance (calculation accuracy) of the adaptive controller, and expanding the calculation processing capability. In particular, the determination of the stability of the adaptive parameters used by the adaptive controller of embodiment 11 as described above is effective for use in each embodiment.

The combination of embodiments 2, 3 and 4 has the effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability.

The combination of embodiment 2 and embodiment 3 with embodiment 5 and embodiment 6 has the effect of improving the control performance (calculation accuracy) of the adaptive controller and enlarging the calculation processing capability.

When the 2 nd and 3 rd embodiments are used in combination with the 8 th and 9 th embodiments, there are effects that it is possible to eliminate the deviation of the air/fuel ratio of the individual cylinder due to the specific operating state of the adaptive controller, to improve the control performance (calculation accuracy) of the adaptive controller, and to expand the calculation processing capability.

The combination of embodiments 2, 3 and 10 can eliminate the deviation of the air/fuel ratio of each individual cylinder, improve the control performance (calculation accuracy) of the adaptive controller, and expand the calculation processing capability.

The combination of embodiments 2, 3 and 11 has the effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability. In addition, in embodiment 11, since the stability of the adaptive controller is determined based on the adaptive parameters used by the adaptive controller, there are effects of improving the control stability of the adaptive controller and expanding the calculation processing capability.

The combination of embodiments 4, 5 and 6 has the effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability.

When the 4 th embodiment is used in combination with the 8 th and 9 th embodiments, the deviation of the air/fuel ratio of each cylinder due to the specific operation state of the self-adaptive controller can be eliminated, and the control performance (calculation accuracy) of the adaptive controller can be improved and the calculation processing capability can be expanded.

If the 4 th and 10 th embodiments are used in combination, the deviation of the air/fuel ratio of each individual cylinder can be eliminated, the control performance (calculation accuracy) of the adaptive controller can be improved, and the calculation processing capability can be expanded.

If the 4 th and 11 th embodiments are used in combination, the deviation of the air/fuel ratio of each individual cylinder can be eliminated, the control performance (calculation accuracy) of the adaptive controller can be improved, and the calculation processing capability can be expanded. In addition, in embodiment 11, since the stability of the adaptive controller is determined based on the adaptive parameters used by the adaptive controller, there are effects of improving the control stability of the adaptive controller and expanding the calculation processing capability.

When the 5 th and 6 th embodiments are used in combination with the 8 th and 9 th embodiments, the deviation of the air/fuel ratio of each individual cylinder due to the specific operating state of the adaptive controller can be eliminated, and the control performance (calculation accuracy) of the adaptive controller can be improved and the calculation processing capability can be expanded.

If the 5 th, 6 th and 10 th embodiments are used in combination, the deviation of the air/fuel ratio of each individual cylinder can be eliminated, the control performance (calculation accuracy) of the adaptive controller can be improved, and the calculation processing capability can be expanded.

The combined use of embodiments 5, 6 and 11 has an effect of improving the control performance (calculation accuracy) of the adaptive controller and expanding the calculation processing capability. In addition, in embodiment 11, since the stability of the adaptive controller is determined based on the adaptive parameters used by the adaptive controller, there are effects of improving the control stability of the adaptive controller and expanding the calculation processing capability.

The combined use of the embodiments 8, 9 and 10 can eliminate the deviation of the air/fuel ratio of each individual cylinder, improve the control performance (calculation accuracy) of the adaptive controller, and expand the calculation processing capability.

When the 8 th, 9 th and 11 th embodiments are used in combination, there are effects that it is possible to eliminate the deviation of the air/fuel ratio of the individual cylinder due to the specific operating state of the adaptive controller, to improve the control performance (calculation accuracy) of the adaptive controller, and to expand the calculation processing capability. In addition, in embodiment 11, since the stability of the adaptive controller is determined based on the adaptive parameters used by the adaptive controller, there are effects of improving the control stability of the adaptive controller and expanding the calculation processing capability.

If the 10 th and 11 th embodiments are used in combination, the deviation of the air/fuel ratio of each individual cylinder can be eliminated, the control performance (calculation accuracy) of the adaptive controller can be improved, and the calculation processing capability can be expanded. In addition, in embodiment 11, since the stability of the adaptive controller is determined based on the adaptive parameters used by the adaptive controller, there are effects of improving the control stability of the adaptive controller and expanding the calculation processing capability.

Claims (15)

1. A fuel injection control device for an internal combustion engine, comprising:

a. fuel injection amount control device for controlling fuel injection amount of multi-cylinder internal combustion engine,

b. An adaptive controller for adaptively matching the fuel injection amount as an operation amount to a desired value,

c. An adaptive parameter adjusting means for calculating an adaptive parameter used by the adaptive controller;

wherein the input to the adaptive parameter adjusting means is synchronized with a specific combustion cycle of the internal combustion engine, and the adaptive parameter adjusting means performs the calculation of the adaptive parameter based on at least one of an air/fuel ratio and an in-cylinder fuel amount in a fuel control cycle of the internal combustion engine.

2. The fuel injection control apparatus for an internal combustion engine according to claim 1, wherein the input to the adaptive parameter adjusting means is synchronized with a specific cylinder fuel control cycle of the internal combustion engine.

3. The fuel injection control apparatus for an internal combustion engine according to claim 1 or 2, wherein the adaptive controller is operated in synchronization with a fuel control cycle of the internal combustion engine.

4. A fuel injection control device for an internal combustion engine, comprising:

a. an air/fuel ratio detecting device for detecting an exhaust air/fuel ratio of an internal combustion engine,

b. A fuel injection amount control device for controlling the fuel injection amount of the internal combustion engine in each fuel control cycle,

c. A controller of a recursive form for making the fuel injection amount equal to a desired value by using the controller of the recursive form as an operation amount based on at least the detected exhaust air/fuel ratio;

the fuel injection control system is characterized in that the recursive controller is operated in synchronization with a cycle longer than the fuel control cycle in a predetermined operation state.

5. The fuel injection control apparatus of an internal combustion engine according to claim 4, wherein the controller of the recursive form is an adaptive controller.

6. The fuel injection control apparatus for an internal combustion engine according to claim 5, wherein the adaptive controller includes adaptive parameter adjusting means for calculating an adaptive parameter used for the adaptive controller, and at least the detected exhaust air/fuel ratio is inputted into the adaptive parameter adjusting means, and the adaptive parameter adjusting means is operated in synchronization with a cycle longer than the fuel control cycle in a predetermined operating state.

7. The fuel injection control apparatus for an internal combustion engine according to any one of claims 4 to 6, characterized in that the period longer than the fuel control period is a value equivalent to an integral multiple of a combustion period.

8. The fuel injection control device for an internal combustion engine according to claim 6, wherein the detected air/fuel ratio input by the adaptive parameter adjusting means is a numerical value based on a plurality of values detected at a cycle shorter than an operation cycle of the adaptive parameter adjusting means.

9. The fuel injection control device of an internal combustion engine according to any one of claims 4 to 8, characterized in that the detected air/fuel ratio input into the recursive controller is a numerical value based on several values detected at a cycle shorter than an operation cycle of the recursive controller.

10. A fuel injection control apparatus for an internal combustion engine is provided

a. A fuel injection amount control device for controlling the fuel injection amount of the internal combustion engine,

b. An adaptive controller for operating the fuel injection amount as an operation amount in accordance with a desired value,

c. An adaptive parameter adjusting means for calculating an adaptive parameter used by the adaptive controller, the adaptive parameter adjusting means comprising:

d. an operating state detecting means for detecting an operating state of the internal combustion engine,

and changing a control cycle of at least one of the adaptive controller and the adaptive parameter adjusting means in accordance with the detected operating state.

11. The fuel injection control apparatus for an internal combustion engine according to claim 10, wherein a control cycle of the adaptive parameter adjusting means is set to be the same as or larger than a control cycle of the adaptive controller.

12. The fuel injection control apparatus for an internal combustion engine according to claim 10, wherein the control cycle of the adaptive parameter adjusting means is set to be an integral multiple of the control cycle of the adaptive controller.

13. The fuel injection control apparatus for an internal combustion engine according to any one of claims 10 to 12, wherein the control cycle of at least one of the adaptive controller and the adaptive parameter adjusting means is changed with a cycle that is an integral multiple of the fuel control cycle.

14. The fuel injection control apparatus of an internal combustion engine according to any one of claims 10 to 13, characterized in that the above-described operating state is at least a rotation speed of the machine.

15. A fuel injection control device for an internal combustion engine, comprising:

a. an air/fuel ratio detection device provided in an exhaust system of an internal combustion engine for detecting an exhaust air/fuel ratio,

b. an operation state detection means for detecting an operation state of the internal combustion engine including at least a machine rotation speed and a machine load,

c. fuel injection quantity determining means for determining a fuel injection quantity for each cylinder from a predetermined crank angle for each cylinder based on at least the detected operating state of the internal combustion engine,

d. a fuel injection device for injecting fuel into each cylinder according to the determined fuel injection quantity,

e. a feedback device for correcting the fuel injection amount, the feedback device comprising an adaptive controller and an adaptive parameter adjusting means for estimating an adaptive parameter, the adaptive controller correcting the fuel injection amount so that a control amount obtained from at least an output of the air/fuel ratio detecting means matches a desired value,

the adaptive parameter adjusting device and the adaptive controller are operated in independent operation cycles.

Images