JP2006266094A - Air-fuel ratio control device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To set optimum control constants for the overall response characteristics (dead time and n-th delay) of an air-fuel ratio sensor. <P>SOLUTION: The variation timing of a fuel supply amount to an internal combustion engine is detected and the behaviors of outputs from the air-fuel ratio sensor before and after a variation in the fuel supply amount are monitored to detect the response characteristics of the air-fuel ratio sensor in two groups, i.e., the dead time between a time point when the fuel supply amount is varied and a time point when the output from the air-fuel ratio sensor is started to vary and the n-th delay characteristics (n is a positive integer) indicating sensor output variation characteristics after the dead time is passed. When the detected response characteristics are deviated from the response characteristics mounted on an ECU (hereafter, referred to as "the mounted response characteristics"), the mounted response characteristics are corrected and the control constants at which the outputs from the air-fuel ratio sensor become optimum behaviors are calculated by the mounted response characteristics after the correction. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an air-fuel ratio control apparatus that feedback-controls the air-fuel ratio of an air-fuel mixture using a model that simulates the response characteristics of a sensor that detects the air-fuel ratio or oxygen concentration of exhaust gas from an internal combustion engine (engine).

  In recent air-fuel ratio feedback control, a sensor for detecting the air-fuel ratio or oxygen concentration of exhaust gas is installed in the exhaust pipe, and the air-fuel ratio of exhaust gas changes after changing the fuel supply amount to the engine to be controlled. The response characteristic is modeled by dead time + first-order lag characteristic (or second-order lag characteristic), and this model is used to feed back the air-fuel ratio (fuel supply amount) of the air-fuel mixture supplied to the engine based on the output of the sensor. The feedback control system to control is designed. For this reason, when the dead time and the first-order lag change due to engine machine differences, sensor individual differences, engine deterioration, sensor deterioration, etc., if control is performed using the mounted control constants as they are, the controllability of the air-fuel ratio will be increased. Will lead to worsening.

Therefore, conventionally, as shown in Patent Document 1 (Japanese Patent Laid-Open No. 2004-190591), a correction coefficient for correcting a control constant (control gain) according to a simple first-order lag model error is uniformly set, and The control constant is corrected by this correction coefficient.
JP 2004-190591 A (first page, etc.)

  However, since the technique of Patent Document 1 corrects the control constant with a correction coefficient that is uniformly set according to the model error, an optimal control constant is set for sensor deterioration, engine differences, and the like. It was difficult to do. In other words, by setting a correction coefficient according to the model error, the control must be stable in the entire area using the correction coefficient, so the control constant on the stable side must be selected, It has been difficult to set a control constant optimized for the entire response characteristics of the sensor (dead time and n-th order delay).

  The present invention has been made in consideration of such circumstances. Therefore, the object of the present invention is optimal for the entire response characteristics (dead time and n-th order delay) of the sensor that detects the air-fuel ratio or oxygen concentration of the exhaust gas. An object of the present invention is to provide an air-fuel ratio control device that can set a control constant.

  In order to achieve the above object, an invention according to claim 1 is directed to an internal combustion engine based on an output of the sensor using a model that simulates a response characteristic of a sensor that detects an air-fuel ratio or oxygen concentration of exhaust gas of the internal combustion engine. In the air-fuel ratio control device having an air-fuel ratio control means for feedback control of the air-fuel ratio of the air-fuel mixture supplied to the engine, the change timing of the fuel supply amount to the internal combustion engine is detected, and the sensor output behavior before and after the fuel supply amount change The response characteristic of the sensor is monitored, and the n-th order lag characteristic (n is a positive integer) representing the dead time from when the fuel supply amount changes until the sensor output starts to change and the subsequent sensor output change characteristic When the detected response characteristic deviates from the response characteristic mounted on the air-fuel ratio control means (hereinafter referred to as “mounting response characteristic”), the mounting response characteristic is detected. It corrects the sensor output implementation response characteristics after the correction is obtained to calculate the control constants for the optimization behavior. In this way, the mounting response characteristic can be corrected appropriately based on the actually detected sensor response characteristic (dead time + n-order delay), and the sensor output can be reduced with respect to sensor deterioration, internal machine engine differences, and the like. It is possible to calculate the control constant that achieves the optimum behavior on-board, and it is possible to avoid deterioration of controllability caused by deterioration of the sensor, internal machine engine difference, or the like.

  In this case, it is preferable that the mounting control constant is corrected by the control constant calculated by the control constant calculating means. In this way, it is possible to appropriately correct the mounting control constant according to sensor deterioration, machine difference of the internal combustion engine, and the like. The mounting control constant correction method may be switched between executing and stopping the air-fuel ratio feedback.For example, during the feedback execution, by gradually changing the mounting control constant toward the calculated control constant, While avoiding sudden changes in the control state of the internal combustion engine due to a sudden change in the mounting control constant, or during the execution of feedback, the mounting control constant is not changed, but is changed at the next feedback stop, While the feedback is stopped, the mounting control constant may be changed to the calculated control constant all at once.

  Further, as described in claim 3, it is preferable to compare the control constant calculated by the control constant calculating means and the mounting control constant to determine whether the sensor is abnormal by the abnormality determining means. Since sensor response characteristics (dead time and n-th order delay) are used when calculating the control constant, if the sensor response characteristics become abnormal, the calculated control constant also becomes abnormal, and the deviation between the calculated control constant and the mounted control constant Becomes abnormally large. From this relationship, if the calculated control constant is compared with the mounting control constant, it is possible to determine whether or not the sensor is abnormal. In this case, there is an advantage that it is possible to detect a low level abnormality (for example, an abnormality in which both the dead time and the response time are slightly different), which is difficult to detect in the conventional system.

  Further, as in claim 4, when the abnormality determining means determines that there is a sensor abnormality, it is preferable to warn the warning means. In this way, when a sensor abnormality occurs, the driver can be immediately informed of the need for repair, etc., and it is possible to avoid leaving the exhaust emission deterioration state due to the sensor abnormality over a long period of time. it can.

  Further, as described in claim 5, when the response characteristic detected by the response characteristic detection means deviates from the mounting response characteristic, the sensor output is optimal using the transfer function and gain margin or phase margin of the air-fuel ratio feedback control system. You may make it calculate the control constant used as a behavior. Since the gain margin and phase margin can be set according to the optimum behavior regardless of the operating conditions and response characteristics, it is not necessary to specify the gain margin and phase margin for each operating condition and response characteristics, and the same gain margin or Equivalent optimum behavior can be achieved with phase margin. However, according to the present invention, when it is desired to change the optimum behavior depending on the operating condition, a gain margin or a phase margin may be set for each operating condition.

  In this case, if the response characteristic detected by the response characteristic detecting means deviates from the mounting response characteristic as in claim 6, the gain margin or the phase margin is changed according to the error of the response characteristic and the above-mentioned claim is made. The control constant may be calculated by the method of item 5. In this way, even if the sensor is deteriorated, an appropriate control constant can be calculated in consideration of the degree of deterioration of the response characteristic, and the stability of control and the deterioration of responsiveness can be prevented. When changing the gain margin or phase margin, it is not necessary to correct the mounting response characteristics.

  According to another aspect of the present invention, the noise component superimposed on the sensor output is detected by the noise component detection means, and the gain margin or the phase margin is changed according to the magnitude of the noise component to transmit the air-fuel ratio feedback control system. A control constant that causes the sensor output to behave optimally may be calculated using the function and the gain margin or the phase margin. In this way, even if a noise component is superimposed on the output of the sensor due to deterioration of the sensor, deterioration of the internal combustion engine, etc., it is possible to calculate an appropriate control constant that eliminates the influence of the noise component and stabilizes control. Deterioration of responsiveness and responsiveness can be prevented. In this case as well, there is no need to correct the mounting response characteristics.

  In addition, as in claim 8, when the response characteristic detected by the response characteristic detection means deviates from the mounting response characteristic, a control constant that simulates the input or output of the plant that simulates the response characteristic of the sensor is simulated. You may make it calculate by. In this way, an ideal response waveform can be designated (optimal plant input / output behavior can be designated at the waveform level). In addition, since the input of the plant can also be restricted, control can be performed without making the control constant too large from the viewpoint of protection of the internal combustion engine.

  In this case, as in the ninth aspect, the control constant may be calculated by simulation including an error between the response characteristic detected by the response characteristic detecting means and the mounting response characteristic. In this way, even if the sensor deteriorates, an appropriate control constant can be calculated in consideration of the degree of deterioration of the response characteristic, and deterioration of stability and responsiveness can be prevented. In this case, it is not necessary to have a correction coefficient or the like for each operating condition, and it is not necessary to correct the mounting response characteristics.

  Further, as in claim 10, the control constant may be calculated by a simulation including the magnitude of the noise component superimposed on the output of the sensor. In this way, even if a noise component is superimposed on the output of the sensor due to deterioration of the sensor, deterioration of the internal combustion engine, etc., it is possible to calculate an appropriate control constant that eliminates the influence of the noise component and stabilizes control. Deterioration of responsiveness and responsiveness can be prevented.

  In this case, as in claim 11, the first calculation means for calculating the control constant using the gain margin or the phase margin, and the dynamic characteristics of the internal combustion engine to be controlled are simulated in consideration of the response characteristics of the sensor. Second control means for calculating, by simulation, a control constant that makes the input or output of the plant the optimum behavior, and the control constant calculated by the first calculation means and the second calculation means The control constant having higher stability may be selected. In this way, a control constant that emphasizes stability can be selected from the two control constants calculated by the two types of methods, and control with a safe and accurate control constant becomes possible.

  Further, according to a twelfth aspect of the present invention, third calculation means for calculating the control constant by simulation including an error between the response characteristic detected by the response characteristic detection means and the mounting response characteristic, and noise superimposed on the output of the sensor A fourth calculation unit that calculates the control constant by a simulation including the magnitude of the component, and among the control constant calculated by the third calculation unit and the control constant calculated by the fourth calculation unit The one having higher stability may be selected. In this way, a control constant that emphasizes more stability can be selected from the two control constants calculated in consideration of all the situations that occur due to sensor degradation, etc., and is safe and accurate. It is possible to control with various control constants.

  Further, as in claim 13, the control constant is calculated under a predetermined operating condition, and the control constant of the other operating condition is estimated based on the relationship between the predetermined operating condition and the calculated control constant. good. In this way, it is not necessary to repeat the process of calculating the optimal control constant for each operating condition many times, and the detection man-hours and calculation man-hours can be reduced. In addition, since the control constants of the entire region can be changed instantaneously, it is possible to instantly cope with exhaust gas deterioration due to failure. Further, it is possible to estimate a control constant in a region that cannot be measured during the operation of the internal combustion engine (a region that is used transiently but is difficult to keep operating constantly).

  Further, as in claim 14, a control constant is calculated based on a gain margin or phase margin under a certain operating condition, and a control constant under other operating conditions is calculated based on the relationship between the gain margin or phase margin under the certain operating condition. May be estimated. In this way, even if you want to change the optimal behavior for each operating condition, it is not necessary to repeat the process of calculating the optimal control constant by changing the gain margin or phase margin for each operating condition. Man-hours and calculated man-hours can be reduced. In addition, the same effect as in the thirteenth aspect can be obtained.

  Several embodiments embodying the best mode for carrying out the present invention will be described below.

  A first embodiment in which the present invention is applied to an intake port injection engine will be described with reference to FIGS. First, a schematic configuration of the entire engine control system will be described with reference to FIG. An air cleaner 13 is provided at the most upstream portion of the intake pipe 12 of the engine 11 that is an internal combustion engine, and an air flow meter 14 that detects the intake air amount is provided downstream of the air cleaner 13. On the downstream side of the air flow meter 14, a throttle valve 15 whose opening is adjusted by the motor 10 and a throttle opening sensor 16 for detecting the throttle opening are provided. Further, a surge tank 17 is provided on the downstream side of the throttle valve 15, and an intake pipe pressure sensor 18 for detecting the intake pipe pressure is provided in the surge tank 17. The surge tank 17 is provided with an intake manifold 19 for introducing air into each cylinder of the engine 11, and a fuel injection valve 20 for injecting fuel is attached in the vicinity of the intake port of the intake manifold 19 of each cylinder. Yes. A spark plug 21 is attached to each cylinder of the engine 11 for each cylinder, and the air-fuel mixture in the cylinder is ignited by spark discharge of each spark plug 21.

  The intake valve 28 of the engine 11 is provided with a variable intake valve timing mechanism 29 that varies the opening / closing timing (intake valve timing) of the intake valve 28, and the exhaust valve 30 has an opening / closing timing ( A variable exhaust valve timing mechanism 31 that varies the exhaust valve timing) is provided.

  On the other hand, the exhaust pipe 22 of the engine 11 is provided with a catalyst 23 such as a three-way catalyst that purifies CO, HC, NOx, etc. in the exhaust gas. An air-fuel ratio sensor 24 for detecting (oxygen concentration) is provided. A cooling water temperature sensor 25 that detects the cooling water temperature and a crank angle sensor 26 that outputs a pulse signal each time the crankshaft of the engine 11 rotates at a constant crank angle are attached to the cylinder block of the engine 11. Based on the output signal of the crank angle sensor 26, the crank angle and the engine speed are detected.

  Outputs of these various sensors are input to an engine control circuit (hereinafter referred to as “ECU”) 27. The ECU 27 is mainly composed of a microcomputer, and executes various engine control programs stored in a built-in ROM (storage medium), thereby setting the air-fuel ratio of the exhaust gas detected by the air-fuel ratio sensor 24 to the target air. The air-fuel ratio feedback correction coefficient is calculated so as to match the fuel ratio, and functions as air-fuel ratio control means for feedback-controlling the air-fuel ratio (fuel supply amount) of the air-fuel mixture supplied to the engine 11. This air-fuel ratio feedback control system changes the response characteristic of the change in the air-fuel ratio of the exhaust gas after changing the fuel supply amount (fuel injection amount of the fuel injection valve 20) to the engine 11 to be controlled to the dead time + 1st order delay. Designed by modeling with characteristics. This response characteristic may be modeled by dead time + second-order lag characteristic. In short, it may be modeled by dead time + n-order lag characteristic (n is a positive integer).

  Further, the ECU 27 executes each routine shown in FIGS. 4 to 6 to be described later, thereby changing the fuel supply amount stepwise during the steady operation in which the engine 11 is stably operated near the target air-fuel ratio. Thus, the response characteristic of the air-fuel ratio sensor 24 functions as a response characteristic detection unit that detects the dead time and the first-order delay characteristic (n-order delay characteristic) separately.

  Hereinafter, a method of detecting the response characteristic of the air-fuel ratio sensor 24 will be described based on the time charts of FIGS. In the time chart of FIG. 2, the ECU 27 determines that the engine operation state is the steady operation state at the time t1, switches the steady operation flag to ON, and then the response characteristic detection execution condition is satisfied at the time t2. Then, the measurement processing flag is turned ON, and response characteristic measurement processing is started. Here, the response characteristic detection execution condition is that the output of the air-fuel ratio sensor 24 (hereinafter simply referred to as “sensor output”) is stable in the vicinity of the target air-fuel ratio in a steady operation state and the fluctuation amount of the sensor output is small. This is because the response characteristic of the air-fuel ratio sensor 24 cannot be detected with high accuracy when the sensor output is unstable.

  During the period (t2 to t3) from the time t2 when the response characteristic is measured until the predetermined time T1 elapses, the average value of the sensor output is smoothed while continuing the air-fuel ratio feedback control and maintaining the steady operation state. It is calculated by an averaging process such as an arithmetic mean. The average value of the sensor output is stored in the memory of the ECU 27 as “the steady value BA of the sensor output before fuel increase”. If the sensor output is sufficiently stable at the steady-state value BA at the response characteristic measurement start time t2, the sensor output sampled at the response characteristic measurement start time t2 is used as it is before the fuel increase. It may be BA.

  Then, at the time t3 when the measurement of the steady value BA of the sensor output before the fuel increase is finished, the air-fuel ratio feedback control is stopped, the fuel increase flag is turned ON, and the fuel supply amount (fuel injection amount) is stepwise. Increase the amount and start measuring the dead time TA. Thereafter, the point where the change gradient of the sensor output exceeds the threshold value is detected as the change start point.

  Various methods can be considered for calculating the change gradient of the sensor output. In the first embodiment, the difference between the current sensor output AF (N) and the previous sensor output AF (N-1) [AF (N ) −AF (N−1)] is used as the change gradient of the sensor output. Note that the difference [AF (N) −AF (N−S)] between the current sensor output AF (N) and the sensor output AF (N−S) before S times may be used as the change gradient of the sensor output.

  The point at which the change gradient [AF (N) −AF (N−1)] of the sensor output first exceeds the threshold value may be detected as the change start point. In order to prevent erroneous detection of the change start point due to the above, detection of the change start point is continued until the change gradient of the sensor output exceeds the threshold value a predetermined number of times.

In this case, the threshold value may be a fixed value set in advance for the sake of simplification of the arithmetic processing, but in the first embodiment, the operating parameter (for example, the engine speed and the engine speed that affects the noise superimposed on the sensor output). The threshold value is set based on the air / fuel ratio. In this way, an appropriate threshold value can be set according to the actual noise level, and the change start point can be detected more accurately while preventing erroneous detection due to noise or the like.
The change start point may be a predetermined rate response point (for example, a 10% response point).

  After detecting the change start point as described above, the time from the fuel increase timing t3 to the change start point t4 is detected as a dead time TA. At the time t4 when the change start point is detected, the change start point detection flag is switched to ON, and the measurement process of the first-order lag characteristic is started.

  After the start of the measurement process of the first-order lag characteristic, the change amount of the sensor output from the steady value BA of the sensor output before the fuel increase becomes a value after the fuel increase during the period from the fuel increase timing t3 until the predetermined time T0 elapses. A response time TB until a predetermined ratio (for example, 63%) of the change amount (AA-BA) of the sensor output to the steady value AA is calculated. This response time TB is used as information for evaluating the first-order lag characteristic.

  At this time, the steady value AA of the sensor output after the fuel increase may be a learning value learned for each operation region. However, in the first embodiment, when the predetermined time T0 has elapsed from the fuel increase timing t3. At t6, the air-fuel ratio feedback control is restarted, and until the predetermined time T2 elapses, the target air-fuel ratio is shifted to the rich side by the amount of fuel increase, and the fuel supply amount is feedback-corrected so that the sensor output is near the target air-fuel ratio Stabilize at (steady value AA). The average value of the sensor output in this period (t6 to t7) is calculated by averaging processing such as smoothing processing or arithmetic averaging, and this average value is used as “steady value AA of sensor output after fuel increase”. If the sensor output is sufficiently stable at the steady value AA at the time t6 when the predetermined time T0 has elapsed from the fuel increase timing t3, the sensor output at the time t6 when the predetermined time T0 has elapsed is the steady value of the sensor output after the fuel increase. AA may be used.

  As described above, at the time t7 when the measurement of the steady value AA of the sensor output after the fuel increase is completed, the measurement process flag is switched to OFF, and the response characteristic measurement process ends. If the steady value AA of the sensor output after fuel increase is known in advance from the learning value or the like, when the sensor output reaches the steady value AA, the measurement processing flag is switched to OFF and the response characteristic You may make it complete | finish a measurement process.

  After the response characteristic measurement process is completed, the target air-fuel ratio is switched to the target air-fuel ratio during normal operation, and air-fuel ratio feedback control is executed. In the first embodiment, the fuel supply amount is increased and the air-fuel ratio is changed to the rich side to detect the response characteristics. On the contrary, the fuel supply amount is decreased and the air-fuel ratio is set to the lean side. It may be changed. Alternatively, the fuel supply amount is decreased (or increased) stepwise at the start of response characteristic measurement, the air-fuel ratio is shifted to the lean side (or rich side), and the steady value BA of the sensor output is calculated. Increase (or decrease) stepwise and shift the air-fuel ratio to the rich side (or lean side) to calculate the steady value AA of the sensor output. From the steady value BA of the sensor output before the fuel increase (or before the decrease) The response time TB until the change amount of the sensor output reaches a predetermined ratio (for example, 63%) of the change amount (AA-BA) up to the steady value AA of the sensor output after the fuel increase (or after the decrease) is calculated. Anyway.

  Further, the ECU 27 executes routines shown in FIGS. 7 to 11 to be described later, so that the detected response characteristics deviate from the response characteristics mounted on the ECU 27 (hereinafter referred to as “mounting response characteristics”) beyond an allowable value. In this case, the mounting response characteristic is corrected, and functions as a control constant calculation means for calculating a control constant (control gain) at which the output of the air-fuel ratio sensor 24 becomes an optimum behavior with the corrected mounting response characteristic. The processing contents of these routines will be described below.

[Air-fuel ratio sensor response characteristic detection main routine]
The air / fuel ratio sensor response characteristic detection main routine of FIG. 4 is periodically executed during engine operation, and serves as response characteristic detection means in the claims. When this routine is started, first, at step 101, it is determined whether or not the response characteristic detection execution condition is satisfied (whether or not the measurement processing flag is ON). Here, the response characteristic detection execution condition is that the engine operation state is a steady operation state (the steady operation flag is ON), and the output (sensor output) of the air-fuel ratio sensor 24 is stable near the target air-fuel ratio. This is because the response characteristic of the air-fuel ratio sensor 24 cannot be detected with high accuracy when the sensor output is unstable.

  If it is determined in step 101 that the response characteristic detection execution condition is not satisfied, this routine is terminated without performing the subsequent processing, but it is determined that the response characteristic detection execution condition is satisfied. For example, the response characteristic detection process is executed as follows.

  First, in step 102, response characteristic measurement is started, and the process waits until a predetermined time T1 elapses (step 103). After the response characteristic measurement is started, time series data of the sensor output, which is data for calculating the steady value BA of the sensor output before the fuel increase, is stored in the memory of the ECU 27. Then, during the period from the start of response characteristic measurement to the elapse of the predetermined time T1, the air-fuel ratio feedback control is continued and the steady operation state is maintained, and the steady value BA (average value) of the sensor output before fuel increase is obtained. Calculated by averaging processing such as mash processing or arithmetic averaging.

  Then, when a predetermined time T1 has elapsed from the start of response characteristic measurement, the routine proceeds to step 104, the air-fuel ratio feedback control is stopped, the fuel increase flag is turned ON, and in the next step 105, the fuel supply amount (fuel injection) Increase the amount in steps. Thereafter, the process proceeds to step 106 and waits until a predetermined time T0 elapses from the fuel increase timing. Here, the predetermined time T0 is set to a time necessary for the sensor output to stabilize at the steady value AA after the fuel increase. During the period from the start of fuel increase to the elapse of the predetermined time T0, sensor output time-series data serving as data for calculating the response time TB is stored in the memory of the ECU 27.

  Thereafter, when a predetermined time T0 has elapsed from the start of fuel increase, the routine proceeds to step 107, where the air-fuel ratio feedback control is resumed and the target air-fuel ratio is increased by the fuel increase until the predetermined time T2 has elapsed (step 108). The sensor output is stabilized in the vicinity of the target air-fuel ratio (steady value AA) by feedback-correcting the fuel supply amount by shifting to the side. The average value of the sensor output during this period is calculated by averaging processing such as smoothing or arithmetic averaging, and this average value is used as the “steady value AA of sensor output after fuel increase”.

  Then, when the predetermined time T2 has elapsed, the routine proceeds to step 109, where the response characteristic measurement is terminated, the target air-fuel ratio is switched to the target air-fuel ratio during normal operation, and air-fuel ratio feedback control is executed.

[Waste time calculation routine]
The dead time calculation routine of FIG. 5 is periodically executed during engine operation, and serves as response characteristic detection means in the claims. When this routine is started, first, at step 201, it is determined whether or not a dead time calculation condition is satisfied. This dead time calculation condition is that (1) the measurement processing flag is ON (response characteristic is being measured), and (2) the dead time TA has not yet been calculated. These two conditions (1) If there is a condition that does not satisfy either one of (2), the dead time calculation condition is not satisfied, and this routine is terminated without performing the subsequent processing.

  On the other hand, if the above two conditions (1) and (2) are satisfied at the same time, the dead time calculation condition is satisfied and the routine proceeds to step 202 and waits until the fuel supply amount is increased. The processing in step 202 serves as detection means in the claims.

  Thereafter, when the fuel supply amount is increased, the routine proceeds to step 203 where the current sensor output change gradient [AF (N) −AF (N−1)] is compared with a threshold value, and the sensor output change is detected. Wait until the gradient [AF (N) −AF (N−1)] exceeds the threshold.

  When the change gradient [AF (N) −AF (N−1)] of the sensor output exceeds the threshold value, the change start point is determined, and the process proceeds to step 204 where the change start point detection counter Count is set. Counting up and proceeding to step 205, it is determined whether or not the count value of the change start point detection counter Count has reached a predetermined value M, and if not, the process returns to step 203 and described above. Repeat the process. Thus, the detection of the change start point is continued until the change gradient [AF (N) −AF (N−1)] of the sensor output exceeds the threshold value a predetermined number of times (M times).

Thereafter, when the change gradient [AF (N) −AF (N−1)] of the sensor output exceeds the threshold value a predetermined number of times (M times) and the count value of the change start point detection counter Count reaches the predetermined value M In step 206, the dead time TA from the fuel increase timing (j) to the change start point (NM) is calculated by the following equation.
TA = {(N−M) −J} × arithmetic cycle Here, (N−M) indicates that the change gradient [AF (N) −AF (N−1)] of the sensor output first exceeded the threshold value. The point is regarded as a change start point, and this change start point is expressed by the cumulative calculation count so far, and J is the fuel increase timing expressed by the cumulative calculation count so far.

[Nth delay characteristic calculation routine]
The n-th order lag characteristic calculation routine of FIG. 6 is periodically executed during engine operation, and serves as response characteristic detection means in the claims. When this routine is started, first, at step 301, it is determined whether or not a response time calculation condition is satisfied. The response time calculation conditions are (1) the measurement processing flag is OFF (response characteristic measurement processing end), (2) the response time TB has not yet been calculated, and these two conditions (1) If there is a condition that does not satisfy either of the conditions (2) and (2), the response time calculation condition is not satisfied, and this routine is terminated without performing the subsequent processing.

  On the other hand, if the above two conditions (1) and (2) are satisfied at the same time, the response time calculation condition is satisfied and the routine proceeds to step 302 and waits until the dead time TA is detected. Thereafter, when the dead time TA is detected, the routine proceeds to step 303 where the sensor output steady value BA before the fuel increase and the sensor after the fuel increase based on the time series data of the sensor output stored in the memory of the ECU 27. The output steady value AA is calculated. Here, the steady value BA of the sensor output before the fuel increase is an average value of the sensor output until the predetermined time T1 elapses from the start of the measurement process, and the steady value AA of the sensor output after the fuel increase is the end of the measurement process. It is the average value of the sensor output of the previous predetermined time T2.

Thereafter, the process proceeds to step 304, and the sensor output AF63 corresponding to a predetermined ratio (for example, 63%) of the change amount (AA-BA) from the steady value BA before the fuel increase to the steady value AA after the fuel increase is calculated by the following equation. calculate.
AF63 = (AA−BA) × predetermined ratio

Thereafter, the process proceeds to step 305, the time series data of the sensor output stored in the memory of the ECU 27 is searched, the time NT when the sensor output reaches AF63 is obtained, then the process proceeds to step 306, and the change start point (N The response time TB from (M) to the time point NT when the sensor output reaches AF63 is calculated by the following equation.
TB = {NT− (NM)} × operation cycle

  In this routine, after completion of the response characteristic measurement process (when the measurement process flag is OFF), the sensor output steady value BA before and after fuel increase is based on the sensor output time-series data stored in the memory of the ECU 27. The steady-state value AA is calculated, but at the start of the response characteristic measurement process (when the measurement process flag is ON), the steady-state values BA and AA of the sensor output before and after the fuel increase are determined in advance by the learning value or the like. In the case of the response characteristic measurement process, AF63 = (AA−BA) × predetermined ratio is calculated before the change start point (dead time TA) is detected, and the change start point (dead time). After the detection of TA), the sensor output may be compared with AF63 every time the sensor output is sampled, and the time point NT when the sensor output reaches AF63 may be detected in real time.

  In this routine, the response time of the first-order lag is calculated, but instead of this, a response time of the second-order lag, or a response time of the third or higher order may be detected. Further, instead of the response time, a coefficient of an nth-order lag transfer function may be calculated.

[Control constant determination main routine]
The control constant determination main routine of FIG. 7 is periodically executed during engine operation, and serves as a control constant calculation means in the claims. When this routine is started, first, at step 401, it is determined whether or not the response characteristic (dead time + n-order delay) of the air-fuel ratio sensor 24 has been detected by the routines of FIGS. If so, it waits until it is detected.

  Thereafter, when the response characteristic is detected, the process proceeds to step 402, where the response characteristic (mounting response characteristic) mounted on the ECU 27 and the response characteristic (detection response characteristic) detected in the routines of FIGS. It is determined whether or not the absolute value of the difference exceeds the allowable value 1. At this time, whether the difference between the mounting response characteristic and the detection response characteristic exceeds the allowable value 1 for the difference between the dead time mounting value and the detection value, and the difference between the mounting value and the detection value for the nth order delay. If both are less than the allowable value 1, it is determined that the absolute value of the difference between the mounting response characteristic and the detection response characteristic does not exceed the allowable value 1 (the mounting response characteristic need not be corrected). The process returns to step 401 and waits until the next response characteristic is detected. Here, the allowable value 1 is set in order to avoid erroneous correction of the control constant due to subtle variations in response characteristics due to the influence of noise or the like superimposed on the output of the air-fuel ratio sensor 24. This allowable value 1 may be set for each of the absolute value, relative value, and operating condition with respect to the response characteristic.

  On the other hand, if the difference between either the dead time or the n-th order delay exceeds the allowable value 1, the absolute value of the difference between the mounting response characteristic and the detection response characteristic exceeds the allowable value 1 (mounting) It is determined that the response characteristic needs to be corrected), and the process proceeds to step 403, where the mounting response characteristic (dead time + n-order delay) is corrected with the detection response characteristic. For example, the implementation value of dead time and n-th order lag is rewritten with the detected value.

  Thereafter, the process proceeds to step 404, and a control constant (control gain) at which the output of the air-fuel ratio sensor 24 becomes the optimum behavior with the corrected mounting response characteristic is calculated. Thereafter, the process proceeds to step 405, and whether or not the absolute value of the difference between the control constant (mounting response characteristic) mounted in the ECU 27 and the control constant (calculated control constant) calculated in step 404 exceeds the allowable value 2 or not. If the absolute value of the difference between the mounting control constant and the calculated control constant does not exceed the allowable value 2, it is determined that there is no need to correct the mounting control constant, and the process returns to step 401 above. Repeat the process.

  On the other hand, if the absolute value of the difference between the mounting control constant and the calculated control constant exceeds the allowable value 2, the process proceeds to step 406 and the mounting control constant correction routine of FIG. Correct.

[Mounting control constant correction routine]
When the mounting control constant correction routine of FIG. 8 is started in step 406 of FIG. 7, first, in step 501, it is determined whether or not the absolute value of the difference between the calculated control constant and the mounting control constant exceeds a predetermined value 1. If the predetermined value 1 is not exceeded, it is determined that there is no need to correct the mounting control constant, and this routine is terminated as it is. This means that when the difference between the calculated control constant and the mounting control constant is small, the influence of noise or the like is considered, so that the mounting control constant is not corrected.

  On the other hand, if the absolute value of the difference between the calculated control constant and the mounting control constant exceeds the predetermined value 1, the process proceeds to step 502 to determine whether or not a prohibition condition for prohibiting correction of the mounting control constant is satisfied. If it is determined that the prohibit condition is satisfied, this routine is terminated as it is. This is to prevent the mounting control constant from being corrected under special conditions.

  On the other hand, if the prohibition condition is not met, the process proceeds to step 503, where it is determined whether the air-fuel ratio feedback is stopped, and the mounting control constant correction method between the air-fuel ratio feedback stop and the execution is as follows. Switch. If the air-fuel ratio feedback is stopped, the process proceeds to step 504, and the mounting control constant is immediately changed to the calculated control constant.

  If air-fuel ratio feedback is being executed, the process proceeds to step 505, where it is determined whether or not the calculated control constant is greater than the mounted control constant depending on whether or not the difference between the calculated control constant and the mounted control constant is greater than zero. If the calculated control constant is larger than the mounting control constant, the process proceeds to step 506, where the mounting control constant is rewritten to a value obtained by adding the predetermined value 3 to the current mounting control constant, and if the calculated control constant is smaller than the mounting control constant, Proceeding to step 507, the mounting control constant is rewritten to a value obtained by subtracting the predetermined value 3 from the current mounting control constant. Thus, during execution of the air-fuel ratio feedback, the mounting control constant is gradually increased toward the calculated control constant by increasing / decreasing the mounting control constant by a predetermined value of 3. This is because if the mounting control constant is suddenly changed during the execution of the air-fuel ratio feedback, the engine control state is suddenly changed and a torque shock or the like occurs.

  Thereafter, the process proceeds to step 508, where it is determined whether or not the absolute value of the difference between the calculated control constant and the mounting control constant is smaller than a predetermined value 2 corresponding to the maximum error on the system. If the absolute value of the difference from the control constant is greater than or equal to the predetermined value 2, the process returns to step 505, and the process of increasing or decreasing the mounting control constant by the predetermined value 3 is repeated. When the absolute value of the difference between the calculated control constant and the mounting control constant becomes smaller than the predetermined value 2, it is determined that the calculated control constant substantially matches the mounting control constant, and this routine is terminated.

Examples of correction of the mounting control constants by the mounting control constant correction routine of FIG. 8 described above are shown in FIGS.
FIG. 9 is a time chart showing an example of correcting the mounting control constant during execution of air-fuel ratio feedback. In the correction example of FIG. 9, when the absolute value of the difference between the calculated control constant and the mounting control constant is determined to exceed the predetermined value 1 during execution of the air-fuel ratio feedback, the calculated control constant and the mounting control constant are changed. Since the difference is equal to or greater than the predetermined value 2, the process of increasing the mounting control constant by the predetermined value 3 by a predetermined period is repeated. This is because if the mounting control constant is changed suddenly during the execution of air-fuel ratio feedback, the engine control state will change suddenly and torque shock etc. will occur, so the mounting control constant will gradually change toward the calculated control constant. A sudden change in the engine control state due to a sudden change in the control constant is avoided. Then, at time t2 when the difference between the calculated control constant and the mounting control constant becomes equal to or smaller than the predetermined value 2, it is determined that the calculated control constant substantially matches the mounting control constant, and the change of the mounting control constant is stopped.

  On the other hand, FIG. 10 is a time chart showing an example of correcting the mounting control constant during air-fuel ratio feedback stop. In the correction example of FIG. 10, when it is determined that the absolute value of the difference between the calculated control constant and the mounted control constant exceeds the predetermined value 1 while the air-fuel ratio feedback is stopped, the mounted control constant is immediately set to the calculated control constant. change. This is because, during the air-fuel ratio feedback stop, even if the mounting control constant is suddenly changed, the effect does not appear in the engine control state.

[Sensor abnormality determination routine]
The sensor abnormality determination routine in FIG. 11 is periodically executed during engine operation, and serves as abnormality determination means in the claims. When this routine is started, first, at step 601, it is determined whether or not a control constant has been calculated. If not, the process waits until it is calculated. Then, when the control constant is calculated, the process proceeds to step 602, a map of the mounting control constant stored in the memory of the ECU 27 is searched, and the mounting control constant corresponding to the current engine operating condition is read.

  Thereafter, the process proceeds to step 603, and an allowable error corresponding to the current engine operating condition is calculated in consideration of, for example, data variation and noise. Then, in the next step 604, it is determined whether or not the difference between the mounting control constant read in step 602 and the calculated control constant is within the allowable error range. Then, it is determined that the air-fuel ratio sensor 24 is normal.

  On the other hand, if the difference between the mounting control constant and the calculated control constant exceeds the allowable error range, the process proceeds to step 606, where it is determined that the air-fuel ratio sensor 24 is abnormal, and the process proceeds to the next step 607, where the abnormality information is displayed. The information is stored in a rewritable non-volatile memory (not shown) such as a backup RAM of the ECU 27, the warning lamp 32 is turned on (or flashes), and a warning is given to a display unit (not shown) of the instrument panel of the driver's seat. Display and warn the driver.

  In the first embodiment described above, the response characteristic of the air-fuel ratio sensor 24 is detected by dividing it into a dead time and an nth-order lag characteristic (n is a positive integer), and the detected response characteristic is a response implemented in the ECU 27. When a deviation from the characteristic (mounting response characteristic) exceeds an allowable value, the mounting response characteristic is corrected, and a control constant is calculated so that the output of the air-fuel ratio sensor 24 becomes an optimum behavior with the corrected mounting response characteristic. Therefore, the mounting response characteristic can be appropriately corrected based on the actually detected response characteristic (dead time + n-order delay), and the air-fuel ratio can be controlled against the deterioration of the air-fuel ratio sensor 24, the machine difference of the engine 11, and the like. A control constant at which the output of the sensor 24 becomes an optimum behavior can be calculated on-board, and deterioration of controllability caused by deterioration of the air-fuel ratio sensor 24, machine difference of the engine 11, or the like can be avoided.

  Moreover, in the first embodiment, since the control constant calculated using the actually detected response characteristic (dead time + n-order delay) is compared with the mounting control constant, it is determined whether there is an abnormality in the sensor. There is an advantage that it is possible to detect a low level abnormality (for example, an abnormality in which both the dead time and the response time are slightly different), which is difficult to detect the abnormality in the conventional system.

  In general, as shown in FIG. 12, the air-fuel ratio feedback control system is composed of an engine 11 that is a control target P (plant) and a controller C (air-fuel ratio control means), and each is expressed by a transfer function shown in FIG. The

Here, the gain margin is expressed by the ratio of the absolute value of the denominator D to the absolute value of the numerator E of the transfer function C * P of the air-fuel ratio feedback control system.
Gain margin = | Denominator D | / | Numerator E | (1)
Molecule E = b1 * FI * z + b2 * FI
Denominator D = z ^{d +3} + A1 * z ^{d +2} + A2 * z ^{d + 1} -b1 * FI * z -b2 * FI

The phase margin is expressed by the phase of the numerator E and the phase of the denominator D.
Phase margin = phase (numerator E) -phase (denominator D) (2)

  In the second embodiment of the present invention, the response characteristic (dead time + n-order delay) of the air-fuel ratio sensor 24 is detected by the same method as in the first embodiment, and the detected response characteristic deviates from the response characteristic mounted on the ECU 27. In this case, the two control constants ω and ζ are calculated using the transfer function C * P and the gain margin or phase margin of the air-fuel ratio feedback control system. At this time, one of the two control constants ω and ζ is set as a fixed value, and the gain margin or phase margin set value is substituted into the above formula (1) or (2), and the other control constant is set. By solving, the other control constant is calculated backward. By performing such processing for the two control constants ω and ζ, respectively, the two control constants ω and ζ are calculated.

  In this case, the gain margin and phase margin can be set according to the optimal behavior regardless of the operating conditions and response characteristics, so there is no need to specify the gain margin and phase margin for each operating condition and response characteristics. The same optimal behavior can be realized with gain margin and phase margin. However, according to the present invention, when it is desired to change the optimum behavior depending on the operating condition, a gain margin or a phase margin may be set for each operating condition.

  In the third embodiment of the present invention, when the detected response characteristic deviates from the mounting response characteristic beyond the allowable value by executing the control constant determination main routine of FIG. Accordingly, the gain margin or the phase margin is changed, and the control constant is calculated by the same method as in the second embodiment.

  In the control constant determination main routine of FIG. 14, first, in step 701, it is determined whether or not the response characteristic (waste time + n-order delay) of the air-fuel ratio sensor 24 has been detected by the same method as in the first embodiment. If not, wait until it is detected.

  After that, when the response characteristic is detected, the process proceeds to step 702, where the absolute value of the difference between the response characteristic mounted on the ECU 27 (mounting response characteristic) and the detected response characteristic (detection response characteristic) is an allowable value 1. It is determined whether or not the value exceeds the allowable value 1. If the allowable value 1 is not exceeded, it is determined that it is not necessary to correct the mounting control constant, and the process returns to step 701 and waits until the next response characteristic is detected.

  On the other hand, if the absolute value of the difference between the mounting response characteristic and the detection response characteristic exceeds the allowable value 1, the process proceeds to step 703, and the difference between the mounting response characteristic and the detection response characteristic is calculated as an “response characteristic error”. In the next step 704, a gain margin or a phase margin corresponding to the response characteristic error is calculated from the map of FIG. The map in FIG. 15 is set so that the gain margin or phase margin increases as the response characteristic error increases (the control constant ω decreases as the gain margin or phase margin increases).

  Thereafter, the process proceeds to step 705, and the control constant is calculated by the same method as in the second embodiment using the gain margin or phase margin calculated in step 704. Thereafter, the process proceeds to step 706, and whether or not the absolute value of the difference between the control constant (mounting response characteristic) mounted in the ECU 27 and the control constant (calculated control constant) calculated in step 705 exceeds the allowable value 2. If the absolute value of the difference between the mounting control constant and the calculated control constant does not exceed the allowable value 2, it is determined that there is no need to correct the mounting control constant, the process returns to step 701, and the above-mentioned Repeat the process.

  On the other hand, if the absolute value of the difference between the mounting control constant and the calculated control constant exceeds the allowable value 2, the process proceeds to step 707 and the mounting control constant is corrected by the same method as in the first embodiment.

  In the third embodiment described above, when the detected response characteristic deviates from the mounting response characteristic beyond the allowable value, the gain margin or the phase margin is changed in accordance with the error of the response characteristic, and the second embodiment. Therefore, even if the air-fuel ratio sensor 24 is deteriorated, an appropriate control constant can be calculated in consideration of the degree of deterioration of the response characteristics, and the stability of control, The deterioration of responsiveness can be prevented. When changing the gain margin or phase margin, it is not necessary to correct the mounting response characteristics.

  In the fourth embodiment of the present invention, when the control constant determination main routine of FIG. 16 is executed by the ECU 27, the noise component superimposed on the output of the air-fuel ratio sensor 24 exceeds the allowable value. The control margin is calculated by the same method as in the second embodiment by changing the gain margin or the phase margin according to the magnitude of.

  In the control constant determination main routine of FIG. 16, first, at step 711, it is determined whether or not a noise component has been detected by the noise component detection means, and if not detected, it waits until it is detected.

  Thereafter, when a noise component is detected, the process proceeds to step 712, where it is determined whether or not the magnitude of the noise component is greater than or equal to the allowable value 1. If not, the mounting control constant needs to be corrected. If it is determined that there is no noise, the process returns to step 711 and waits until the next noise component is detected.

  On the other hand, if the magnitude of the noise component is greater than or equal to the allowable value 1, the process proceeds to step 713, where the magnitude of the noise component is calculated, and in the next step 714, the gain margin or phase margin corresponding to the magnitude of the noise component is calculated. Is calculated from the map of FIG. The map of FIG. 15 is set so that the gain margin or phase margin increases as the noise component increases (the control constant ω decreases as the gain margin or phase margin increases).

  Thereafter, the process proceeds to step 715, and the control constant is calculated by the same method as in the second embodiment, using the gain margin or phase margin calculated in step 714. Thereafter, the process proceeds to step 716, and whether or not the absolute value of the difference between the control constant (mounting response characteristic) mounted in the ECU 27 and the control constant (calculated control constant) calculated in step 715 exceeds the allowable value 2. If the absolute value of the difference between the mounting control constant and the calculated control constant does not exceed the allowable value 2, it is determined that there is no need to correct the mounting control constant, the process returns to step 711, and the above-mentioned Repeat the process.

  On the other hand, if the absolute value of the difference between the mounting control constant and the calculated control constant exceeds the allowable value 2, the process proceeds to step 717 and the mounting control constant is corrected by the same method as in the first embodiment.

  In the fourth embodiment described above, the gain margin or phase margin is changed according to the magnitude of the noise component superimposed on the output of the air-fuel ratio sensor 24, and the transfer function and gain margin or phase margin of the air-fuel ratio feedback control system are changed. Since the control constant is calculated using the above, even if the noise component is superimposed on the output of the air-fuel ratio sensor 24 due to the deterioration of the air-fuel ratio sensor 24, the deterioration of the engine 11, or the like, it is possible to eliminate the influence of the noise component. Control constants can be calculated, and control stability and responsiveness can be prevented from deteriorating. In this case as well, there is no need to correct the mounting response characteristics.

  The gain margin or the phase margin may be changed based on both the response characteristic error and the magnitude of the noise component.

  In the fifth embodiment of the present invention, the response characteristic (dead time + n-order delay) of the air-fuel ratio sensor 24 is detected in the same manner as in the first embodiment, and the detected response characteristic deviates from the response characteristic mounted on the ECU 27. In this case, a control constant that optimizes the input or output of the plant simulating the response characteristic of the air-fuel ratio sensor 24 is calculated by simulation. As shown in FIG. 17, the simulation model is composed of a controller and a plant (control target). The controller (control logic implemented in the ECU 27) has an output of the air-fuel ratio sensor 24 and an air-fuel ratio feed pack correction coefficient. The plant is configured to have an ideal response waveform, and the plant uses the response characteristic model mounted in the ECU 27 as it is.

  In the fifth embodiment, the control constant is calculated by simulation by executing the control constant calculation routine of FIG. 18 by the ECU 27. The control constant calculation routine of FIG. 18 is a routine for fixing one control constant ζ and optimizing the other control constant ω. When optimizing the control constant ζ, the control constant ω is fixed and the same processing as the control constant calculation routine of FIG. 18 may be performed.

  In the control constant calculation routine of FIG. 18, first, in step 801, the process waits until the mounting response characteristic is changed (corrected). The mounting response characteristic is changed by the same method as in the first embodiment. Then, when the mounting response characteristic is changed, the process proceeds to step 802 to search for the mounting control constants ω and ζ under the changed operating condition, and in the next step 803, as shown in FIG. A fuel step input (fuel disturbance inj) is added to the above, and one of the mounting control constants ω is fixed to the value searched in the above step 802, and the simulation is performed.

  Thereafter, the process proceeds to step 804, where the response and convergence of the output λ (air-fuel ratio) of the plant are detected. Here, as shown in FIG. 19, the responsiveness is detected in the time from the occurrence of the fuel step input until the plant output λ first reaches within, for example, ± 10% of the target value. It is detected in the time from when the fuel step input occurs until the output λ of the plant finally converges within ± 10% of the target value, for example.

  Thereafter, the process proceeds to step 805, and whether or not the output λ of the plant is an ideal waveform (see FIG. 20) depending on whether or not the detected responsiveness and convergence match, that is, the mounting control constant ω is It is determined whether it is smaller or larger than the ideal value. As a result, if it is determined that the detected responsiveness and convergence are the same, it is determined that the mounting control constant ω is smaller than the ideal value, and the process proceeds to step 806, where the mounting control constant ω (n) is set to the previous value ω ( n-1) is updated to a value obtained by adding the predetermined value 2. Thereafter, the process proceeds to step 807, the fuel step input is added with the current mounting control constant ω (n), and the simulation is performed again. In the next step 808, the response and convergence of the output λ of the plant are detected. Proceeding to step 809, it is determined whether or not the detected responsiveness and convergence are inconsistent. If the responsiveness and convergence are not inconsistent, the process returns to step 805. As a result, the mounting control constant ω (n) is increased by a predetermined value 2 by 2 until the response and the convergence do not match, and the fuel step input is added to detect the response and convergence of the plant output λ. Repeat the process.

  After that, when the detected responsiveness and the convergence do not coincide with each other, the process proceeds to step 714 to determine the previous mounting control constant ω (n−1) as the final mounting control constant ω (n). End the routine.

On the other hand, if it is determined in step 805 that the detected responsiveness and the convergence are not the same, it is determined that the mounting control constant ω is larger than the ideal value, and the process proceeds to step 710 where the mounting control constant ω ( n) is updated to a value obtained by subtracting the predetermined value 1 from the previous value ω (n−1). Thereafter, the process proceeds to step 717, the fuel step input is added with the current mounting control constant ω (n), and the simulation is performed again. In the next step 712, the response and convergence of the output λ of the plant are detected. Proceeding to step 713, it is determined whether or not the detected responsiveness and convergence are matched. If the responsiveness and convergence are not matched, the process returns to step 805. As a result, the mounting control constant ω (n) is decreased by a predetermined value 1 by 1 until the responsiveness and the convergence match, and the fuel step input is added to detect the responsiveness and convergence of the plant output λ. Repeat the process. Thereafter, when the detected responsiveness and convergence are matched, this routine is terminated.
By performing such processing, the mounting control constant ω (n) is set so that the plant input does not overshoot.

  In the fifth embodiment described above, every time the detected response characteristic deviates from the mounting response characteristic and the mounting response characteristic is changed (corrected), control constants that optimize the plant input / output behavior are calculated by simulation. As a result, an ideal response waveform can be specified (the optimal behavior of plant input / output can be specified at the waveform level). In addition, since the plant input can also be restricted, control can be performed from the viewpoint of protecting the engine 11 without making the control constant too large.

  In the sixth embodiment of the present invention, when the detected response characteristic deviates from the mounting response characteristic beyond the allowable value by executing the control constant determination main routine of FIG. The control constant is calculated by a simulation including this.

  In the control constant determination main routine of FIG. 21, when the absolute value of the difference between the mounting response characteristic and the detection response characteristic exceeds the allowable value 1 in the same manner as in the third embodiment (FIG. 14), the response characteristic is obtained. Are calculated (steps 701 to 703).

  Thereafter, the process proceeds to step 705, where an error of the response characteristic is input to the plant, a simulation is performed, and a control constant is calculated. Thereafter, the process proceeds to step 706, and whether or not the absolute value of the difference between the control constant (mounting response characteristic) mounted in the ECU 27 and the control constant (calculated control constant) calculated in step 705 exceeds the allowable value 2. If the absolute value of the difference between the mounting control constant and the calculated control constant does not exceed the allowable value 2, it is determined that there is no need to correct the mounting control constant, the process returns to step 701, and the above-mentioned Repeat the process.

  On the other hand, if the absolute value of the difference between the mounting control constant and the calculated control constant exceeds the allowable value 2, the process proceeds to step 707 and the mounting control constant is corrected by the same method as in the first embodiment.

  In the sixth embodiment described above, the control constant is calculated by simulation including an error between the detection response characteristic and the mounting response characteristic. Therefore, even if the air-fuel ratio sensor 24 deteriorates, the degree of deterioration of the response characteristic is determined. In consideration of this, an appropriate control constant can be calculated, and deterioration of stability and responsiveness can be prevented. In this case, it is not necessary to have a correction coefficient or the like for each operating condition, and it is not necessary to correct the mounting response characteristics.

  In the seventh embodiment of the present invention, when the control constant determination main routine of FIG. 22 is executed by the ECU 27, when the noise component superimposed on the output of the air-fuel ratio sensor 24 exceeds the allowable value, the magnitude of the noise component The control constant is calculated by a simulation including

  In the control constant determination main routine of FIG. 22, when the magnitude of the noise component detected by the noise component detection means exceeds the allowable value 1 in the same manner as in the fourth embodiment (FIG. 16), The size is calculated (steps 711 to 713).

  Thereafter, the process proceeds to step 715, where the magnitude of the noise component is input to the plant and a simulation is performed to calculate a control constant. Thereafter, the process proceeds to step 716, and whether or not the absolute value of the difference between the control constant (mounting response characteristic) mounted in the ECU 27 and the control constant (calculated control constant) calculated in step 715 exceeds the allowable value 2. If the absolute value of the difference between the mounting control constant and the calculated control constant does not exceed the allowable value 2, it is determined that there is no need to correct the mounting control constant, the process returns to step 711, and the above-mentioned Repeat the process.

  On the other hand, if the absolute value of the difference between the mounting control constant and the calculated control constant exceeds the allowable value 2, the process proceeds to step 717 and the mounting control constant is corrected by the same method as in the first embodiment.

  In the seventh embodiment described above, the control constant is calculated by the simulation including the magnitude of the noise component superimposed on the output of the air-fuel ratio sensor 24. Therefore, due to the deterioration of the air-fuel ratio sensor 24, the deterioration of the engine 11, etc. Even if a noise component is superimposed on the output of the sensor, an appropriate control constant that eliminates the influence of the noise component can be calculated, and deterioration of control stability and responsiveness can be prevented.

  Two control constants may be calculated by the method of two of the first to seventh embodiments, and the control constant having the higher stability may be selected by comparing the two control constants.

  For example, in the eighth embodiment of the present invention shown in FIG. 23, the first calculation means 901 for calculating the control constant G1 using the gain margin or the phase margin by the same method as the second embodiment, And a second calculation unit 902 that calculates a control constant G2 having an optimum behavior of the input or output of the plant simulating the response characteristic of the air-fuel ratio sensor 24 by the same method. The first calculation unit 901 includes: The calculated control constant G1 is compared with the control constant G2 calculated by the second calculating means 902 to select the one having higher stability (903). In this way, it is possible to select a control constant that emphasizes more stability from the two control constants G1 and G2 calculated by the two methods, and control with a safe and accurate control constant is possible. It becomes.

  In the ninth embodiment of the present invention shown in FIG. 24, the third calculation means 911 that calculates the control constant G3 by simulation including an error between the detection response characteristic and the mounting response characteristic by the same method as in the sixth embodiment, A fourth calculation unit 912 that calculates a control constant G4 by a simulation including the magnitude of the noise component superimposed on the output of the air-fuel ratio sensor 24 in the same manner as in the seventh embodiment, and the third calculation unit 911 includes The calculated control constant G3 and the control constant G4 calculated by the fourth calculating means 912 are compared to select the one having higher stability (913). In this way, it is possible to select a control constant that emphasizes more stability from the two control constants G3 and G4 calculated in consideration of all the situations that occur due to deterioration of the air-fuel ratio sensor 24 or the like. This enables safe and accurate control constant control.

  In the tenth embodiment of the present invention shown in FIG. 25, a fifth calculation unit 921 that calculates a control constant G5 that optimizes the plant output by simulation and a control constant G6 that optimizes the plant input by simulation are calculated. A sixth calculating means 922, and comparing the control constant G5 calculated by the fifth calculating means 921 with the control constant G6 calculated by the sixth calculating means 922, which has a higher stability. Select (923). In this way, a highly stable control constant can be set in consideration of both plant input and output.

  In the tenth embodiment of the present invention shown in FIG. 26, control constants are calculated under predetermined operating conditions (engine load, engine speed, etc.), and other values are calculated based on the relationship between the predetermined operating conditions and the calculated control constant. The control constant of the operating condition is estimated.

  In this way, it is not necessary to repeat the process of calculating the optimal control constant for each operating condition many times, and the detection man-hours and calculation man-hours can be reduced. In addition, since the control constants of the entire region can be changed instantaneously, it is possible to instantly cope with exhaust gas deterioration due to failure. Furthermore, it is possible to estimate a control constant in a region that cannot be measured during engine operation (a region that is used transiently but is difficult to continue to operate steadily).

  In Example 11 of the present invention shown in FIG. 27, a control constant is calculated based on a gain margin or phase margin under a certain operating condition, and based on the relationship between the gain margin or phase margin under the certain operating condition and the calculated control constant. Thus, control constants under other operating conditions are estimated.

  In this way, even if you want to change the optimal behavior for each operating condition, it is not necessary to repeat the process of calculating the optimal control constant by changing the gain margin or phase margin for each operating condition. Man-hours and calculated man-hours can be reduced. In addition, the same effects as those of the eleventh embodiment can be obtained.

[Other Examples]
Three or more control constants are calculated by the method of three or more of the embodiments described above, and the control constant having the highest stability is selected from the three or more control constants. good.

  Also, keep the mounting control constant to a constant with the maximum margin, and change it only in the direction of narrowing the margin (increasing the control constant) according to the response characteristic error and noise component size. Also good.

  Alternatively, the mounting control constant is set to a constant with no margin (that is, it is assumed that the error / noise component of the response characteristic is 0), and a margin is provided in accordance with the magnitude of the error / noise component of the response characteristic. You may make it change only to.

Further, the control constant may be changed only in the stable direction without moving in the direction of increasing the control constant (in the direction of narrowing the margin).
In addition, as a prohibition condition for prohibiting the change of the control constant, a condition such as during feedback may be used when the control constant is far different from the surrounding control constant.

The detected response characteristic may use an average value of N detections.
In order to reduce the processing load on the microcomputer, the relationship between noise and sensor response characteristic errors and control constants may be calculated and implemented in advance, and these constants may be used. (A map may be mounted in advance to reduce the processing load of the microcomputer, and a constant may be drawn from the map.)

It is a schematic block diagram of the whole engine control system in Example 1 of this invention. 3 is a time chart illustrating a response characteristic detection method according to the first embodiment. 3 is a time chart illustrating an abnormality diagnosis method for the air-fuel ratio sensor according to the first embodiment. 6 is a flowchart showing a process flow of an air-fuel ratio sensor response characteristic detection main routine according to the first embodiment. 6 is a flowchart illustrating a flow of processing of a dead time calculation routine according to the first embodiment. 6 is a flowchart illustrating a flow of processing of an n-th order lag characteristic calculation routine according to the first embodiment. 7 is a flowchart illustrating a flow of processing of a control constant determination main routine according to the first embodiment. 6 is a flowchart illustrating a flow of processing of a mounting control constant correction routine according to the first exemplary embodiment. 6 is a time chart illustrating an example of correcting a mounting control constant during execution of air-fuel ratio feedback according to the first embodiment. 6 is a time chart illustrating an example of correcting the mounting control constant during the air-fuel ratio feedback stop according to the first embodiment. 6 is a flowchart illustrating a flow of processing of a sensor abnormality determination routine according to the first embodiment. 6 is a block diagram of an air-fuel ratio feedback control system of Embodiment 2. FIG. It is a figure explaining the relationship between the transfer function of the air-fuel ratio feedback control system of Example 2, and a gain margin or a phase margin. 12 is a flowchart illustrating a flow of processing of a control constant determination main routine according to a third embodiment. It is a figure which shows an example of the map which calculates a gain margin or a phase margin according to the error (size of a noise component) of a response characteristic. 14 is a flowchart illustrating a flow of processing of a control constant determination main routine according to a fourth embodiment. FIG. 10 is a diagram illustrating a simulation model of Example 5. FIG. 10 is a flowchart illustrating a process flow of a control constant determination main routine according to a fifth embodiment. It is a time chart explaining the behavior of the output λ of the plant after fuel step input. It is a time chart explaining the ideal waveform of plant output λ. FIG. 18 is a flowchart illustrating a process flow of a control constant determination main routine according to a sixth embodiment. FIG. 18 is a flowchart illustrating a process flow of a control constant determination main routine according to a seventh embodiment. FIG. 20 is a flowchart illustrating a process flow of a control constant selection main routine according to an eighth embodiment. It is a flowchart which shows the flow of a process of the control constant selection main routine of Example 9. It is a flowchart which shows the flow of a process of the control constant selection main routine of Example 10. It is a figure explaining the estimation method of the control constant of Example 11. FIG. FIG. 20 is a diagram illustrating a control constant estimation method according to a twelfth embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Engine (internal combustion engine), 12 ... Intake pipe, 15 ... Throttle valve, 18 ... Intake pipe pressure sensor, 19 ... Intake manifold, 20 ... Fuel injection valve, 22 ... Exhaust pipe, 23 ... Catalyst, 24 ... Air-fuel ratio sensor 27 ... ECU (air-fuel ratio control means, response characteristic detection means, control constant calculation means, abnormality determination means, detection means), 32 ... warning lamp (warning means)

Claims (14)

Air-fuel ratio control means for feedback-controlling the air-fuel ratio of the air-fuel mixture supplied to the internal combustion engine based on the output of the sensor using a model simulating the response characteristic of the sensor for detecting the air-fuel ratio or oxygen concentration of the exhaust gas of the internal combustion engine In an air-fuel ratio control device comprising:
Detecting means for detecting a change timing of a fuel supply amount to the internal combustion engine;
The behavior of the sensor output before and after the change of the fuel supply amount is monitored, and the response characteristic of the sensor indicates the dead time until the sensor output starts to change from the time when the fuel supply amount changes and the subsequent sensor output change characteristic. a response characteristic detecting means for detecting the n-th delay characteristic (n is a positive integer);
When the response characteristic detected by the response characteristic detection unit deviates from the response characteristic mounted on the air-fuel ratio control unit (hereinafter referred to as “mounting response characteristic”), the mounting response characteristic is corrected, and the corrected mounting response An air-fuel ratio control apparatus comprising: control constant calculation means for calculating a control constant that causes the sensor output to exhibit an optimum behavior in terms of characteristics.   The air-fuel ratio control apparatus according to claim 1, wherein the control constant calculation unit corrects the mounting control constant based on the calculated control constant.   3. The sky determination unit according to claim 1, further comprising an abnormality determination unit that compares the control constant calculated by the control constant calculation unit with the mounting control constant to determine whether the sensor is abnormal. Fuel ratio control device.   4. The air-fuel ratio control apparatus according to claim 3, further comprising warning means for warning when the abnormality determination means determines that the sensor is abnormal. Air-fuel ratio control means for feedback-controlling the air-fuel ratio of the air-fuel mixture supplied to the internal combustion engine based on the output of the sensor using a model simulating the response characteristic of the sensor for detecting the air-fuel ratio or oxygen concentration of the exhaust gas of the internal combustion engine In an air-fuel ratio control device comprising:
Detecting means for detecting a change timing of a fuel supply amount to the internal combustion engine;
The behavior of the sensor output before and after the change of the fuel supply amount is monitored, and the response characteristic of the sensor indicates the dead time until the sensor output starts to change from the time when the fuel supply amount changes and the subsequent sensor output change characteristic. a response characteristic detecting means for detecting the n-th delay characteristic (n is a positive integer);
When the response characteristic detected by the response characteristic detection unit deviates from the response characteristic mounted on the air-fuel ratio control unit, the sensor output is optimized using the transfer function and gain margin or phase margin of the air-fuel ratio feedback control system An air-fuel ratio control apparatus comprising: control constant calculating means for calculating a control constant that becomes a behavior.   And a means for changing the gain margin or the phase margin according to an error of the response characteristic when the response characteristic detected by the response characteristic detection means deviates from the mounting response characteristic. Item 6. The air-fuel ratio control device according to Item 5. Air-fuel ratio control means for feedback-controlling the air-fuel ratio of the air-fuel mixture supplied to the internal combustion engine based on the output of the sensor using a model simulating the response characteristic of the sensor for detecting the air-fuel ratio or oxygen concentration of the exhaust gas of the internal combustion engine In an air-fuel ratio control device comprising:
Noise component detection means for detecting a noise component superimposed on the sensor output;
By changing the gain margin or phase margin according to the magnitude of the noise component detected by the noise component detection means, the sensor output is optimally operated using the transfer function of the air-fuel ratio feedback control system and the gain margin or phase margin. An air-fuel ratio control apparatus comprising: control constant calculating means for calculating a control constant to be Air-fuel ratio control means for feedback-controlling the air-fuel ratio of the air-fuel mixture supplied to the internal combustion engine based on the output of the sensor using a model simulating the response characteristic of the sensor for detecting the air-fuel ratio or oxygen concentration of the exhaust gas of the internal combustion engine In an air-fuel ratio control device comprising:
Detecting means for detecting a change timing of a fuel supply amount to the internal combustion engine;
The behavior of the sensor output before and after the change of the fuel supply amount is monitored, and the response characteristic of the sensor indicates the dead time until the sensor output starts to change from the time when the fuel supply amount changes and the subsequent sensor output change characteristic. a response characteristic detecting means for detecting the n-th delay characteristic (n is a positive integer);
When the response characteristic detected by the response characteristic detection unit deviates from the response characteristic mounted on the air-fuel ratio control unit, the control constant that simulates the input or output of the plant that simulates the response characteristic of the sensor is simulated. An air-fuel ratio control apparatus comprising: control constant calculating means for calculating   9. The air-fuel ratio control apparatus according to claim 8, wherein the control constant calculating unit calculates the control constant by a simulation including an error between the response characteristic detected by the response characteristic detecting unit and the mounting response characteristic. .   9. The air-fuel ratio control apparatus according to claim 8, wherein the control constant calculation unit calculates the control constant by a simulation including a magnitude of a noise component superimposed on the sensor output.   The control constant calculation means includes a first calculation means for calculating a control constant using a gain margin or a phase margin, and a control constant that optimizes an input or output of a plant that simulates the response characteristics of the sensor by simulation. A second calculating means for calculating, wherein the control constant selected by the first calculating means and the control constant calculated by the second calculating means are selected with higher stability. The air-fuel ratio control apparatus according to any one of claims 5 to 10.   The control constant calculating means includes a third calculating means for calculating the control constant by simulation including an error between the response characteristic detected by the response characteristic detecting means and the mounting response characteristic, and a noise component superimposed on the sensor output. And a fourth calculation means for calculating the control constant by a simulation including the magnitude of the control constant, the control constant calculated by the third calculation means and the control constant calculated by the fourth calculation means 11. The air-fuel ratio control apparatus according to claim 8, wherein the one having higher stability is selected.   The control constant calculating means calculates the control constant under a predetermined operating condition, and estimates a control constant of another operating condition based on a relationship between the predetermined operating condition and the calculated control constant. The air-fuel ratio control apparatus according to any one of claims 1 to 12.   The control constant calculating means calculates the control constant based on the gain margin or phase margin under a certain operating condition, and calculates the control constant under other operating conditions based on the relationship between the gain margin or phase margin under the certain operating condition. The air-fuel ratio control apparatus according to claim 5, wherein the air-fuel ratio control apparatus estimates the air-fuel ratio.

Images