DE19655340B4 - Catalytic converter for IC engine - is monitored by sensors on input and output and with abnormality control using reference values for typical sensor readings - Google Patents

Abstract

The catalytic converter (27) has an input sensor upstream and an output sensor downstream. The sensor outputs are monitored and provide a measure of the effectiveness of the catalytic converter in real time. The separate outputs of the sensors are compared with typical reference values to detect any abnormalities ie. to detect if either sensor is faulty. The typical values for the sensors are the central values of the sensor signals when the engine is operating at a steady rate. The control system detects the rich and lean settings of the fuel/air mixture on both sensors and controls the correction for the fuel input.

Description

The The present invention relates to an abnormality detector for a Air-fuel ratio control system, based on the state of a catalyst for purifying exhaust gases recorded on output values of fuel ratio sensors, the each on the upstream and downstream sides of the catalytic converter and also the state of the air-fuel ratio sensors detected.

For detecting the degree of deterioration of a catalyst with respect to exhaust gas purification, Japanese Unexamined Patent Application Publication No. Hei 6-17640 teaches accordingly DE 43 22 341 A1 A saturation absorption amount detecting device for a catalyst. In this detecting apparatus, the fuel injection quantity (target air-fuel ratio) is continuously varied by a predetermined correction quantity for a predetermined time so that the air-fuel ratio of the exhaust gas approaches a rich or lean region until the initial value of an air-fuel ratio sensor disposed on the downstream side of a catalyst. has reached a saturation determination level. A saturated absorption quantity (maximum absorption quantity) of a catalyst is calculated based on the correction quantity for a fuel injection amount and a time period wherein the correction of the fuel injection amount is performed when the output value of the air-fuel ratio sensor disposed on the downstream side has reached the saturation determination level. In this case, the catalyst is already saturated before the initial value of the air-fuel ratio sensor reaches the saturation determination level. That is, a time delay between the time when the output value of the air-fuel ratio sensor reaches the saturation determination level and the time when the rich or lean component of the exhaust gas that can not be absorbed by the catalyst flows toward the downstream side of the catalyst. As a result, a maximum absorption quantity of the catalyst can not be correctly detected due to this time delay.

In an air-fuel ratio control system, the precision and accuracy of the air-fuel ratio control is significantly deteriorated when an air-fuel ratio sensor can not correctly detect the air-fuel ratio. For this reason, it is desirable to accurately diagnose a condition of an air-fuel ratio sensor. For example, Japanese Unexamined Patent Publication No. Sho 62-225943 discloses "a method for detecting the abnormality of a corresponding oxygen concentration sensor" ( DE 37 10 154 C2 ). According to this method, an abnormality of an air-fuel ratio sensor of the limiting current type (oxygen concentration sensor) is detected based on the relationship between an applied voltage and a detected current.

Although the malfunction such as decoupling (loose contact) and a short circuit of a circuit connected to an air-fuel ratio sensor connected, can be detected, however, in this Case the deterioration of an air-fuel ratio sensor not be recorded. This means, that even if the deterioration of an air-fuel ratio sensor has occurred, a computer as long as the initial value of a Air-fuel ratio sensor is within a normal range, an engine on the basis of an air-fuel ratio controls that this air-fuel ratio sensor has been recorded. As a result, the precision of the Air-fuel ratio control be significantly worsened.

From the already introduced, DE 37 10 154 C2 For example, a method for detecting an abnormality of an oxygen concentration sensor is known. This also provides for monitoring a sensor voltage as a function of a pumping current. If the sensor voltage drops at a high pumping current, the monitoring circuit closes for a malfunction of the oxygen concentration sensor.

The It is therefore an object of the present invention to provide an abnormality detecting apparatus for a Air-fuel ratio control system to its loss of function as a result of deterioration an air-fuel ratio sensor to avoid. This object is achieved by an abnormality detecting device according to claim 1 solved.

According to the present invention, claim 1, a central value for an air-fuel ratio on the upstream side of a catalyst is prepared based on the output value of an upstream air-fuel ratio sensor provided on the upstream side of the catalyst. In addition, a central value for an air-fuel ratio becomes on the downstream side of the catalyst sators based on the output value of a downstream air-fuel ratio sensor disposed on the downstream side of the catalyst. An error between the central values of the air-fuel ratio on the upstream and downstream sides of the catalyst is worked out. It is determined whether the upstream air-fuel ratio sensor or the downstream air-fuel ratio sensor is abnormal based on whether the failure is in a predetermined range.

The is called, that the central value for the air-fuel ratio on the upstream Side of the catalyst and that on the downstream side are substantially the same if the upstream and downstream air-fuel ratio sensors is normal. If the error is outside the predetermined range Thus, each of the upstream and downstream air-fuel ratio sensors becomes as damaged or incorrectly determined. As a result, an abnormality of the upstream air-fuel sensor or the downstream Air-fuel sensor accurate be determined.

These and other features and characteristics of the present invention by studying the following detailed description, the appended claims and the drawings, all of which are the subject of the application are. The drawings show:

1 FIG. 3 is a schematic block diagram illustrating an entire engine control system according to a first embodiment of the present invention; FIG.

2 10 is a sectional view showing the detailed structure of an air-fuel ratio sensor on the downstream side of a catalyst;

3 the voltage-current characteristic of an air-fuel ratio sensor on the upstream side of a catalyst,

4 a block diagram for explaining the control principle of an air-fuel ratio control system,

5 FIG. 4 is a flowchart showing a fuel injection amount calculation routine. FIG.

6 a diagram for setting a target air-fuel ratio,

7 FIG. 4 is a flowchart showing an air-fuel ratio central value calculation routine. FIG.

8th a flowchart showing a sensor abnormality determination routine,

9 a first flowchart showing an abnormality determination routine for an upstream air-fuel ratio sensor,

10 a second flowchart showing an abnormality determination routine for the upstream air-fuel ratio sensor,

11A to 11E Time courses for explaining the operation for determining the abnormality of the air-fuel ratio sensor on the upstream side according to a first embodiment;

12 a voltage-current characteristic for clarifying an abnormal output value of the air-fuel ratio sensor on the upstream side,

13 a flow chart showing an abnormal sensor selection routine;

14 FIG. 3 is a flowchart showing an exhaust gas component inflow calculation routine. FIG.

15A to 15E Time sequences which show the process sequence for the calculation of the inflow of an exhaust gas constituent,

16 FIG. 10 is a flowchart illustrating an exhaust component exhaust calculation routine. FIG shows,

17 FIG. 4 is a flow chart showing a calculation routine for an amount of absorbed lean component; FIG.

18 FIG. 4 is a flowchart showing a calculation routine for an amount of absorbed rich ingredient; FIG.

19 FIG. 3 is a flow chart showing a catalyst deterioration detection routine (lean component); FIG.

20 FIG. 4 is a flow chart showing a detection routine of catalyst deterioration (with respect to the rich component); FIG.

21A to 21E Time for explaining the operation for determining the abnormality of the air-fuel ratio sensor on the upstream side according to a second embodiment,

22 a flowchart showing a main fuel injection routine,

23 FIG. 10 is a flowchart showing an abnormality determination routine for an upstream air-fuel ratio sensor in the second embodiment; FIG.

24A to 24C Time courses for explaining the operation regarding the determination of the abnormality in the air-fuel ratio sensor on the upstream side according to a third embodiment,

25 Flowchart showing an abnormality determination routine of an upstream air-fuel ratio sensor in the third embodiment,

26 FIG. 10 is a flowchart showing a control routine for a downstream air-fuel ratio sensor according to a fourth embodiment; FIG.

27 FIG. 10 is a flowchart showing an exhaust component exhaust flow calculation routine according to the fourth embodiment; FIG.

28 FIG. 10 is a flowchart showing an exhaust-gas-component inflow calculation routine according to a fifth embodiment; FIG.

29 FIG. 10 is a flowchart showing an exhaust gas component outflow calculation routine in the fifth embodiment; FIG.

30A to 30F Time sequences which show the process flow for the quantity calculation of an exhaust gas component,

31 FIG. 10 is a flowchart showing the exhaust gas component inflow calculation routine according to a sixth embodiment; FIG.

32 a graph showing the concentration for each component versus the air-fuel ratio A / F of the exhaust gas,

33 a graph showing the concentration of a lean component (O ₂ ) and that of a rich component (Co + H ₂ ) versus an absolute value | λluft | shows λluft at excess air ratio,

34 a kL card,

35 a kR card,

36 FIG. 4 is a flow chart showing a procedure for setting a deterioration determination period (1); FIG.

37 FIG. 4 is a flowchart showing a procedure for setting a determination deterioration period (2); FIG.

38 FIG. 4 is a flow chart illustrating a procedure for setting a deterioration determination period (3); FIG.

39 FIG. 4 is a flowchart showing the exhaust component exhaust flow calculation routine. FIG.

40 FIG. 6 is a graph illustrating a conversion table for converting the output voltage of a downstream air-fuel ratio sensor into an air-fuel ratio RA / F; FIG.

41 FIG. 4 is a flowchart showing a temperature correction routine for a downstream air-fuel ratio sensor element. FIG.

42 FIG. 6 is a graph schematically illustrating the correction of the linearity characteristic of the conversion table based on the temperature of the element of the downstream air-fuel ratio sensor; FIG.

43 FIG. 4 is a flowchart showing a correction routine performed when the fuel supply is interrupted. FIG.

44 FIG. 6 is a graph schematically illustrating the correction of the linearity characteristic of a lean component in the conversion table when the fuel supply is cut off. FIG.

45 FIG. 10 is a flow chart showing an upstream air-fuel ratio condition determining routine; FIG.

46 FIG. 10 is a flowchart showing a downstream air-fuel ratio condition determination routine; FIG.

47 FIG. 4 is a first flowchart showing an exhaust-component-inflow calculation routine having an erasure function. FIG.

48 FIG. 2 is a second flowchart showing an exhaust component inflow calculation routine having an erase function. FIG.

49A to 49F Timines showing a flowchart in a deterioration determination period

50A to 50F Timings that show an operation for the deletion or removal of the inflow of the exhaust gas component,

51 FIG. 3 is a flowchart showing a counting routine for counting a frequency, wherein a catalyst is saturated by a rich / lean component; FIG.

52 FIG. 3 is a flowchart showing a deterioration detection routine with respect to a lean component catalyst. FIG.

53 FIG. 10 is a flowchart showing a deterioration detection routine regarding a rich component catalyst, and FIG

54 a flowchart showing a fuel injection supply delay routine.

(First embodiment)

A first embodiment of the present invention will be described below with reference to FIGS 1 to 13 described in more detail. First, with respect to the 1 the schematic structure of an entire engine control system described. An air purifier or filter 13 is in the most upstream portion of an intake pipe 12 an engine 11 provided, which is designed as an internal combustion engine. An inlet air temperature sensor 14 for detecting the intake air temperature Tam is at the downstream side of the air filter 13 intended. A throttle valve 15 and a throttle opening sensor 16 for detecting the opening position TH of the throttle valve 15 are on the downstream side of the intake air temperature sensor 14 intended. In addition, there is an intake pipe pressure sensor 17 for detecting the pressure PM in the intake pipe 12 on the downstream side of the throttle valve 15 arranged. A pressure equalization tank 18 is on the downstream side of the intake pipe pressure sensor 17 arranged. An intake manifold 19 for feeding air into each cylinder of the engine 11 is to the surge tank 18 connected. Injectors 20 for the injection of fuel are to the branch lines of the intake manifold or manifold 19 connected.

spark 21 are on each cylinder in the engine 11 attached. A high-voltage current that is in an ignition circuit 22 is generated, is to the spark plugs 21 via a distributor 23 created. A crankshaft angle sensor 24 for example, to generate 24 Pulse signals in 720 ° CA (corresponding to two revolutions of a crankshaft) is at the distributor 23 intended. An engine speed Ne is determined based on intervals of output pulse signals of the crankshaft angle sensor 24 detected. A coolant temperature sensor 38 for detecting the temperature Thw of the engine coolant is on the engine 11 attached.

In the meantime, there is an exhaust pipe 24 (an exhaust passage) to an exhaust port (not shown) of the engine 11 via an exhaust or exhaust manifold 25 connected. A catalyst 27 such as the catalytic converter rhodium for reducing harmful components (such as Co, HC, NOx and others) in the exhaust gas is in the exhaust pipe 26 intended. An upstream air-fuel ratio sensor 28 and a downstream air-fuel ratio sensor 29 which generates a linear air-fuel ratio signal (limiting current) in response to the air-fuel ratio in the exhaust gas are on the upstream and downstream sides of the catalyst 27 each arranged.

The upstream and downstream air-fuel ratio sensors 28 and 29 have a structure like in the 2 is shown. The construction will be described in detail below. The air fuel ratio sensors 28 and 29 are at the exhaust pipe 26 attached so that they are inside in the exhaust pipe 26 protrude. The air fuel ratio sensors 28 and 29 are each by an oxygen concentration detection element 51 for generating a limit current corresponding to the oxygen concentration in an air-fuel ratio lean region or according to the carbon monoxide (CO) concentration in an air-fuel ratio rich region, a heater 52 for heating the oxygen concentration detecting element 51 from the inside, a cover 53 for covering the oxygen concentration detecting element 51 and a large number of small holes or holes 54 formed, through which the exhaust gas flows and those in the cover 53 are formed continuously.

With reference to the 3 Next, the voltage-current characteristic of the air-fuel ratio sensors 28 and 29 described. When the air fuel ratio sensors 28 and 29 are operated at a temperature of T1, then they are stable, as represented by a characteristic L1, which in 3 is shown as a solid line. In this case, the rectilinear portion of the characteristic line L1, which is parallel to a voltage axis V, shows the limit current. The limiting current increases or decreases depending on the air-fuel ratio in the exhaust gas. That is, if the air-fuel ratio is lean, then the limit current increases, and on the other hand, if the air-fuel ratio is rich, then the limit current decreases.

With respect to this voltage-current characteristic, a range of the voltage below the voltage of the rectilinear portion that is parallel to the voltage axis V represents a resistance-dominating region. The gradient of the characteristic L1 in the resistance-dominating region becomes fixed by the internal resistance Electrolytic layer of the air fuel ratio sensors 28 and 29 certainly. The internal resistance of the solid electrolyte layer changes depending on the temperature. Consequently, when the temperature of the oxygen concentration detecting element 51 is reduced, then the gradient of the characteristic L1 is reduced due to the reduction of the internal resistance. That is, when the temperature T of the oxygen concentration detecting element 51 T2 is which is lower than T1, then the voltage-current characteristic falls to a characteristic L2, which is shown as a dashed line in 3 is shown. A straight line portion of the characteristic line L2, which is parallel to the voltage axis V, indicates the limit current. Like in the 3 is shown, the limiting current of the characteristic L1 is substantially equal to the characteristic L2. When a positive voltage Vpos is applied to the solid electrolytic layer with respect to the characteristic line L1, the current flowing in the oxygen concentration detecting element becomes 51 flows on the river Ipos limited. When a negative voltage Vneg is applied to the solid electrolytic layer, then, a negative current Ineg is formed, which is proportional only to the temperature without depending on the oxygen concentration passing through the oxygen concentration detecting element 51 flows (see point Pb according to 3 ). In this embodiment, reference is made to a circuit for monitoring the temperature of the air-fuel ratio sensors 28 and 29 waived.

The output values of the aforementioned sensors 28 and 29 are from an electronic control circuit 30 via an input connection 31 read. The electronic control circuit 30 is essentially a CPU 32 , a ROM 33 , RAM 34 as well as a backup RAM 35 educated. The electronic control circuit 30 controls a fuel injection amount TAU, an ignition timing Ig, and other units that need the output values of the sensors (parameters related to the engine operating condition) and generates signals according to the result of the operation of the injector 20 and the ignition circuit 22 via an output connection 36 to the engine 11 to control.

In addition, the electronic control circuit performs 30 a sensor abnormality detection procedure which will be described below to diagnose whether the air-fuel ratio sensor 28 or 29 is abnormal and gives a flare to a warning lamp 37 over the output terminal 36 off to the warning light 37 to light up and warn a driver if the air-fuel ratio sensor 28 or 29 is abnormal.

Next, a method for controlling the air-fuel ratio generated by the electronic control circuit will be described 30 is carried out.

[1] Modeling the too controlling object:

As a model of a system for controlling the air-fuel ratio λ of an engine 11 is a first-degree autodegressive average motion model provided with a dead time of P = 3. The model can be approximated by the following equation (1): λ (k) = a * λ (k-1) + b * FAF (k-3) (1)

in this connection FAF denotes an air-fuel ratio correction factor, "a" and "b" denote a model constant for Determine the responsibility of the model and "k", "k-1" and "k-3" indicate sampling times.

In the above equation (1), in view of the disturbance d, the model of a control system can be approximated by the following equation (2): λ (k) = a * λ (k-1) + b * FAF (k-3) + d (k-1) (2)

It is easy, the above mentioned Model constants "a" and "b" by discretization by sampling with a rotation period (360 ° CA) using a response to determine a discontinuity this means a transfer function G of the control system for to work out the control of the air-fuel ratio λ.

[2] A method for expressing a State variables X (however X denotes a vector quantity).

When the above-mentioned equation (2) is expressed using an expression "a state variable X (k) = [X1 (k), X2 (k), X3 (k), X4 (k)] ^T (T denotes a transposed matrix), is revised, then the determinant is generated, which is represented by the following equation (3):

The following equations can be obtained from the determinant (3) described above: X1 (k + 1) = a × X1 (k) + b × X2 (k) + d (k) = λ (k + 1) (4) X2 (k + 1) = FAF (k - 2) (5) X3 (k + 1) = FAF (k-1) (6) X4 (k + 1) = FAF (k) (7)

[3] Structure of a regulator:

When a regulator is constructed based on the above equations (3) to (7), the air-fuel ratio correction factor FAF is expressed by the following equation (8) using an optimal feedback gain K = [K1, K2, K3, K4] and a state variable X T (k) = [λ (k), FAF (k-3), FAF (k-2), FAF (k-1)]: FAF (k) = K * X T (k) = K1 * λ (k) + K2 * FAF (k-3) + K3'FAF (k-2) + K4'FAF (k-1) (8)

Further, when an integral term ZI (k) for absorbing an error is added to the aforementioned equation (8), the air-fuel ratio correction factor FAF can be obtained by the following equation (9): FAF (k) = K1 * λ (k) + K2 * FAF (k-3) + K3 * FAF (k-2) + K4 * FAF (k-1) + ZI (k) (9)

The integral term ZI (k) is determined by a deviation between the target air-fuel ratio λTG and a current air-fuel ratio λ (k). For this reason, an integration constant Ka can be calculated according to the following equation (10): ZI (k) = ZI (k-1) + Ka {λTG -λ (k)} (10)

4 Fig. 10 is a block diagram illustrating an air-fuel ratio control system whose model is defined as described above. According to the 4 the air-fuel ratio correction factor FAF (k) is expressed using the Z ^-1 converter to produce FAF (k-1). However, the air-fuel ratio correction factor FAF (k-1) in the RAM becomes 34 is stored and read at the next control time. In the present case, FAF (k-1) denotes the last air-fuel ratio correction factor, FAF (k-2) the second-last air-fuel ratio correction factor, and FAF (k-3) the third-last air-fuel ratio correction factor.

According to the 4 For example, a block P1 surrounded by an alternate long and two short dashes line determines a state variable X (k) to a state in which the air-fuel ratio λ (k) is feedback-controlled to approximate the target air-fuel ratio λTG. A block P2 relates to a memory for calculating the integral term ZI (k), and a block P3 calculates a current air-fuel ratio correction factor FAF (k) based on the state variable X (k) determined by the block P1, where the integral term ZI ( k) is calculated by the block P2.

[4] Determination of an optimal Feedback gain K and an integration constant Ka.

An optimal feedback gain K and an integration constant Ka can be set by minimizing a weighting function J obtained by the following equation (11):

In the above equation (11), the value function J minimizes a deviation between the air-fuel ratio λ (k) and a target air-fuel ratio λTG while limiting the influence of the air-fuel ratio correction factor FAF (k). The limit on the air fuel ver ratio correction factor FAF (k) can be controlled by changing the values of the parameters Q and R by weight. For this reason, the function is repeated until an optimum control characteristic is obtained by changing the values with respect to the parameters Q and R of the weight to set an optimal feedback gain K and an integration constant Ka.

Furthermore hang an optimal feedback gain K and an integration constant Ka from the above model constants "a" and "b". Therefore, the stability (Robustness) of a system against the fluctuation of the system Control of the current air-fuel ratio λ (the variations of the parameters) It is necessary to guarantee the optimal feedback gain K and an integration constant Ka in anticipation of the fluctuation of each To fix model constants "a" and "b". That's why the simulation will be performed in anticipation of the fluctuation which currently for each model constant "a" and "b" may occur to provide optimal feedback gain K and to set an integration constant Ka, which is a stability of the system fulfill.

[1] Modeling a control object, [2] a method of expressing a state variable, [3] determining a controller, and [4] determining an optimal feedback gain and an integration constant have been described in detail above. However, they are considered as already set with respect to the air-fuel ratio control system according to this embodiment. The electronic control circuit 30 controls the air-fuel ratio using only the equations (9) and (10) described above.

Next, the operation of the air-fuel ratio control system constructed as described above will be described. The 5 FIG. 12 is a flow chart showing a fuel injection quantity calculation routine executed by a CPU. FIG 32 in the electronic control circuit 30 is performed. This routine or subroutine will be all 360 ° CA in synchronization with the rotation of the motor 11 carried out. When the operation is started by this routine, first, the basic injection quantity Tp is determined based on an intake pressure PM, engine speed Ne, and other boundary conditions in step 101 calculated. The process in step 101 represents a basic injection amount calculating means. In the subsequent step 102 It is determined whether the feedback condition of the air-fuel ratio λ is satisfied or not. The feedback condition is satisfied, for example, when the temperature Thw of the engine coolant is equal to or higher than a predetermined temperature value and the engine 11 not rotated at a high speed and with a large load.

If the feedback condition is met in the present case, then the process goes to step 103 continued. In step 103 An air-fuel ratio correction factor FAF for approximating a detected air-fuel ratio λ is set to a target air-fuel ratio λTG. That is, an air-fuel ratio correction factor FAF is calculated on the basis of the target air-fuel ratio λTG and the air-fuel ratio λ (K) obtained by an upstream air-fuel ratio sensor 28 is detected according to the above equations (9) and (10). The target air-fuel ratio λTG is obtained by using, for example, a map as shown in FIG 6 is shown. According to this map, the target air-fuel ratio λTG is set to a theoretical air-fuel ratio of 14.7 (λ = 1.0) in a range in which the engine 11 rotated at a high speed and a great load and in an area where the engine 11 rotated at a low speed and a low load. In a middle range, the target air-fuel ratio λTG is set to a lean value (λ> 1.0). The process in step 103 FIG. 10 illustrates an air-fuel ratio correction adjusting means. FIG.

If in between in step 102 the feedback condition is not met, then the process goes to step 104 continued. In step 104 For example, an air-fuel ratio correction factor FAF is set to 1.0. Setting the factor FAF to 1.0 means that the air-fuel ratio λ is not corrected. As a result, a so-called open-loop control is performed.

This will be in step 103 or 104 set the air-fuel ratio correction factor FAF, whereupon the process moves to step 105 progresses. In step 105 The fuel injection amount TAU is calculated on the basis of the basic injection amount Tp, an air-fuel ratio correction factor FAF, and another correction factor FALL according to the following equation (12). TAU = Tp · FAF · CASE (12 )

Then, a control signal corresponding to the above-mentioned injection amount TAU to the injectors 20 output. As a result, the time in which the valve of the injector 20 is open (fuel injection time) controlled. As a result, the air-fuel ratio λ is regulated to become the target air-fuel ratio λTG. The process in step 105 represents an injection control means.

Next, the process of determining whether either the upstream air-fuel ratio sensor 28 or the downstream air-fuel ratio sensor 29 is abnormal or unspecified. In this abnormality determination process, first, the center values of the air-fuel ratio on the upstream side of the catalyst become 27 and the air-fuel ratio on the downstream side of the catalyst 27 each by both air-fuel ratio sensors 28 and 29 are calculated by means of an air-fuel ratio central value processing routine as shown in FIG 7 is shown. The error between the air-fuel ratio central value on the upstream side of the catalyst 27 and the air-fuel ratio central value on the downstream side of the catalyst 27 is prepared by a sensor abnormality determination routine as described in U.S.P. 8th is shown. It is determined whether the upstream air-fuel ratio sensor 28 or the downstream air-fuel ratio sensor 29 is abnormal or not on the basis of whether or not the error is in a predetermined range. This process will be described in detail below.

[Calculation of a central value of the air-fuel ratio:]

With reference to the 7 the operation of an air-fuel ratio central value calculating routine for calculating the air-fuel ratio central value on the upstream and downstream sides of the catalyst will be described 27 described in more detail below. In this routine, first in step 111 determines if the engine 11 is operated steadily or not. That is, it is determined whether the operating condition and the load of the engine 11 are substantially constant. An accurate average air-fuel ratio can not be calculated when the engine is running 11 is not steady. For this reason, the routine is terminated without performing the further steps. If, in the meantime, the operation is steady, then the process goes to step 112 where it is determined whether the output RA / F of the downstream air-fuel ratio 29 is in a predetermined range (KRB <RA / F <KRU: where KRB and KRU are predetermined values). If the air-fuel ratio center value can not be accurately calculated, if the output RA / F is outside the predetermined range, then the routine is terminated without performing any further steps.

If, in the meantime, the output value RA / F of the downstream air-fuel ratio sensor 29 is in the predetermined range, then the process goes to step 113 where the average value of the RA / F value is calculated according to the following equation in step 114 until in step 113 it is determined that the first predetermined time has elapsed: The average value of RA / F = {RA / F (current value) + average value of RA / F (last value)} / 2

When the first predetermined time has elapsed, the process goes to step 115 continued. The average value of RA / F obtained in the first predetermined time is set as a center value of RA / F (as an air-fuel ratio central value on the downstream side of the catalyst 27 ). This will be in step 116 an upper limit R (+) and a lower limit R (-) of the central value of RA / F are respectively calculated according to the following equations: Upper limit R (+) = RA / F central value + K1 (K1: constant) Lower limit (R (-) = RA / F central value - K2 (K2: constant)

Next, in the steps 117 and 120 the central value of FA / F (an air-fuel ratio center value on the upstream side of the catalyst 27 ) calculated based on the output value FA / F of the upstream air-fuel ratio sensor 28 in the same way as described above. That is, when the output value FA / F of the upstream air-fuel ratio sensor 28 is in a predetermined range (KFB <FA / F <KFU: KFB and KFU are predetermined values), then the average value of FA / F obtained in the second predetermined time is set as a central value of FA / F (steps 117 to 119a and 119b ). This will be in step 120 an upper limit F (+) and a lower limit F (-) of the center value of FA / F are respectively calculated according to following equations then this routine is terminated: Upper limit F (+) = FA / F central value + K3 (K3: constant) Lower limit F (-) = FA / F central value - K4 (K4: constant)

A sensor abnormality determination routine according to the 8th represents a sensor abnormality determination means. In this routine, in step 121 an error (= FA / F central value - RA / F central value) between the central value of FA / F (the air-fuel ratio central value on the upstream side of the catalyst 27 ) and the central value of RA / F (the air-fuel ratio center value on the downstream side of the catalyst 27 ). In the next step 122 becomes an absolute value of the error calculated in step 121 , compared with a predetermined abnormality determination value k. If | error | ≤ k, since the error is small, it is determined that both the upstream air-fuel ratio sensor 28 and the downstream air-fuel ratio sensor 29 are normal. Then the routine is ended.

If | error | > k, then this means that either the upstream air-fuel ratio sensor 28 or the downstream air-fuel ratio sensor 29 deteriorate or malfunction. The output of the deteriorating sensor indicates an incorrect air-fuel ratio. Therefore, in this case, the process goes to step 123 after which a sensor abnormality detection process is performed.

The 9 and 10 FIG. 15 shows an abnormality determination routine for an upstream air-fuel ratio sensor for determining whether the upstream air-fuel ratio sensor 28 is abnormal or not. With reference to the season tickets according to the 11A to 11E First, a method for determining the abnormality of an upstream air-fuel ratio sensor 28 described in more detail.

If, as in the season tickets according to the 11A to 11E A target air-fuel ratio λTG is suddenly varied toward the lean region at time t1, then an air-fuel ratio correction factor FAF changes so that the value FAF becomes small. When an injection amount is decreased according to the change of the air-fuel ratio correction factor FAF, the air-fuel ratio λ caused by the upstream air-fuel ratio sensor changes 28 is detected, toward the lean area. Further, if the target air-fuel ratio λTG is suddenly changed, then values KCT1 and KCT2 are respectively set to a first count value CT1 and a second count value CT2. The first count value CT1 is decremented as a time elapsed after a time t1, and the second count value CT2 is decremented after a time t2 at which the air-fuel ratio correction factors FAF converge to a certain value. Then, at time t3 in which the value of the first count CT1 is zero, it is determined whether the upstream air-fuel ratio sensor 28 is abnormal based on whether the ratio of the variation ΔλTG of the target air-fuel ratio λTG and the variation ΔFAF of the air-fuel ratio correction factor FAF is within a predetermined range.

That is, when the upstream air-fuel ratio sensor 28 is normal, because an air-fuel ratio correction factor FAF has been set so that the current air-fuel ratio λ coincides with the target air-fuel ratio λTG, then it is considered that a suitable air-fuel ratio correction factor FAF is set according to a deviation between the detected air-fuel ratio λ and the target air-fuel ratio λTG has been. Thus, an air-fuel ratio correction factor FAF varies in accordance with the variation ΔλTG of the target air-fuel ratio λTG. However, if a detected air-fuel ratio λ is not accurate because of the upstream air-fuel ratio sensor 28 is abnormal, then, no appropriate air-fuel ratio correction factor FAF corresponding to the change ΔλTG of the target air-fuel ratio λTG is set. In this case, it can be determined that the upstream air-fuel ratio sensor 28 is abnormal.

The 12 FIG. 10 illustrates embodiments of the output value (restriction current) of the upstream air-fuel ratio sensor. FIG 28 if this sensor 28 is abnormal. In 12 is the characteristic of the sensor output in the case where the sensor is normal, indicated by "La", and the characteristic or the characteristic of the sensor output in the event that the sensor 28 is abnormal due to the deviation of the sensor element 51 or a fault of the heater 52 , shown by "Lb" and "Lc". If the actual air-fuel ratio λ becomes "16", then the air-fuel ratio sensor generates 28 the limiting current Ipa if the air-fuel ratio sensor 28 is normal, which is equal to the current air-fuel ratio (A / F = 16). Meanwhile, the threshold currents Ipb and Ipc are generated when the sensor 28 is abnormal, unlike the limiting current Ipa generated when the sensor 28 is normal. As a result, the actual air-fuel ratio can not be detected. Since no appropriate air-fuel ratio correction factor FAF corresponding to the change ΔλTG is set in such a state, it is determined that the upstream air-fuel ratio sensor 28 is abnormal.

An abnormality determination routine for an upstream air-fuel ratio sensor according to the 9 and 10 is executed in synchronization with a fuel injection operation of the injectors 20 , In this routine, first in step 201 determines whether the difference between the current target air-fuel ratio λTG and the final target air-fuel ratio λTG (i-1) is equal to or greater than a predetermined determination value KλTG, that is, whether the target air-fuel ratio λTG is suddenly changed. If (λTG-λTG (i-1) | <K λTG), then it is determined that the target air-fuel ratio λTG is not changed suddenly, in which case the operation goes to step 205 It is determined whether or not the value of the first count value CT1 has reached zero. If the target air-fuel ratio λTG remains unchanged at the value as before time t1 as in FIG 11A is displayed, the value of the first count value CT1 is kept at zero (an initial value), and this routine is ended without advancing to further steps.

If this is followed by | λTG - λTG (i - 1) | ≥ KλTG (at time t1 according to 11A ), then it is determined that the target air-fuel ratio λTG suddenly changes. The process then proceeds to step 202 wherein a predetermined value KCT1 is set to the first count CT1. The value KCT1 is a value equal to, for example, fifteen times fuel injection. In step 203 is the last target air-fuel ratio λTG (i-1) subtracted from the current target air-fuel ratio λTG to calculate the change ΔλTG of the target air-fuel ratio λTG (ΔλTG = λTG-λTG (i-1)). This will be in step 204 an air-fuel ratio correction factor FAF at this time in the RAM 34 stored as a correction factor FAFBF before the deviation.

The process then goes to step 211 The first count value CT1 is decremented by one. In the next step 212 It is determined whether or not the value of the first count CT1 is zero. Until the value of the first count CT1 is determined to be zero, in step 212 , the count value CT1 continuously becomes one in step 211 decremented.

After the target air-fuel ratio λTG is suddenly varied (after time t1 in FIG 11A ), then a negative determination in step 201 carried out. When in step 205 CT1> 0, then the process goes to step 206 continued. In step 206 the difference between the actual target air-fuel ratio λTG and the final target air-fuel ratio λTG (i-1) is added to the value ΔλTG to renew the value ΔλTG. This will be in step 207 determines whether the absolute value of the difference between the current air-fuel ratio correction factor FAF and the last air-fuel ratio correction factor FAF (i-1) is equal to or smaller than a predetermined value KFAF. That is, it is determined whether the air-fuel ratio correction factor FAF is approaching a certain value.

If | FAF - FAF (i - 1) | > KFAF, that is, before the air-fuel ratio correction factor FAF converges (from the time t1 to the time t2 according to FIG 11B ), then in step 207 "NO" is selected, after which the process on step 208 progresses. In step 208 a predetermined value KCT2 is set to the second count CT2. The value KCT2 is a value equivalent to, for example, 15 times fuel injection.

Meanwhile, | FAF - FAF (i - 1) | ≦ KFAF, that is, after the air-fuel ratio correction factor FAF approaches a predetermined value (after time t2 in FIG 11B ), then in step 207 "YES" is selected, after which the process on step 209 progresses. In step 209 the second count value CT2 is decremented by one. In step 210 It is determined whether the value of the second count CT2 is zero or not. If the value of the second count value CT2 is not zero, the process proceeds to the above-mentioned step 211 continued. Thereafter, the second count value CT2 until the value CT2 is determined to be zero, is continued in step 209 decremented by one.

Then, when the value of both the count value CT1 and the count value CT2 is zero (at the time t3 according to FIG 11D ), then the process goes to step 213 in 10 continued. In step 213 the correction factor FAFBF before the change in the RAM 34 is subtracted from the current air-fuel ratio correction factor FAF to obtain the deviation ΔFAF of the air-fuel ratio correction factor FAF (ΔFAF = FAF-FAFBF). In the following step 214 The values of both counters CT1 and CT2 are set to zero.

This will be in step 215 determines whether the ratio of the absolute value of ΔFAF and that of the value ΔλTG is in a predetermined range (from KCGL to KCGH, for example, KCGL = 0.9, and KCGH = 1.1). If an air-fuel ratio correction factor FAF varies according to the change of the target air-fuel ratio λTG, in step 215 "YES" selected. That is, when the upstream air-fuel ratio sensor 28 generates a normal output signal corresponding to the change of the target air-fuel ratio λTG, then an air-fuel ratio correction factor FAF is changed in a normal manner according to the result of the normal output signal. As a result, the ratio of the absolute value of ΔFAF and that of the value ΔλTG is nearly one. In this case, the upstream air-fuel ratio sensor becomes 28 classified as normal, taking in step 216 an upstream abnormality determination flag XERAF is set to zero and this routine is ended.

Meanwhile, when the air-fuel ratio correction factor FAF excessively varies or is hardly varied according to the change in the target air-fuel ratio λTG, the ratio of the absolute value of ΔFAF and that of the value ΔλTG deviates significantly from the value of one. For this reason, in step 215 "NO" selected. In this case, the upstream fuel ratio sensor becomes 28 classified as abnormal, the process being step by step 217 wherein it is determined whether or not the upstream abnormality determination flag XERAF has already been set to "1". If XERAF = 0, then in step 218 "1" is set for the XERAF flag, after which the routine is ended. In the next process of this routine, if it is determined again that the upstream air-fuel ratio sensor 28 is abnormal, then a predetermined abnormality process in step 219 executed (for example, an alarm lamp 37 lit or feedback control based on the detected air-fuel ratio is disabled).

The 13 FIG. 15 shows a sensor abnormality selection routine for selecting which of the upstream and downstream air-fuel ratio sensors 28 and 29 is abnormal. In this routine, first in step 221 determines whether the upstream air-fuel ratio sensor 28 is abnormal or not. This determination is made based on the result of the abnormality determination performed by the abnormality determination routine for the upstream air-fuel ratio sensor as shown in FIGS 9 and 10 is shown. That is, if the upstream abnormality determination flag XERAF is set to the value "1", then in step 222 determines that the upstream air-fuel ratio sensor 28 is abnormal.

Meanwhile, if XERAF = 0, the process goes to step 223 determining whether it is either the upstream or the downstream air-fuel ratio sensor 28 or 29 is abnormal. This determination is made based on the result of the abnormality determination performed by the sensor abnormality determination routine according to FIG 8th is carried out. If both air fuel ratio sensors 28 and 29 are considered abnormal, then the process goes to step 224 wherein the air-fuel ratio sensor 29 is determined to be abnormal. That is, if in step 221 "NO" is selected (meaning that the upstream air-fuel ratio sensor 28 is determined to be non-abnormal) and in step 223 both air fuel ratio sensors 28 and 29 It can be concluded from this that the abnormal air-fuel ratio sensor is the downstream air-fuel ratio sensor 29 is. If in step 223 "NO" is selected if none of the air-fuel ratio sensors 28 and 29 are abnormal, then this routine is terminated without performing any further process steps.

In this embodiment, it is determined whether the upstream air-fuel ratio sensor 28 or the downstream air-fuel ratio sensor 29 is abnormal based on whether an error between the air-fuel ratio center values is detected by both air-fuel ratio sensors 28 and 29 is in a predetermined range. The air-fuel ratio central value on the upstream side of the catalyst 27 and the air-fuel ratio central value on the downstream side of the catalyst 27 are assumed to be substantially equal if the upstream and downstream air-fuel ratio sensors 28 and 29 are normal. Based on the error between the central values, which are caused by both sensors 28 and 29 Consequently, not only the malfunction of a circuit such as a mismatch or a short circuit but also an abnormality due to a damage of a sensor can be detected. This can not only affect the catalyst 27 can be accurately detected, but also an air-fuel ratio control can be adequately performed using the air-fuel ratios provided by the air-fuel ratio sensors 28 and 29 are detectable.

Based on the error between the central values, which are caused by both sensors 28 and 29 he can not be determined, however, which of the two namely the upstream Kraftstoffverhältnissensors 28 and the downstream fuel ratio sensor 29 has a malfunction.

For this reason, in the first embodiment, it is determined whether the upstream air-fuel ratio sensor 28 is abnormal based on the result of a comparison of the change ΔλTG of the target air-fuel ratio λTG and the deviation ΔFAF of the air-fuel ratio correction factor FAF, if the target air-fuel ratio suddenly changes. As a result, the abnormality of the upstream air-fuel ratio sensor may 28 be determined precisely. Moreover, the sensor having a malfunction may be specified by the sensor abnormality selecting routine as shown in FIG 13 is shown.

[Calculation of the inflow of a Exhaust component]

With reference to the 14 Next, an exhaust-gas-component-inflow calculation routine for calculating the inflow amount of an exhaust gas component introduced into the catalyst will be described 27 flows. This routine is used on every combustion in every cylinder of the engine 11 activated. In this routine, first in the steps 321 and 322 the output value FA / F of the upstream air-fuel ratio sensor 28 with the upper limit F (+) and the lower limit F (-) of the central value FA / F. If F (-) ≦ FA / F ≦ F (+), that is, if FA / F can be regarded as the central value of FA / F, then this routine is terminated without further steps being performed.

Meanwhile, when FA / F> F (+), that is, when the air-fuel ratio on the upstream side of the catalyst 27 has deviated to the side of a lean component from the central value FA / F, then the process goes to step 323 continued. In step 323 For example, a deviation ΔA / F = FA / F-F (+) between FA / F and the upper limit value F (+) of the central value is calculated. In the next step 324 For example, an area on the side of a lean component ΣΔA / F is calculated by integrating the deviation ΔA / F. (please refer 15 ). This will be in step 325 an average value Gav of the amount of exhaust gas which enters the catalyst 27 During the calculation of the lean component side range, ΣΔA / F is calculated based on the operating state of the engine 11 , In the following step 326 becomes the sum FL of a lean component amount in the exhaust gas which enters the catalyst 27 inflow calculated according to the following equation: FL = ΣΔA / F · Gav · k (k: coefficient)

Meanwhile, if FA / F <F (-), that is, if the air-fuel ratio on the upstream side of the catalyst 27 to the side of a rich component from the central value FA / F, the process goes to step 327 continued. In step 327 a deviation ΔA / F (= FA / F-F (-)) between the value FA / F and the lower limit value F (-) of the central value is calculated. In the following step 328 For example, an area on the side of a rich component ΣΔA / F is calculated by integrating the deviation ΔA / F (see FIG 15 ). This will be in step 329 an average value Gav of the amount of exhaust gas which enters the catalyst 27 flows during calculation of the rich component range ΣΔA / F calculated based on the operating condition of the engine 11 , In the next step 330 is the sum FR of a fat component amount in the exhaust gas which is in the catalyst 27 flows, calculated according to the following equation. FR = ΣΔA / F · Gav · k (k: coefficient)

The 15A to 15E show the relationship between ΣΔA / F, Gav: k, FL and FR, which is obtained by the exhaust gas component inflow calculation routine.

[Calculation of the outflow quantity of a Exhaust gas component]

With reference to the 16 For example, an exhaust gas component outflow calculating routine for calculating the exhaust amount of an exhaust gas component will be described below, which is the catalyst 27 flows. This routine is activated at every combustion in each cylinder. In this routine, first in the steps 331 and 332 the output value RA / F of the downstream air-fuel ratio sensor 29 compared with the upper limit F (+) and the lower limit F (-) of the central value of RA / F. If R (-) ≦ RA / F ≦ R (+), that is, if RA / F can be regarded as the central value of RA / F, then this routine is terminated without further steps being performed.

If RA / F> R (+), that is, if the air-fuel ratio on the downstream side of the catalyst 27 on the side of a lean component deviates from the central value of RA / F, then the process goes to step 333 continued. In step 333 a deviation ΔA / F (= RA / F - R (+)) between RA / F and the upper limit R (+) of the central value is calculated. In the following step 334 For example, an area on the side of a lean component ΣΔA / F is calculated by integrating the deviation ΔA / F. This will be in step 335 an average value Gav of the amount of from the catalyst 27 calculated exhaust gas during the calculation of the area on the side of a lean component ΣΔA / F calculated based on the operating condition of the engine 11 , In the following step 336 becomes the sum RL of the lean component quantity of the catalyst 27 outgoing exhaust gas calculated according to the following equation: RL = ΣΔA / F · Gav · k (k: coefficient)

If RA / F <R (-), that is, if the air-fuel ratio on the downstream side of the catalyst 27 to the side of a fat component deviates from the central value RA / F, the process goes to step 337 continued. In step 337 a deviation ΔA / F (= RA / F - R (-)) between RA / F and the lower limit R (-) of the central value is calculated. In the next step 338 For example, an area on the side of a fat component ΣΔA / F is calculated by integrating the deviation ΔA / F. This will be in step 339 the average Gav of the amount of from the catalyst 27 outflowing exhaust gas during the calculation of the range on the part of a fat component ΣΔA / F calculated based on the operating state of the engine 11 , In the next step 340 becomes the sum RR of a fat component amount in the catalyst 27 calculated outgoing exhaust gas according to the following equation: RR = ΣΔA / F · Gav · k (k: coefficient)

[Calculation of the quantity on absorbed lean components]

With reference to the 17 In the following, a lean-component-absorbed-amount calculating routine for calculating the amount of a lean component of an exhaust gas stored in the catalyst will be described below 27 has been absorbed. In this routine is in a first step 341 the sum RL of the lean component quantity of the catalyst 27 effluent exhaust gas from the sum FL of the lean component quantity of the catalyst 27 inflowing exhaust gas subtracts the quantity of CATAL one through the catalyst 27 to obtain absorbed lean component. In the next step 342 is the sum RL of the catalyst 27 outflowing amount of the lean component compared with a predetermined value KL. If RL> KL, then the process goes to step 343 with the amount of CATAL one through the catalyst 27 absorbed lean component is set as a maximum quantity (the saturated quantity) OSImax which can be maximally absorbed before the catalyst. If RL ≤ KL, the process goes to a step 344 where the quantity CATAL is set as a current quantity OSI.

[Calculation of the quantity of absorbed Fat ingredient]

With reference to the 18 For example, an absorbed fat component quantity calculating routine for calculating the amount of an exhaust gas rich component passing through the catalyst is described 27 has been absorbed. In this routine, first in step 351 the sum RR of fat component quantity in the catalyst 27 effluent exhaust gas from the sum FR of the fat component quantity in the catalyst 27 inflowing exhaust gas is subtracted to the amount of CATAR one through the catalyst 27 to receive absorbed fat component. In the next step 352 is the absolute value of the sum RR of the catalyst 27 outflowing fat component amount compared with a predetermined value KR. The reason why the absolute value of the sum RR of the fat component amount is used is that the value of the sum RR of the fat component amount is negative.

If | RR | > KR, then the process moves to step 353 with the amount of CATAR one through the catalyst 27 absorbed fat component is set as a maximum amount (the saturated amount) OSImin, which by the catalyst 27 can be absorbed as much as possible. If | RR | ≤ KR, then the process goes to step 354 with the amount of CATAR set as a current amount of OSI.

[Detection of the deterioration of a catalyst 27 based on the maximum absorbed amount of a lean component]

With reference to the 19 Next, a catalyst deterioration determination routine for detecting deterioration of a catalyst will be described below 27 based on the maximum absorbed amount OSImax of the lean component described in more detail. This routine is used whenever the maximum absorbable amount of OSImax of the lean component in step 343 according to the 17 calculated, executed. In this routine, first in step 361 the maximum absorbed amount OSImax of the lean component divided by a maximum absorbed amount OSIOmax of the lean component, which is determined by the catalyst 27 could be absorbed before the effect of the catalyst 27 deteriorates to obtain a catalyst deterioration degree determination value LDETERIO. In a next step 362 the catalyst deterioration degree determination value LDETERIO is compared with a predetermined deterioration determination value F1. Further, in step 363 determines if a positive determination in step 362 has been performed continuously predetermined times. If a positive determination in step 363 has been carried out, then it is determined that the catalyst 27 has deteriorated. If a negative determination in step 363 has been obtained, then this routine is terminated without a deterioration of the catalyst 27 was determined.

[Detection of deterioration of a catalyst based on the maximum absorbable quantity of a fat component]

With reference to the 20 Next, a catalyst deterioration detecting routine for detecting catalyst degradation will be described below 27 based on the maximum absorbable amount of OSImin of the fat component. This routine is performed each time the maximum absorbable amount of OSImin of the fat component in step 353 according to the 18 is calculated. In this routine, first in step 371 the current maximum absorbed amount of OSImin of the fat component divided by a maximum absorbable amount of OSImin of the fat component derived from the catalyst 27 could be absorbed before the effect of the catalyst 27 deteriorates to obtain a catalyst deterioration degree determination value RDETERIO. In the following step 372 the catalyst deterioration degree determination value RDETERIO is compared with a predetermined deterioration determination value F2. In addition, in step 373 determines if in step 372 a positive determination has been obtained continuously predetermined times. If a positive determination in step 373 was then determined that the catalyst 27 has deteriorated. If a negative determination in step 373 was received, then the routine is terminated without the deterioration of the catalyst 27 has been found.

Only one of the catalyst deterioration determination routines according to the 19 and 20 can be used. If both routines according to the 19 and 20 are carried out, then a method for determining the deterioration of a catalyst 27 if the deterioration of the catalyst 27 is detected by one of the routines and another method for determining the deterioration of a catalyst 27 if the deterioration of the catalyst 27 simultaneously detected by both routines is conceivable.

According to the first embodiment, in 7 the central value of FA / F and the central value of RA / F are calculated. The inflow of an exhaust gas component is in 14 calculated using these central values, wherein the outflow of an exhaust gas component according to 16 is calculated using the same values. However, a sensor output value that is equivalent to the theoretical air-fuel ratio may also be used instead of the central values of FA / F and RA / F to calculate the inflow amount and the outflow amount of an exhaust gas component. If the central values of FA / F and RA / F are used, then the amount of exhaust gas component, which is in and out of the catalyst 27 flows, be accurately calculated even if the air-fuel ratio sensor 28 or 29 worsens and the sensor 28 or 29 can not generate a sensor output signal at the moment when it should produce such that is equivalent to the theoretical air-fuel ratio.

Second Embodiment

A second embodiment according to the present invention, which in the 21A to 23 is detected, detects an abnormality of an upstream air-fuel ratio sensor 28 in a different manner to the first embodiment. In the second embodiment, the Abnormality of an upstream air-fuel ratio sensor 28 detected based on the behavior of a signal output value from the upstream air-fuel ratio sensor 28 when a fuel injection amount is increased due to a low temperature or a high load. The principle of detecting the abnormality will be described below with reference to the time charts according to the 21A to 21E described in more detail.

As in the 21A to 21E is shown, the engine is 11 started by pressing an ignition key at time t10. At this time, a cooling temperature correction factor FWL is set to be larger than 1.0 to increase the fuel injection amount due to the low temperature of the engine coolant. Thereafter, when the temperature of the engine coolant gradually increases and reaches a predetermined value, for example, 40 ° C, then feedback control based on an air-fuel ratio is started at time t11. An air-fuel ratio correction factor FAF is set to a small value against the increase of the fuel injection amount in response to the low temperature of the engine coolant. As a result, the air-fuel ratio correction factor FAF is increased as the coolant temperature correction factor FWL is decreased. The air-fuel ratio correction factor FAF approximates the value 1.0 at time t12, in which the engine 11 is heated, wherein the increase of the fuel injection amount is terminated in response to the low temperature.

At time t10 to time t11, the air-fuel ratio λ (which is determined by the upstream air-fuel ratio sensor 28 detected) is shifted to the side of a rich component in response to the increase of the fuel injection amount. It is diagnosed whether the upstream air-fuel ratio sensor 28 is abnormal based on the change of the air-fuel ratio λ in response to the increase in the fuel injection amount. From the time t11 to the time t12, the increase of the fuel injection amount is continuously continued. However, since an air-fuel ratio correction factor FAF is reduced, the air-fuel ratio λ becomes close to a target air-fuel ratio λTG (λTG = 1.0 in 21D ) held. During this period, it is also diagnosed whether the upstream air-fuel ratio sensor 28 is abnormal based on the change of the air-fuel ratio λ in response to the increase of the fuel injection amount and the decrease of the value FAF.

If a vehicle is in the running state, the fuel injection amount becomes in response to a high load of the engine 11 increased, based on a vehicle acceleration, which is shown at time t13. At this time, the air-fuel ratio feedback control is temporarily disabled, whereby so-called "open-loop control" is performed. During this open-loop control, the air-fuel ratio correction factor FAF is maintained at 1.0. A load correction factor FOTP is increased, wherein the air-fuel ratio λ (which is determined by the upstream air-fuel ratio sensor 28 is detected) is shifted to the side of a fat component. From time t13 to time t14, it is diagnosed whether the upstream air-fuel ratio sensor 28 is abnormal based on the change in the air-fuel ratio λ in response to the increase in the fuel injection amount due to the high load.

If Then, at time t14, a vehicle starts to run, its speed reduce, then the increase the amount of fuel injection due to the high load stops, the load correction factor FOTP returns to the value 1.0. At time t14, as a result of this, the fuel becomes is interrupted, the air-fuel ratio λ temporarily significantly on the Moved page of a lean component. The feedback control of the air-fuel ratio will be resumed after the fuel cut.

The operation described above is performed by a main fuel injection routine according to the 22 and an abnormality determination routine with respect to an upstream air-fuel ratio sensor according to the 23 carried out. The main fuel injection routine as shown in FIG 22 is performed in synchronism with the fuel injection operation of the injectors 20 , When the process is started by this routine, in step 421 Then, the injection amount calculation routine is executed according to 5 executed to calculate the fuel injection amount TAU. This will be in step 422 an average value λAV with respect to the air-fuel ratio λ is calculated with 1/64 averaging operation according to the following equation: λAV = (63 × λAV (i-1) + λ) / 64

In the following step 423 The fuel injection amount TAU is divided by a basic injection amount Tp and an air-fuel ratio leaning value FKG to obtain a fuel correction amount FOTHER for the injection amount TAU. FOTHER = TAU / (Tp · FKG)

The fuel correction amount FOTHER is equivalent to a total correction amount except for the air-fuel ratio lean value FKG, and is obtained as a correction factor including, for example, a coolant temperature correction factor FWL, a load correction factor FOTP, and an air-fuel ratio correction factor FAF. That is, the basic injection amount Tp which varies according to the operating state of the engine 11 is calculated (engine speed Ne, intake pressure PM, etc.) initially set so that the air-fuel ratio λ is controlled so as to take a theoretical air-fuel ratio λth (λth = 1). The dispersion of the fuel injection amount due to differences in the characteristics of engines of the same type and due to aging of an engine is compensated by the air-fuel ratio empirical value FKG. For this reason, in step 423 the total correction quantity required for realizing the air-fuel ratio λth (λth = 1) can be obtained by dividing TAU by the therm tp · FKG.

This will be in step 424 an average value FAV of the correction factor FOTHER is calculated according to the following equation: FAV = (63 * FAV (i-1) + FOTHER) / 64

In the next step 425 the abnormality determination routine is performed on the upstream air-fuel ratio sensor as shown in FIG 23 is pictured.

In step 401 It is determined whether a cooling temperature correction factor FWL exceeds a predetermined determination value KFWL. For example, when the fuel injection amount is increased due to a low temperature (at time t10, as in FIG 21 is shown), then FWL is greater than KFWL. As a result, in step 401 "YES" is selected, the process being step by step 403 jumps. If, however, FWL ≤ KFWL, then in step 402 determines whether a load correction factor FOTP exceeds a predetermined determination value KFOTP. For example, when the fuel injection amount is increased due to a high load (at time t13 according to FIG 21 ), then FOTP is larger than KFOTP. As a result, in step 402 "YES" is selected, the process being step by step 403 progresses. In step 403 It is determined whether the process for determining an air-fuel ratio in the total area in which the engine is completed 11 is operated. If, however, the fuel injection amount is not increased due to the low temperature or the high load ("NO" in the steps 401 and 402 selected) or if the learning of the air-fuel ratio in the total area has not been completed ("NO" becomes in step 403 selected), then the process goes to step 404 continued. In the step 404 For example, a value relating to a count CAFER is reset to zero, and then this routine is ended. That is, when the learning of an air-fuel ratio has not been completed, the dispersion of the injection quantity due to differences in the characteristics of various engines of the same type and due to aging of an engine can not be corrected in a range in which the learning has not been completed is. Only when the learning has been completed is it diagnosed whether an air-fuel ratio sensor 28 is abnormal or not.

If the fuel injection amount is increased, due to the low temperature and the high load and the learning of the fuel ratio is completed (that is, if either in step 401 or 402 "YES" is selected and in step 403 "YES" is selected), then the process goes to a step 405 continued. In step 405 it is determined whether the value of the count CAFER is greater than zero. If CAFER = 0 (initial value) at the beginning, then "NO" in step 405 selected, the process to step 406 progresses. In step 406 For example, a predetermined value KCAFER (for example, a value equivalent to 15 times of fuel injection) is set in the count CAFER, and this routine is ended.

After the value KCAFER has been set in the count CAFER, in step 406 , then in step 405 "YES" is selected if this routine is subsequently executed. For this reason, the process moves to step 407 continues, wherein the count CAFER is decremented by one. In the following step 408 it is determined whether the value of the count CAFER is zero. If it is not zero, the routine is terminated without further action being taken. The process described above is repeated until CAFER = 0.

If then CAFER = 0, then the process goes from step 408 on step 409 continued. An error λAV - 1.0 between the average value λAV of the fuel ratio, which is determined in step 422 and the target air-fuel ratio λTG (λTG = 1.0 in this embodiment) and an error FAV-1.0 between the average value FAV of the correction factor FOTHER calculated in step 424 and a reference value (1.0) are calculated. Furthermore, the ratio (λAV-1.0) / (FAV-1.0) of both errors is worked out, and then it is determined whether the ratio is in a predetermined range (from KFL to KFH, for example KFL = -0.8, KFH = -1.2) ,

If in step 409 " YES "is selected, then the output signal of the upstream air-fuel ratio sensor becomes 28 determined as normal. The process then proceeds to step 410 where an upstream abnormality determination flag XERAF is set to zero and then this routine is terminated.

If in step 409 "NO" is selected, then the upstream air-fuel ratio sensor becomes 28 determined as abnormal. The process goes to step 411 where it is determined whether an upstream abnormality determination flag XERAF has already been set to "1". If XERAF = 0, then the flag XERAF becomes "1" in step 412 set and exit this routine. If the upstream air-fuel ratio sensor 28 again in the next provision in step 409 has been determined to be abnormal, then the process goes to step 413 where a predetermined operation is performed against a sensor abnormality (for example, an alarm lamp will be lit) 37 lit or disabled feedback control based on a detected air-fuel ratio).

According to the second embodiment, as the total fuel correction amount FOTHER for the basic injection amount Tp calculated based on the engine speed Ne and a load applied to the engine 11 is applied (inlet pressure PM) in step 423 furthermore, its mean value FAF is also calculated in step 424 is obtained. Based on the result of a comparison of the average value FAF of the total correction amount FOTHER and the change of the air-fuel ratio λ, which is determined by the upstream air-fuel ratio sensor 28 is detected, whether the upstream air-fuel ratio sensor 28 is abnormal. For this reason, the abnormality of the upstream air-fuel ratio sensor may 28 be determined in a precise manner and immediately as in the first embodiment.

In the second embodiment, it is determined whether the upstream air-fuel ratio sensor 28 is abnormal in response to the increase in the fuel injection amount based on a low temperature or a high load when the engine 11 is operated. However, if the fuel injection amount is decreased, it may also be determined whether the upstream air-fuel ratio sensor 28 is abnormal. For example, in an air-fuel ratio control system having a vapor discharge mechanism for discharging fuel vapor (vaporized fuel) generated in a fuel tank into the intake passage of the engine 11 the amount of fuel coming from the injectors 20 is injected, and reduced in accordance with a discharged fuel quantity. In such a state, it is determined based on the change of the air-fuel ratio λ (the output of the upstream air-fuel ratio sensor 28 ), if the fuel injection amount is decreased, whether the upstream air-fuel ratio sensor 28 is abnormal.

(Third Embodiment)

The 24A to 24C and 25 show a third embodiment for determining the abnormality of an upstream air-fuel ratio sensor 28 by a method different from those of the above-described embodiments. In the third embodiment, it is determined based on the behavior of the air-fuel ratio λ (detected by the upstream air-fuel ratio sensor 28 ) in a temporary operation of the engine 11 Whether the upstream air-fuel ratio sensor 28 is abnormal. The principle of determination of the abnormality of the air-fuel ratio sensor 28 According to the third embodiment will be described below with reference to the time cards according to the 24A to 24C described in more detail.

As in the 24A to 24C is shown, a vehicle is suddenly accelerated at time t21. As a result, the air-fuel ratio λ temporarily fluctuates to the sides of lean and rich components. If a vehicle is suddenly decelerated at time t23, then also takes place a significant fluctuation of the air-fuel ratio λ instead. In this case, based on a difference between a lean peak value λ L and a rich peak value λ R (amplitude of the air-fuel ratio λ), if the air-fuel ratio λ actually fluctuates, it is determined whether the air-fuel ratio sensor 28 is abnormal.

The operation described above is performed by an abnormality determination routine for an upstream air-fuel ratio sensor as shown in FIG 25 is shown. First, in step 501 determines if an operating condition of the engine 11 is steady. The determination as to whether the engine is steady is made based on whether a vehicle is being accelerated or decelerated, whether a target air-fuel ratio λTG has been suddenly changed, or an air-fuel ratio correction factor FAF has suddenly been changed. If the operating condition of the engine 11 is steady, the process moves to step 502 determining whether a value with respect to the count value CAFDT exceeds zero. If the value of the count value CAFDT is zero (an initial value) at the start of the diagnostic operation regarding the abnormality, then "NO" in step 502 This routine is then terminated without further steps being taken.

For example, if a vehicle suddenly accelerates and the engine 11 is in a temporary operation, then in step 501 "NO" (at time t21 according to the 24 ). The process then proceeds to step 503 where a predetermined value KCAFDT is set in the count CAFDT. This will be in step 404 determines whether the current air-fuel ratio λ is greater than a stored lean peak value λ L (which is on the side of a lean component and more than the stored lean peak value λ L). Only if λ> λL, then the process goes to step 505 where the lean peak λ L is renewed by the current air fuel ratio λ. The process then goes to step 506 where it is determined whether the current air-fuel ratio λ is smaller than a stored rich-peak value λR (it is still on the side of a fat component as λR). Only when λ <λR, the process proceeds to a step 507 where the rich peak λR is renewed by the current air-fuel ratio λ. As described above, the lean peak value λ L and a rich peak value λ R become during a period of transient operation of the engine 11 renewed.

If this is the operating condition of the engine 11 returns to a steady state, then the process proceeds to the steps 501 . 502 and 508 where the count CAFDT is decremented by one. In the next step 509 It is determined whether the value of the count value CAFDT is zero. If CAFDT is not zero, the process proceeds to the above-described step 504 continued. That is, a lean peak λ L and a rich peak λ R in steps 504 to 507 during a period in which the count CAFDT is decremented to zero. (from time t21 to t22 according to the 24 ).

If CAFDT = 0 (at time t22 according to the 24 ), then in step 509 "YES" selected. In the next step 510 It is determined whether the difference between a lean peak value λL and a rich peak value λR is equal to or smaller than a predetermined value KAFWD. If λL - λR> KAFWD, then the process goes to step 511 with an upstream abnormality determination flag XELER being set to zero. That is, if λL - λR> KAFWD, then this means that an increased amount of fuel injection due to a sudden acceleration normally corresponds to the output signal of the upstream air-fuel ratio sensor 28 has influenced. As a result, the upstream air-fuel ratio sensor becomes 28 classified as normal. In the next step 515 For example, the lean peak value λL and the rich peak value λR are respectively reset to 1.0 for the next determination of the abnormality, and this routine is finally terminated.

If the result of the determination in step 510 λL - λR ≤ KAFWD, then the process goes to step 512 where it is determined whether the upstream abnormality determination flag XELER has already been set to "1". If the upstream abnormality determination flag XELER has not been set to "1", the process goes to step 513 where the flag XELER is set to "1". If at the next determination in step 510 is determined again on an abnormality, then in step 514 the process against the abnormality of the sensor 28 carried out.

If according to the third embodiment of the engine 11 is in a transient operation, then the amplitude λL - λR of the air-fuel ratio λ, detected by the upstream air-fuel ratio sensor 28 , worked out. Based on the amplitude λL-λR, it is diagnosed whether the upstream air-fuel ratio sensor 28 is abnormal. As a result, it can be determined precisely and immediately, as in the embodiments described above, whether the upstream air fuel ver hältnissensor 28 is abnormal.

(Fourth Embodiment)

In a fourth embodiment, a downstream air-fuel ratio sensor operates 29 as a linear air-fuel (A / F) sensor that generates a linear air-fuel ratio signal according to an air-fuel ratio of the exhaust gas when an operating voltage is applied thereto. However, the downstream air-fuel ratio sensor operates 29 as an oxygen sensor that detects only whether the air-fuel ratio of the exhaust gas is on the lean component side or on the rich component side (whose output signal is inverted to the side of a lean or rich component) when no operating voltage is applied thereto.

[Control of the downstream air-fuel ratio sensor 29]

With reference to the 26 Next, a control routine for a downstream air-fuel ratio sensor for turning on and off an operating voltage with respect to the downstream air-fuel ratio sensor will be described below 29 described in more detail. In this routine, first in step 701 determines whether a value of a count COUNTER which in step 707 is incremented according to the following description, reaches a predetermined value D. That is, it is determined whether a predetermined time has elapsed or not. If the predetermined time has not elapsed, the application of the operating voltage in step 702 blocked. If the predetermined time has elapsed, the operating voltage to the downstream air-fuel ratio sensor 29 in step 708 created.

While applying the operating voltage in step 703 is disabled, an output signal VOX2 from the downstream air-fuel ratio sensor 29 read as an oxygen sensor. In the next step 704 it is determined whether the output signal VOX2 from the oxygen sensor 29 is in a predetermined range (K5 <VOX2 <K6, K5, K6: predetermined value). If the output signal VOX2 is in the predetermined range, a center value operation execution flag FLAG becomes "1" in step 705 set. If the output signal is out of the predetermined range, the central value operation execution flag FLAG in step 706 reset to zero. This will be in step 707 the time count COUNTER is incremented and this routine ends.

While the operating voltage in step 709 is applied, an output signal RA / F from the downstream air-fuel ratio sensor 29 read in as a linear A / F sensor. In the next step 710 It is determined whether the central operation flag FLAG is "1". If FLAG = 1, the central value of RA / F in step 710 calculated. The central value of RA / F is calculated by the same method as in the steps 113 to 115 according to the 7 , In the next step 712 becomes a correction factor KRA / F for absorbing a change of a static characteristic due to the aging process of the downstream air-fuel ratio sensor 29 calculated according to the following equation:

RA / FO is the central value of RA / F when the static characteristic of the downstream air-fuel ratio sensor 29 not varied as a result of aging. The correction factor KRA / F is obtained by dividing the central value RA / FO by the central value of the current RA / F generated in step 711 has been obtained. This will be in step 713 the time count COUNTER is cleared and this routine ends. If the central operation flag FLAG in step 710 is set to "0", the process goes to step 713 continue without the steps 711 and 712 be executed, where. the time count COUNTER is cleared and this routine is ended.

While the application of the operating voltage is disabled, only when the output signal VOX2 from the downstream air-fuel ratio sensor 29 when an oxygen sensor is stably in a predetermined range, the central value of RA / F in step 711 calculated. The RA / F static characteristic correction factor KRA / F is calculated based on the central value of RA / F.

[Calculation of the outflow of a Exhaust gas component]

A calculation routine relating to exhaust gas component outflow according to 27 adds the step 300 to the routine according to 16 added. In this routine will be in step 300 first, the output signal RA / F of the downstream air-fuel ratio sensor 29 with the correction factor KRA / F for multiplying an absorbance of the change of the static characteristic, which in step 712 has been obtained. This obtains the accurate value of RA / F in which an error due to the change of the static characteristic is eliminated. The process in each step after step 300 is the same as that according to the 16 , The lean component quantity RL and the rich component quantity RR of the exhaust gas flow from the catalyst 27 are calculated precisely using the value RA / F, which in step 300 has been corrected.

According to this embodiment, the process according to the 26 and 27 different process the same as that in the first embodiment.

(Fifth Embodiment)

In the first embodiment, when the amount of an exhaust gas component that is in / out of the catalyst 27 flows, is calculated as in the 14 and 16 is shown, then the integrated value of ΣΔA / F (range) of the deviation ΔA / F of the air-fuel ratio is multiplied by the average value Gav of the exhaust gas amount. However, in the fifth embodiment, according to the 28 to 30 calculates an amount of an exhaust gas component each time a deviation ΔA / F is calculated. The method of calculation will be described in detail below.

[Calculation of the inflow amount of a Exhaust gas component]

With reference to the 28 In the following, an exhaust-gas-component-inflow calculation routine for calculating the inflow amount of an exhaust gas component into the catalyst will be described below 27 flows. In this routine, first in the steps 721 and 722 the output signal FA / F of the upstream air-fuel ratio sensor 28 with the upper limit F (+) and the lower limit F (-) of the central value of FA / F. If F (-) ≦ FA / F ≦ F (+), that is, if FA / F can be regarded as the central value of FA / F, this routine is ended without further steps being executed.

If FA / F> F (+), that is, if the air-fuel ratio is on the upstream side of the catalyst 27 If, on the side of a lean component, it is outside the central value of FA / F, then the process goes to step 723 continued. A deviation ΔA / F (= FA / F-F (+)) between FA / F and the upper limit value F (+) is determined in step 723 calculated. In the next step 724 the amount of G becomes one in the catalyst 27 flowing exhaust gas based on the operating condition of the engine 11 calculated. This will be in step 725 a lean component amount FL of the catalyst 27 inflowing exhaust gas is calculated according to the following equation: ΔFL = ΔA / F × G × k (k: coefficient)

This will be in step 726 the lean component amount ΔFL of the exhaust gas is integrated to obtain the total lean component amount FL of the exhaust gas.

If FA / F <F (-), that is, if the air-fuel ratio is on the upstream side of the catalyst 27 is outside on the side of a fat component of the central value of FA / F, the process goes to step 727 continued. A deviation ΔA / F = (FA / F - F (-)) between FA / F and the lower limit value F (-) becomes in step 727 calculated. In the next step 728 becomes the amount G of the exhaust gas based on the operating state of the engine 11 calculated. This will be in step 729 a rich component amount ΔFR of the exhaust gas which enters the catalyst 27 inflow calculated according to the following equation: ΔFR = ΔA / F × G × k (k: coefficient)

This will be in step 730 the rich component amount .DELTA.FR of the exhaust gas is integrated to obtain a total amount of exhaust gas FR of the exhaust gas, which is in the catalyst 27 flows.

The 30A to 30C show the relationship among the total lean and rich component amounts FL and FR of the exhaust gas and the air-fuel ratio FA / F obtained by the upstream-side fuel-ratio sensor 28 was detected.

[Calculation of the outflow of a Exhaust gas component]

With reference to the 29 In the following, an exhaust-gas-outflow calculation routine for calculating the exhaust gas discharge amount of an exhaust gas component resulting from the catalyst will be described 27 flows. In this routine, first in the steps 731 and 732 the output value RA / F of the downstream air-fuel ratio sensor 29 with the upper limit R (+) and the lower limit R (-) of the central value of RA / F. If R (-) ≦ RA / F ≦ R (+), that is, if RA / F can be regarded as the central value of RA / F, this routine is terminated without further steps being taken.

If RA / F> R (+), that is, if the air-fuel ratio is on the downstream side of the catalyst 27 is outside of a lean component of the RA's central value, then the process moves to step 737 continued. A deviation ΔA / F = (RA / F - R (+)) between RA / F and the upper limit R (+) becomes in step 733 calculated. In the next step 734 becomes the amount G of the exhaust gas based on the operating state of the engine 11 calculated. This will be in step 735 a lean component amount ΔRL of the catalyst 27 outgoing exhaust gas calculated according to the following equation: ΔRL = ΔA / F × G × k (k: coefficient)

This will be in step 736 the lean component amount ΔRL of the exhaust gas is integrated to give a total lean component amount RL of the catalyst 27 to obtain effluent exhaust gas.

If RA / F <R (-), that is, if the air-fuel ratio is on the downstream side of the catalyst 27 is outside of a fat component from the center of RA / F, then the process goes to step 737 continued. A deviation ΔA / F (= RA / F - R (-)) between RA / F and the lower limit R (-) becomes in step 737 calculated. In the next step 738 becomes a quantity G of the exhaust gas based on the operating state of the engine 11 calculated. This will be in step 739 a rich component amount ΔRR of the catalyst 27 outgoing exhaust gas calculated according to the following equation: ΔRR = ΔA / F × G × k (k: coefficient)

This will be in step 740 the rich component quantity .DELTA.RR of the exhaust gas is integrated to give a total fat component amount RR of the catalyst 27 to obtain effluent exhaust gas.

The 30D to 30F show the relationship between the total lean and rich component amounts RL and RR of the exhaust gas and the air-fuel ratio RA / F flowing through the downstream air-fuel ratio sensor 29 has been detected.

(Sixth Embodiment)

From the first and fifth embodiments, it is known that for the downstream air-fuel ratio sensor 29 a linear A / F sensor is used which generates a linear air-fuel ratio signal according to the air-fuel ratio of the exhaust gas. However, according to a sixth embodiment, as shown in FIGS 31 to 53 is shown as a downstream air-fuel ratio sensor 29 an oxygen sensor whose output signal is inverted depending on the air-fuel ratio of the exhaust gas which is on the side of a rich or lean component. The output voltage of the oxygen sensor is linearized by an output signal linearizing means (operation according to step 859 is determined by the 39 described later) to be converted into an air-fuel ratio.

[Calculation of the inflow amount of a Exhaust gas component]

With reference to the 31 is an exhaust gas component inflow calculation routine for calculating the inflow amount of one into the catalyst 27 inflowing exhaust gas component described below. In this routine, first in step 800 from determining deterioration period setting routines (1) to (3), as will be described later, determines whether the operating state of the engine 11 is within one of the deterioration determination periods in which the deterioration of the catalyst 27 is determined. In this deterioration determination Period setting routine (1) to (3), it is determined whether the operating condition of the engine 11 is within one of the deterioration determination periods based on the value of a period display flag set by each of the routines.

If the operating condition of the engine 11 is in the deterioration determination period, the process goes to step 801 continued. The total inflow amount of an exhaust gas component in the deterioration determination period becomes according to the steps 801 to 810 calculated.

First, in the steps 801 and 802 the output signal FA / F of the upstream air-fuel ratio sensor 28 with the upper limit F (+) and the lower limit F (-) of the central value of FA / F. If F (-) ≦ FA / F ≦ F (+), that is, if FA / F can be regarded as the central value of FA / F, this routine is terminated without performing any further steps.

If FA / F> F (+), that is, if the air-fuel ratio is on the upstream side of the catalyst 27 is outside of a lean component of the central value of FA / F, the process goes to step 803 continued. A deviation ΔA / F = (FA / F-F (+)) between FA / F and the upper limit value F (+) becomes in step 803 calculated. In the next step 804 becomes the amount G of the exhaust gas based on the operating state of the engine 11 calculated. This will be in step 805 the molar value .DELTA.FL of a lean component of the exhaust gas, which in the catalyst 27 inflow calculated according to the following equation:
In the above equation, "k '" denotes a coefficient for converting the lean component amount ΔA / F · G into a molar value and "kl" denotes a lean component correction coefficient. The reason why this lean component correction factor kl is used is: The 32 shows the concentration of each component versus the air-fuel ratio A / F of the inflowing exhaust gas and the 33 shows the relationship between the concentration of a lean component (O2) and that of a fat component (CO + H2) versus the excess air ratio λair. In a graph according to the 33 HC is not contained in a fat component. However, as in the 32 HC is lower compared to CO and H2, this can be ignored. Like from the 32 and 33 is apparent, even if the air-fuel ratio A / F is outside of a lean or rich component by the same amount ΔA / F, the molar value of a lean component (O ₂ ) and that of a fat component (CO + H ₂ ) differently.

If the molar value ΔFL of the lean component of the exhaust gas is calculated based on the error ΔA / F of the air-fuel ratio A / F from a theoretical air-fuel ratio, the molar-value conversion coefficient k 'is corrected based on the molar concentration ratio of the lean and rich components. Such a correction is performed by the following two methods. In a first method, a lean component correction factor kl is compared with a deviation | Δλ | of the overhang air ratio λ air of 1 is elaborated with respect to a fat component as the reference (kR = 1) using one of 34 shown kL card. According to a second method, a fat component correction factor kR is compared with a deviation | Δλ | of the excess air ratio λ air of FIG. 1 with respect to a lean component as a reference (k L = 1) using an in 35 worked out kR card shown.

After applying an equation "ΔFL = ΔA / F × G × k 0" as already described above, then in step 806 this ΔFL is integrated to obtain a total molar value FL of a lean component of the exhaust gas which enters the catalyst 27 flows.

If FA / F <F (-), that is, if the air-fuel ratio FA / F is on the upstream side of the catalyst 27 if outside is on the side of a fat component from the central value of FA / F, then the process goes to step 807 continued. A deviation ΔA / F (= FA / F-F (-)) between FA / F and the lower limit value F (-) is determined in step 807 calculated. In the next step 808 becomes the amount G of the exhaust gas based on the operating state of the engine 11 calculated. This will be in step 809 the molar value ΔFR of a fat component quantity of the exhaust gas which enters the catalyst 27 flows in, calculated using the equation below:
In the above equation, "k '" denotes a molar value conversion coefficient and "kR" denotes a rich component correction coefficient. As described above, the fat component correction coefficient kR is calculated by using the kR map according to the 35 fixed when a lean component is set as the reference value (kL = 1). On the other hand, if a fat component is used as a reference, then the fat component correction coefficient kR is set to 1.

In the next step 810 will be the one in step 809 calculated molar value ΔFR integrated to a total molar value FR of the fat component quantity of the catalyst 27 to receive incoming exhaust gas.

The routine is performed by interrupting each combustion in each cylinder to calculate the total molar value FL of the lean component amount and the total molar value FR of the rich component amount during the deterioration determination period. When the deterioration determination period ends, the process proceeds from step 800 to step 811 continued. The total molar value FL of the lean component amount and the total molar value FR of the rich component amount are respectively set as the period total inflow amounts FLtotal and FRtotal of the exhaust gas component. Then the stored molar values ΔFL and ΔFR of the lean and rich component quantities of each exhaust gas are reset and this routine is ended.

Time cards according to the 49A to 49F show an example of the process for calculating FL and FR.

[Establishment of Deterioration Determination Periods]

A deterioration determination period during which the deterioration of the catalyst 27 is determined, is a period between the time at which the engine 11 is started until the time in which the warm-up of the catalyst 27 is terminated, that is, until the temperature of the catalyst 27 has increased to a temperature at which the catalyst 27 operates efficiently or is a period from the time the engine is in operation 11 has taken its operation in a steady state after complete heating until a predetermined time kdt has elapsed. For a method of determining whether the warm-up of the catalyst 27 is completed, there are two methods. One of them uses an elapsed time after the engine has been started, the other method uses an integrated value, after which the combustion energy of the engine 11 is integrated after the engine 11 has been started. For this reason, for a method of setting a deterioration determination period, the following three methods exist.

[Establishing a deterioration determination period (1)]

A deterioration determination period setting routine (1) according to 36 sets a deterioration determination period based on an elapsed time from when the engine starts 11 has been started. This routine is executed by interruption within each predetermined interval. In step 821 determines if the engine 11 has been started. If the engine 11 has not been started, the process goes to step 825 continued. In step 825 a period indication flag is set to zero, which means that the operating state of the engine 11 corresponds to a state after or before the deterioration determination period, and then this routine is ended.

If the engine 11 is started, the process proceeds from step 821 on step 822 continued, wherein an elapsed time T from the start time of the engine 11 is detected at. In the next step 823 It is determined whether the elapsed time T has reached a predetermined time T0, which is for warming up the catalyst 27 is required. If the elapsed time T does not reach the predetermined time value T0, the process goes to step 824 continued. In step 824 the period indication flag is set to "1", which means that the operating state of the engine 11 corresponds to a state for the deterioration determination period, and then this routine is ended.

If this is the elapsed time T from the start of the engine 11 reaches from the predetermined time T0, then the warm-up of the catalyst 27 considered completed. The process then proceeds to step 825 where the period indicator flag is set to "1".

By the above-described operation, a period becomes from the time of starting the engine 11 is set to a predetermined elapsed time T0 as a deterioration determination period. The period display flag is kept "1" during this deterioration determination period.

[Setting a deterioration determination period (2)]

A deterioration determination period setting routine (2) according to 37 sets a deterioration determination period based on the integrated value of combustion energy of Motors 11 from the start time of the engine 11 from firm. The routine is by interrupting each combustion of each cylinder of the engine 11 carried out. In step 831 determines if the engine 11 has been started. If the engine 11 has not been started, the process goes to step 835 where a period indication flag is set to "0" and this routine is ended.

If the engine 11 has started, the process proceeds from step 831 on step 832 where the combustion energy E is integrated to an integrated value ΣE since the starting time of the engine 11 to receive. The combustion energy E is calculated using a map (not shown) in which at least one of the following values, namely, the air-fuel ratio, the engine speed, the pressure in an intake passage, and an intake air amount is used as a parameter. In the next step 833 is determined whether the integrated value ΣE of the combustion energy E reaches a predetermined value E0, which is for warming up the catalyst 27 is required. If the integrated value ΣE does not reach the predetermined value E0, the process goes to step 834 continued. In step 834 the period display flag is set to "1" and this routine is ended.

Then, if the integrated value ΣE of the combustion energy E reaches the predetermined value E0, then the warm-up operation of the catalyst becomes 27 considered completed. As a result, the routine proceeds to step 835 where the period display flag is set to "0" and this routine is ended.

By the above-described operation, a period becomes from the time of starting the engine 11 until the integrated value ΣE of the combustion energy reaches the predetermined value EO, set as a deterioration determination period. The period display flag is held at "1" during this deterioration determination period.

[Establishing a deterioration determination period (3)]

A deterioration determination period detection routine (3) according to 38 sets a period from the time the engine is running 11 is operated in a stable state after being warmed up to a predetermined elapsed time kdt, as a deterioration determination period. This routine is performed by interrupting each predetermined time interval. In step 841 determines if the engine 11 is operated in a steady state based on whether the engine speed, the pressure in an intake passage, the intake air amount, or other values are stable. If the engine 11 is operated in a steady state, the process goes to step 842 continued. In step 842 is determined based on an engine cooling temperature, whether the engine 11 completely warmed up. If the engine 11 is not operated in a steady state or has not fully heated, the process goes to step 846 continued. In step 846 the period display flag is set to zero and the routine is ended.

If the engine 11 is operated in a steady state and has fully warmed up, the deterioration determination period is started. The process goes to step 843 , wherein an elapsed time Δt is counted from the start of the deterioration determination period. In the next step 844 It is determined whether the elapsed time Δt is larger than a predetermined time kdt. If the elapsed time Δt is not larger than the predetermined time kdt, the process goes to step 845 continued. In step 845 the period display flag is set to "1" and the routine is ended. Then, when the elapsed time Δt is larger than the predetermined time kdt, the process proceeds 846 continued. In step 846 the period display flag is set to zero and the routine is ended.

By the above-described operation, a period from the time when the engine becomes 11 it starts to operate in a steady state after its complete heating until the time when a predetermined time kdt has elapsed, set as a deterioration determination period. During this period, the period display flag is held at "1".

It may be sufficient one of the above-described deterioration determination period setting routines (1) to use (3), respectively in the

36 to 38 are shown. However, the routines (1) and (3) or the routines (2) and (3) can also be shared.

[Calculation of the outflow of a Exhaust gas component]

With reference to the 39 In the following, an exhaust-gas-outflow calculation routine for calculating the total outflow quantity of exhaust gas components consisting of a catalyst will be described 27 during the deterioration determination period. In this routine, first in step 858 the output voltage of the downstream oxygen sensor 29 detected. In the next step 859 becomes the output voltage of the downstream oxygen sensor 29 in an air-fuel ratio RA / F according to a preset conversion table (see 40 ) converted. This will be in step 860 determines if the operating condition of the engine 11 corresponds to a state during a deterioration determination period. If a positive determination in step 860 is received, the process goes to step 861 , wherein the outflow of an exhaust gas component through the process of step 861 to 870 is calculated.

In the steps 861 and 862 becomes the output value RA / F of the downstream air-fuel ratio sensor 29 with the upper limit R (+) and the lower limit R (-) of the central value of RA / F. If R (-) ≦ RA / F ≦ R (+), that is, if the output RA / F can be considered the central value of RA / F, this routine is terminated without performing any further steps.

If RA / F> R (+), that is, if the air-fuel ratio RA / F on the downstream side of the catalyst 27 is outside of a lean component of the RA's central value, then the process moves to step 863 continued. A deviation ΔA / F = (RA / F - R (+)) between RA / F and the upper limit R (+) becomes in step 863 calculated. In the next step 864 The amount G of the exhaust gas is calculated based on the operating state of the engine 11 , This will be in step 865 the molar value ΔRL of a lean component of the exhaust gas passing through the catalyst 27 flows, calculated according to the following equation: ΔRL = ΔA / F × G × k0 k0 = k '× kL

In the above equation, "k '" denotes a molar value conversion coefficient and "kR" denotes a lean component correction coefficient (see FIG 34 ). In the next step 866 is the molar value .DELTA.L of the lean component of the exhaust gas integrated to obtain a total molar value RL of the lean component of the exhaust gas, which from the catalyst 27 flows.

If RA / F <R (-), that is, if the air-fuel ratio RA / F on the downstream side of the catalyst 27 If outside is on the side of a fat component from the central value of RA / F, then the process goes to step 867 continued. A deviation ΔA / F (= RA / F - R (-)) between RA / F and the lower limit R (-) becomes in step 867 calculated. In the next step 868 becomes the amount G of exhaust gas based on the operating state of the engine 11 calculated. This will be in step 869 the molar value ΔRR of a fat component amount of the exhaust gas which enters the catalyst 27 out, calculated using the equation below: ΔRR = ΔA / F × G × k1 k0 = k '× kR

In the above equation, "k '" denotes a molar value conversion coefficient and "kL" denotes a rich component correction coefficient (see FIG 35 ). In the next step 870 the molar value ΔRR of the rich component amount of the exhaust gas is integrated to obtain a total molar value RR of the rich component amount of the exhaust gas resulting from the catalyst 27 flows.

The routine is performed by interrupting each combustion in each cylinder to calculate the total molar value RL of the lean component amount and the total molar value RR of the rich component amount during the deterioration determination period. When the deterioration determination period ends, the process proceeds from step 860 to step 871 continued. The total molar value RL of the lean component amount and the total molar value RR of the rich component amount are respectively set as the period total discharge amounts RLtotal and RRtotal of the exhaust gas component. Then, the stored molar values ΔRL and ΔRR of the lean and rich component amounts of each exhaust gas are reset and this routine is ended.

Time cards according to the 49A to 49F show an example of the above-described process for calculating RL and RR.

[Correction of the temperature an element of a downstream Oxygen sensor]

A temperature correction routine with respect to a downstream oxygen sensor element according to the 41 estimates the temperature of an element 51 a downstream oxygen sensor 29 and corrects the linearity characteristic of a conversion panel for converting the output voltage of the downstream oxygen sensor 29 in an air-fuel ratio RA / F corresponding to the estimated temperature of the element 51 , The reason for this is that the output characteristic of the downstream oxygen sensor 29 according to the temperature of the element 51 varied.

In this routine will be in step 881 First, a saturation temperature table having an engine speed Ne and a pressure Pm in an intake pipe taken as a parameter and the saturation temperature of the downstream oxygen sensor 28 which corresponds to the heat quantity of the exhaust gas, which is essentially determined by the parameters Ne and Pm. This will be in step 882 the saturation temperature is processed with a first order delay and thereby the temperature of the element 51 the downstream oxygen sensor 29 estimated. The reason for this is because the change in the temperature of the element 51 the downstream oxygen sensor 29 caused by the change (the change of Ne and Pm) of the heat quantity of the exhaust gas can be approximated with a first degree delay. In the next step 883 becomes the linearity characteristic of the lean and rich components in the conversion panel for converting the output voltage of the downstream oxygen sensor 29 to the air-fuel ratio Ra / F according to the temperature of the element 51 the downstream oxygen sensor 29 corrected (see 42 ).

[Correction when the fuel supply is interrupted]

A correction routine that works in 43 is shown, corrects the linearity characteristic of a lean component in a conversion board for converting the output voltage of a downstream oxygen sensor 29 in a fuel ratio RA / F when the fuel supply is interrupted. In this routine, first in step 891 determines if the fuel is interrupted (F / C). If the fuel is interrupted, the process goes to step 892 wherein it is determined whether an elapsed time since the beginning of the fuel cut exceeds a predetermined time value t10. If the fuel is not interrupted or if the elapsed time does not reach the predetermined time t10, this routine is terminated without performing any further steps.

If the elapsed time reaches the predetermined time t10, the process goes to step 893 wherein the output voltage of the downstream oxygen sensor 29 is read. This will be in step 894 corrects the linearity characteristic of a lean component in the conversion table (see 44 ).

[Determination of air-fuel ratio on the upstream Page]

A determination routine regarding an upstream air-fuel ratio condition, as in FIG 45 which determines which of the rich, lean or stoichiometric air-fuel ratios FA / F on the upstream side of the catalyst 27 true. This determination is made using the upper limit F (+) and the lower limit F (-), which are determined in step 120 according to the 7 be calculated.

In this routine, first in the steps 901 and 902 determines whether the air-fuel ratio FA / F on the upstream side is equal to or greater than the upper limit value F (+), or equal to or lower than the lower limit value F (-). If FA / F ≥ F (+), then the air-fuel ratio FA / F is determined to be lean. In this case, the process goes to step 903 with an upstream air-fuel ratio flag set to "1". If FA / F ≦ F (-), then the air-fuel ratio FA / F is determined to be rich. In this case, the process goes to step 904 with the upstream air-fuel ratio flag set to "-1". If F (+)> FA / F> F (-), then the air-fuel ratio is determined to be stoichiometric. In this case, the process goes to step 904 with the upstream air-fuel ratio flag set to "0".

The season tickets according to the 49A to 49F show an example of the above Shifting sequence with respect to the upstream air-fuel ratio flag.

[Determination of the air-fuel ratio on the downstream Page]

A determination routine regarding a downstream air-fuel ratio state according to 46 determines whether on the downstream side of the catalyst 27 a rich, lean or stoichiometric air-fuel ratio RA / F prevails. This determination is made using the upper limit R (+) and the lower limit R (-) which are determined in step 116 according to the 7 have been calculated.

In this routine, first in the steps 911 and 912 determines whether the air-fuel ratio RA / F on the downstream side is equal to or greater than the upper limit value R (+) or equal to or smaller than the lower limit value R (-). If RA / F ≥ R (+), then the air-fuel ratio RA / F is determined to be lean. In this case, the process goes to step 913 with a downstream air-fuel ratio flag set to "1". If RA / F ≤ R (-), then the air-fuel ratio RA / F is determined to be rich. In this case, the process goes to step 914 with the downstream air-fuel ratio flag set to "-1". If R (+)> RA / F> R (-), then the air-fuel ratio RA / F is determined to be stoichiometric. In this case, the process goes to step 914 with the downstream air-fuel ratio flag set to "0".

Time cards, as in the 49A - 49F 11 show an example of the above-described switching timing with respect to the downstream air-fuel ratio flag.

[Erasable calculation of the inflow of a Exhaust gas component]

An erasable calculation routine relating to an exhaust gas component inflow according to 47 and 48 adds an erase function to the exhaust component injection calculation routine in accordance with FIG 31 added. If, for this reason, an operation according to the 47 and 48 is chosen, then the process as he is in 31 shown is no longer required.

The in the 47 and 48 shown routines are performed by interrupting operations with respect to each combustion of each cylinder. When a lean component through the upstream air-fuel ratio sensor 28 during the deterioration determination period, or when a predetermined time Δtk has not elapsed since the upstream air-fuel ratio sensor 28 has stopped detecting a lean component, and if a lean component is not exhausted by the downstream air-fuel ratio sensor 29 has been detected, this means that the catalyst 27 is not saturated. For this reason, a lean component quantity of the exhaust gas, which is based on the output value of the upstream air-fuel ratio sensor 28 has been calculated (a clear flag is set to "1"), so that the lean component quantity no longer has an effect of detecting the deterioration of a catalyst.

In detail, in step 920 First determines if the operating condition of the engine 11 corresponds to a state within the deterioration determination period. If a negative determination in step 920 is done, this routine is terminated without performing any further steps. If a positive determination in step 920 is performed, the process goes to step 921 continued. In step 921 It is determined whether an elapsed time dt has elapsed a predetermined time Δtk since the air-fuel ratio on the upstream side has shifted from a lean region to a stoichiometric region. If the elapsed time dt does not reach the predetermined time Δtk, the process goes to step 922 continued. In step 922 It is determined whether a count flag has been set to "1" with respect to an elapsed time (dt). In the present case, the dt-count flag is kept at "1" since the time from when counting the elapsed time dt has been started until the elapsed time dt reaches the predetermined time Δtk.

When in step 922 it is determined that the dt count flag is "1", the process goes to step 923 continued. In step 923 the elapsed time dt is measured. In the next step 924 It is determined whether the downstream air-fuel ratio flag is "1", which means that the air-fuel ratio RA / F on the downstream side is lean. If the downstream air-fuel ratio flag is "1", the process goes to step 925 continued. The clear flag is set to "0", which means that the lean component quantity FL of the exhaust gas is not cleared. If the downstream air-fuel ratio flag is not "1" in step 924 That's how the step will be 925 skipped and the process moves to step 926 continued. Of the initial value of the deletion flag is set to "1", which means that the lean component quantity FL of the exhaust gas is deleted.

If the dt count flag is not at "1" in step 922 is set, the process also proceeds to step 926 continued. In step 926 It is determined whether the upstream air-fuel ratio flag is set to "1", which means that the air-fuel ratio FA / F on the upstream side is lean. If the upstream air-fuel ratio flag is "1", the process goes to step 927 continued. In step 927 becomes a molar value ΔFL of a lean component amount of the catalyst 27 inflowing exhaust gas is calculated according to the following equation: ΔFL = ΔA / F · G · k0 ΔA / F = FA / F - F (+) k0 - k '· kl

at In the above equation, G denotes the amount of exhaust gas called "k" a molar value conversion coefficient and "kl" denotes one Lean component correction coefficients.

In the next step 928 It is determined whether the downstream air-fuel ratio flag is set to "1", which means that the air-fuel ratio RA / F on the upstream side is lean. If the downstream air-fuel ratio flag is "1", the process goes to step 929 continued. In step 929 the clear flag is set to "0" so as not to cause extinction of the lean component quantity FL of the exhaust gas. However, if the downstream air-fuel ratio flag is not set to "1", this routine will be executed without executing the step 929 completed.

If in the above step 926 the upstream air-fuel ratio flag is not set to "1", the process goes to step 930 continued. In step 930 It is determined whether the last upstream air-fuel ratio flag is set to "1", that is, whether the air-fuel ratio on the upstream side was lean in the last operation. In step 930 "YES" is selected immediately after the upstream air-fuel ratio sensor 28 has completed the acquisition process of a lean component. In this case, in step 931 the molar value ΔFL of the lean component amount is integrated while the upstream air-fuel ratio sensor detects a lean component to a lean component quantity FL of the catalyst 27 inflowing exhaust gas, whereby ΣΔFL is deleted. This will be in step 932 the dt count flag is cleared to "1" and this routine ends. If the last upstream air-fuel ratio flag in step 930 is not set to "1", that is, if the air-fuel ratio on the upstream side was not lean in the last operation, this routine is terminated without the steps 931 and 932 be executed.

Then, when the elapsed time dt has reached the predetermined time Δtk from the time when the air-fuel ratio on the upstream side has been shifted from a lean region to a stoichiometric region, then 921 "YES" selected. The process goes to step 933 according to the 48 and it is determined whether the deletion flag is "1", which means that the lean component quantity FL of the exhaust gas should be deleted. If the deletion flag is "1", the process goes to step 934 proceeding, wherein the lean component quantity FL of the exhaust gas is deleted. If the deletion flag is "0", meaning that the lean component quantity FL of the exhaust gas does not need to be cleared, the process goes to step 935 without execution of the step 934 continued.

In step 935 the current lean component amount FL of the exhaust gas is added to a last total flow amount FLtotal of a lean component in the purge determination period to renew the total inflow amount FLtotal of the lean component.

After the deletion flag in step 936 is set to "1", a time dt in step 937 deleted. In the next step 938 the dt count flag is set to "0" and this routine is ended.

The deletion process described above, as described in US Pat 47 and 48 is related to the lean component amount FL of the exhaust gas. However, the same method may be used for the rich component amount FR of the exhaust gas. Time cards according to the 50A to 50F show an example of the deletion process.

[Frequency count of the saturation by a fat / lean component]

A frequency count routine relating to rich / lean component saturation in accordance with 51 counts a frequency in which the catalyst 27 is saturated by a lean component and a frequency in which it is saturated by a fat component during a deterioration determination period.

In this routine, first in step 941 determines if the operating condition of the engine 11 corresponds to a state in the deterioration determination period based on the period display flag. If the period display flag is "1", which means the deterioration determination period, the process goes to step 942 continued. In step 942 is determined (1) whether the downstream air-fuel ratio flag is set to "1" and (2) whether the current value of the downstream air-fuel ratio flag is different from the latter value. If the above conditions (1) and (2) are reached simultaneously, then the catalyst becomes 27 classified as saturated by a lean component. The process goes to step 943 with a count for lean component saturation being incremented by one. However, if the catalyst 27 is not saturated by a lean component, the process moves to step 944 without incrementing the lean count.

In step 944 it is determined (1) whether the downstream air-fuel ratio flag is set to "-1" and (2) whether the current value of the downstream air-fuel ratio flag is different from the latter value. If the above-mentioned conditions (1) and (2) are simultaneously achieved, the catalyst becomes 27 classified as saturated by a fat component. The process then proceeds to step 945 , wherein a fat count for a fat component saturation is incremented by one. However, if the catalyst 27 is not saturated by a fat component, the routine is terminated without incrementing the fat count.

Then, when the deterioration determination period ends, the process of step S10 941 on the steps 946 and 947 continued. The value of the lean count value as well as the value of the rich count value are set to nL and nR, respectively, to be used in a catalyst deterioration determination routine, which is determined from FIG 52 and 53 will be described below.

[Determination of deterioration of the catalyst 27 on the side of a lean component]

A deterioration determination routine relating to a lean component side catalyst according to the 52 is executed immediately after a deterioration determination period ends (step 951 ). Immediately after the deterioration determination period ends, the process goes to step 952 continued. In step 952 For example, the average value MAXOSIav of the maximum amount in which a lean component is absorbed in a deterioration determination period is calculated according to the following equation, where the total inflow amount FLtotal is an exhaust lean component, the total exhaust amount Rltotal of an exhaust lean component, and a frequency nL in which a lean component catalyst during the deterioration determination period becomes saturated, using: MAXOSIav = (Fltotal - RLtotal) / nL

In the next step 953 becomes the current average value MAXOSIav of the maximum amount in which a lean component is absorbed by the average value MAXIOSIOav of the maximum amount of a lean component which is absorbed when the catalyst 27 is new, to obtain a catalyst deterioration degree determination value LDETERIO. In the next step 954 the catalyst deterioration degree determination value LDETERIO is compared with a predetermined deterioration determination value KL. If LDETERIO <kL then the catalyst 27 in step 955 classified as deteriorated. On the other hand, if LDETERIO ≥ kL, this routine will be without classification of the catalyst 27 finished as deteriorated.

[Detection of the deterioration of a catalyst 27 on the side of a fat component]

A deterioration determination routine relating to a grease component side catalyst according to the 53 is executed immediately after the deterioration determination period ends (step 961 ). Immediately after the deterioration determination period ends, the process goes to step 962 continued. In step 962 the average value MINOSIav of the maximum amount in which a fat component is absorbed in the deterioration determination period is calculated according to the following equation in which the total ice flow amount FRtotal of an exhaust gas rich component, the total outflow amount RR totally an exhaust gas component and a frequency nR, in which the catalyst 27 is saturated with a fat component in the deterioration determination period, using: MINOSIav = (FRtotal - RRtotal) / nR

In the next step 963 For example, the present average MINOSIav of the maximum amount in which a fat component is absorbed by the average value MINOSIOav becomes the maximum amount of a fat component absorbed when the catalyst 27 is new, to obtain a catalyst deterioration degree determination value RDETERIO. In the next step 965 the catalyst deterioration degree determination value RDETERIO is compared with a predetermined deterioration determination value kR. If RDETERIO <kR, then the catalyst becomes 27 in step 965 classified as deteriorated. On the other hand, if RDETERIO ≥ kR, this routine will be without classification of the catalyst 27 finished as deteriorated.

Only one of the catalyst deterioration determination routines according to 52 and 53 can be further executed. If both routines according to the 52 and 53 is a method for determining the deterioration of the catalyst 27 when the deterioration of the catalyst 27 is detected by one of the routines as well as another method for determining the deterioration of the catalyst 27 when the deterioration of the catalyst 27 is detected simultaneously by both routines, conceivable.

(Further embodiment)

[Supply of an injection jitter signal]

In a fuel cell signal supply routine according to the 54 will step in first 971 determines if the operating condition of the engine 11 corresponds to a state in a deterioration determination period. If it corresponds to a state in the deterioration determination period, the process goes to step 972 wherein a fuel cell signal supply in which the fuel injection amount is varied so that the air-fuel ratio of the exhaust gas alternately fluctuates between the sides of a rich component and a lean component by a predetermined value is performed. If it does not correspond to the state in the deterioration determination period, no fuel cell signal supply is executed. The average values MAXOSIav and MINOSIOav of the maximum amount of exhaust gas component during the deterioration determination period by the catalyst 27 can be accurately calculated by performing the fuel cell signal supply during the deterioration determination period.

Claims (9)

An abnormality detecting device for an air-fuel ratio control system in the upstream and downstream air-fuel ratio sensors ( 28 and 29 ) on an upstream side and a downstream side of a catalyst ( 27 ) installed in an exhaust pipe ( 26 ) of an engine ( 11 ) is arranged to purify an exhaust gas, with a device (step 119b ) for calculating a central value of the air-fuel ratio on the upstream side of the catalyst based on an output signal of the upstream air-fuel ratio sensor, a device (step 121 ) for calculating an error between the central value of the air-fuel ratio on the upstream side of the catalyst and the central value of the air-fuel ratio on the downstream side of the catalyst, and a sensor abnormality determination device (steps 122 and 123 ) for determining an abnormality of the upstream air-fuel ratio sensor or the downstream air-fuel ratio sensor based on whether the failure is in a predetermined range. An abnormality determination device for an air-fuel ratio control system according to claim 1, characterized by an upstream sensor abnormality determination device (steps 201 to 218 ) for determining an abnormality of the upstream air-fuel ratio sensor. An abnormality detecting device for an air-fuel ratio control system according to claim 1, characterized in that the upstream sensor abnormality determination device (step 201 to 218 ) determines the abnormality of the upstream air-fuel ratio sensor based on a Change in the air-fuel ratio detected by the upstream air-fuel ratio sensor or a control parameter that also changes according to the change in the air-fuel ratio. An abnormality detecting device for an air-fuel ratio control system according to claim 3, characterized by a basic injection amount calculating device (step 101 ) for calculating a basic injection amount (Tp) corresponding to an operating state of the engine, an air-fuel ratio correction value setting device (steps 103 and 104 for setting an air-fuel ratio correction value (FAF) according to a deviation between an air-fuel ratio detected by the upstream air-fuel ratio sensor and a target air-fuel ratio and an injection control device (step 105 ) for controlling the fuel injection amount delivered to the engine based on the basic injection amount and the air-fuel ratio correction value, wherein the upstream sensor abnormality determination device (steps 201 to 218 ) determines the abnormality of the upstream air-fuel ratio sensor based on a change in the air-fuel ratio detected by the upstream air-fuel ratio sensor when the basic injection amount suddenly changes. An abnormality detecting device for an air-fuel ratio control system according to claim 3, characterized by a basic injection amount calculating device (step 101 ) for calculating a basic injection amount according to an operating state of the engine, an air-fuel ratio correction value setting device (steps 103 and 104 for setting an air-fuel ratio correction value corresponding to a deviation between an air-fuel ratio detected by the upstream air-fuel ratio sensor and a target air-fuel ratio, and an injection control device (step 105 ) for controlling the fuel injection amount supplied to the engine based on the basic injection amount and the air-fuel ratio correction value, wherein the upstream sensor abnormality determination device (step 201 to 218 ) determines the abnormality of the upstream air-fuel ratio sensor based on a change of an air-fuel ratio correction value set by the air-fuel ratio correction value setting device if the basic injection amount suddenly varies. An abnormality detecting device for an air-fuel ratio control system according to claim 4, characterized by a target air-fuel ratio adjusting device (step 103 ) for setting a target air-fuel ratio according to an operating state of the engine, wherein the upstream sensor abnormality determination device (step 215 ) for determining an abnormality of the upstream air-fuel ratio sensor by comparing a change in the air-fuel ratio correction value set by the air-fuel ratio correction value setting device with a change in the target air-fuel ratio set by the target air-force ratio adjusting device if the target air-fuel ratio suddenly changes. An abnormality detecting device for an air-fuel ratio control system according to claim 4, characterized by an injection quantity correcting device (step 421 ) for increasing or decreasing an injection amount according to an operating condition of the engine and a total correction amount calculating means (steps 423 and 424 ) for calculating a total correction amount for the basic injection amount calculated by the basic injection amount calculating device when the injection amount is corrected by the injection amount correcting device, wherein the upstream sensor abnormality detecting device (step 409 ) determines the abnormality of the upstream air-fuel ratio sensor by comparing the total correction amount calculated by the total correction amount calculating device and a change in the air-fuel ratio detected by the upstream air-fuel ratio sensor when the injection amount is increased or decreased by the injection amount correcting device. An abnormality detecting device for an air-fuel ratio correcting system according to claim 4, characterized by an amplitude detecting device (steps 504 to 507 ) for detecting a fluctuation amplitude of the Air fuel ratio, which has been detected by the upstream air-fuel ratio sensor, wherein the abnormality determination device for the upstream sensor (step 510 ) for determining the abnormality of the upstream air-fuel ratio sensor based on the fluctuation amplitude of the air-fuel ratio detected by the amplitude-detecting device when an operating state of the engine is in a transient state. An abnormality detecting device for an air-fuel ratio control system according to claim 2, characterized in that a determining device (steps 221 to 224 ) for determining which of the sensors, namely, the upstream air-fuel ratio sensor and the downstream air-fuel ratio sensor is abnormal, based on a result obtained by the sensor abnormality determination device and a result obtained by the upstream sensor abnormality determination device.