Knocking Control System for Internal Combustion Engines

This invention relates to a knocking control system for internal combustion engines, and particularly to a system of this kind for an internal combustion engine equipped with a device which varies the opening and closing timing or lift of inlet valves and exhaust valves of the engine depending on operating conditions of the engine in order to prevent knocking.

An internal combustion engine is conventionally known, e.g. from Japanese Patent Publication (Kokoku) No. 49-33289, which is capable of changing the valve timing (valve opening period or valve lift) of inlet valves and/or exhaust valves between low speed valve timing suitable for a lower engine rotational speed region and high speed valve timing suitable for a higher engine rotational speed region, in order to enhance the charging efficiency or combustion efficiency.

Further, a control method for internal combustion engines is known, e.g. from Japanese Patent Publication (Kokoku) 57-30980, in which knocking occurring in the engine is detected, and the ignition timing is retarded upon detection of knocking to thereby eliminate the knocking.

However, if knocking control is effected in an internal combustion engine equipped with the valve timing device, there occur the following problems:

Knocking is caused by abnormal combustion within engine cylinders. The magnitude (knocking noise level) and frequency of knocking, as well as the crank angle at which knocking occurs vary if the valve timing and/or lift of the inlet and exhaust valves and/or the compression ratio vary. Further, the noise level (background level) of the output of a knocking sensor, on the basis of which a knocking discrimination level is set, varies at the time of changeover of the valve timing when changeover of rocker arms and oil passages, ete. is effected. Also, the ignition timing at which knocking occurs and the optimal ignition timing (MBT) are different between the low speed valve timing and the high speed valve timing. Therefore, knocking is not eliminated to thereby cause damage to the engine and degraded driveability and hence marketability of same, unless the retarding amount, retarding speed, advancing amount, and advancing speed of the ignition timing are set to respective appropriate values depending on the actual valve timing when the ignition timing is retarded upon detection of knocking to eliminate same and when it is advanced after elimination of the knocking.

In order to solve the above problems, if the values of discrimination parameters used by knocking discriminating means or the values of control parameters used by knocking control means are varied in accordance with changeover of the valve timing responsive to engine operating conditions, knocking control can be properly carried out irrespective of the actual valve timing. However, if a failure occurs in the valve timing control means in such a case, the values of the discrimination parameters or the control parameter cannot be varied correctly in accordance with changeover of the value timing, resulting in occurrence of knocking.

It is the object of the invention to provide a knocking control system for an internal combustion engine, which is capable of properly effecting knockign control even if a failure has occurred in valve timing control means.

To attain the above object, the present invention provides a knocking control system for an internal combustion engine having inlet valves and exhaust valves, at least one of the inlet valves and the exhaust valves having valve timing thereof controlled by valve timing control means depending on operating conditions of the engine.

According to a first aspect of the invention, the knocking control system is characterized by comprising:

valve timing detecting means for detecting the valve timing controlled by the valve timing control means;

knocking parameter detecting means for detecting a detection parameter indicative of knocking occurring in the engine;

knocking discriminating means for determining whether or not knocking has occurred in the engine on the basis of the detection parameter detected by the knocking parameter means and at least one discrimination parameter;

knocking control means responsive to an output from the knocking discriminating means indicative of a determination result that knocking has occurred, for controlling an operation of the engine by the use of at least one control parameter, so as to eliminate knocking;

parameter value selecting means for selecting a value of at least one of the at least one discrimination parameter and at at least one control parameter, which corresponds to the valve timing detected by the valve timing detecting means;

abnormality detecting means for detecting abnormality in the valve timing control means; and

failsafe means for holding the valve timing at predetermined valve timing irrespective of an operating condition in which the engine is operating, and selecting a value of at least one of the at least one discrimination parameter and the at least one control parameter, which correspond to the held predetermined valve timing, when abnormality in the valve timing control means is detected by the abnormality detecting means.

According to a second aspect of the invention, the knocking control system is characterized by comprising:

knocking discrimination means for determining whether or not knocking has occurred in the engine;

knocking control means responsive to the valve timing controlled by the valve timing control means and an output from the knocking discriminating means, for controlling an operation of the engine so as to eliminate knocking;

abnormality detecting means for detecting abnormality in the valve timing control means; and

failsafe means for inhibiting operation of the knocking control means when abnormality in the valve timing control means is detected by the abnormality detecting means.

The above and othe objects, features, and advantages of the invention will become more apparent from the ensuing detailed description taken in conjunction with the accompanying drawings.

FIG. 1 is a schematic diagram showing the whole arrangement of a fuel supply control system for an internal combustion engine incorporating a knocking control system according to the invention;

FIG. 2 is a longitudinal cross-sectional view of essential parts of the engine;

FIG. 3 is a transverse cross-sectional view of a connection-changeover mechanism;

FIGS. 4, 4A and 4B are schematic diagrams showing an oil-feeding system and an oil pressure-changeover device;

FIG. 5 is a diagram showing set basic fuel injection period characteristics for low speed valve timing and high speed valve timing;

FIG. 6 is an enlarged view of a portion encircled in FIG. 5;

FIG. 7 is a diagram showing a T.sub.VT table;

FIG. 8 is a diagram showing lower speed and higher speed valve timing regions;

FIGS. 9, 9A, 9B, and 9C are flowcharts of a program for controlling the changeover of the valve timing;

FIG. 10 is flowchart of a subroutine for obtaining basic fuel injection period values Ti.sub.L and Ti.sub.H from respective Ti.sub.L and Ti.sub.H maps;

FIG. 11 is a flowchart of a subroutine for calculating a value T.sub.VT ;

FIG. 12 is a flowchart of a program for controlling ignition timing whch is executed in the knocking control system of the invention;

(A) and (B) of FIG. 13 are block diagrams illustrating an arrangement for determining occurrence of knocking;

FIG. 14 is a flowchart of a program for controlling knocking according to a first aspect of the invention, which is executed in the knocking control system of the invention;

FIG. 14A is a flowchart of a program for controlling knocking according to a second aspect of the invention;

FIG. 15 is a flowchart of a subroutine corresponding to a step 1406 of the program of FIG. 14;

FIG. 16 is a flowchart of a subroutine corresponding to a step 1407 of the program of FIG. 14;

FIG. 17A is a graph showing a multiplier term and FIG. 17B is a graph showing an addend term to be used for calculation of a knocking discrimination level executed at a step 1404 in FIG. 14;

FIG. 18 is a graph showing the knocking discrimination level calculated at the step 1404 in FIG. 14;

FIG. 19 is a graph showing output from a knocking sensor appearing in FIG. 1;

FIG. 20 is a flowchart of a subroutine for fail-safe operation to be carried out at a step 1409 in FIG. 14;

FIG. 21 is a flowchart of a subroutine for fail-safe operation to be carried out at a step 1409' in FIG. 14A; and

FIG. 22 is a graph showing a predetermined value .theta..sub.RETF/S of a correction variable .theta.ig.sub.KN appearing in FIG. 21.

The knocking control system according to the invention will now be described in detail with reference to the drawings showing an embodiment thereof.

Referring first to FIG. 1, there is shown the whole arrangement of a fuel supply control system for an internal combustion engine incorporating the knocking control system according to the invention. In the figure, reference numeral 1 designates an internal combustion engine of DOHC in-line 4 cylinder type for automotive vehicles, in which two pairs of inlet and exhaust valves are provided for each cylinder. Connected to the cylinder block of the engine 1 is an intake pipe 2 across which is arranged a throttle body 3 accommodating a throttle valve 3' therein. A throttle valve opening (.theta..sub.TH) sensor 4 is connected to the throttle valve 3' for generating an electric signal indicative of the sensed throttle valve opening and supplying same to an electronic control unit (hereinafter called "the ECU") 5.

Fuel injection valves 6, only one of which is shown, are inserted into the interior of the intake pipe at locations intermediate between the cylinder block of the engine 1 and the throttle valve 3' and slightly upstream of respective intake valves, not shown. The fuel injection valves 6 are connected to a fuel pump, not shown, and electrically connected to the ECU 5 to have their valve opening periods controlled by signals therefrom.

Ignition plugs 22 provided for respective cylinders of the engine 1 are connected via a driving circuit 21 to the ECU 5 which controls the ignition timing .theta.ig of the ignition plugs 22.

Further, an electromagnetic valve 23 for valve timing changeover control, described hereinafter, is connected to the output of the ECU 5 which controls opening and closing of the electromagnetic valve 23.

On the other hand, an intake pipe absolute pressure (P.sub.BA) sensor 8 is provided in communication with the interior of the intake pipe 2 at a location immediately downstream of the throttle valve 3' for supplying an electric signal indicative of the sensed absolute pressure within the intake pipe 2 to the ECU 5. An intake air temperature (T.sub.A) sensor 9 is inserted into the intake pipe 2 at a location downstream of the intake pipe absolute pressure sensor 8 for supplying an electric signal indicative of the sensed intake air temperature T.sub.A to the ECU 5.

An engine coolant temperature (T.sub.W) sensor 10, which may be formed of a thermistor or the like, is mounted in the cylinder block of the engine 1, for supplying an electric signal indicative of the sensed engine coolant temperature T.sub.W to the ECU 5. An engine rotational speed (Ne) sensor 11 and a cylinder-discriminating (CYL) sensor 12 are arranged in facing relation to a camshaft or a crankshaft, not shown, of the engine 1. The engine rotational speed sensor 11 generates a pulse as a TDC signal pulse at each of predetermined crank angles whenever the crankshaft rotates through 180 degrees, while the cylinder-discriminating sensor 12 generates a pulse at a predetermined crank angle of a particular cylinder of the engine, both of the pulses being supplied to the ECU 5. A knocking sensor 24 is mounted in a peripheral wall of the cylinder block of the engine 1 at a location in the vicinity of the top-dead-center of an engine cylinder, for sensing vibration of the engine and supplying an electric signal indicative of the sensed vibration to the ECU 5. The knocking sensor 24 is adapted to resonate with the frequency of knocking taking place within the cylinder. Such knocking sensors may be provided for respective cylinder to detect knocking more accurately.

A three-way catalyst 14 is arranged within an exhaust pipe 13 connected to the cylinder block of the engine 1 for purifyng noxious components such as HC, CO, and NOx. An O.sub.2 sensor 15 as an exhaust gas ingredient concentration sensor is mounted in the exhaust pipe 13 at a location upstream of the three-way catalyst 14, or sensing the concentration of oxygen present in exhaust gases emitted from the engine 1 and supplying an electrical signal indicative of the sensed oxygen concentration to the ECU 5.

Further electrically connected to the ECU 5 are a vehicle speed sensor 16, a gear position sensor 17 for detecting the shift lever position of a transmission, and an oil pressure sensor 18 for detecting oil pressure in oil feeding passages (88a, 88e in FIG. 2), referenced to hereinafter, of the engine 1. Signals from these sensors are supplied to the ECU 5.

The ECU 5 comprises an input circuit 5a having the functions of shaping the waveforms of input signals from various sensors, shifting the voltage levels of sensor output signals to a predetermined level, converting analog signals from analog-output sensors to digital signals, and so forth, a central processing unit (hereinafter called "the CPU") 5b, memory means 5c storing various operational programs which are executed in the CPU 5b and for storing results of calculations therefrom, etc., and an output circuit 5d which outputs driving signals to the fuel injection valves 6, the driving circuit 21, and the electromagnetic valve 23.

The CPU 5b operates in response to the above-mentioned signals from the sensors to determine operating conditions in which the engine 1 is operating such as an air-fuel ration feedback control region for controlling the air-fuel ratio in response to oxygen concentration in exhaust gases and open-loop control regions, and calculates, based upon the determined operating conditions, the valve opening period or fuel injection period T.sub.OUT over which the fuel injection valves 6 are to be opened, by the use of the following equation in synchronism with inputting of TDC signal pulses to the ECU 5.

where Ti represents a basic fuel amount, more specifically a basic fuel injection period of the fuel injection valves 6, which is determined based upon the engine rotational speed Ne and the intake pipe absolute pressure P.sub.BA. As the Ti map for determining the Ti value, a Ti.sub.L map for low speed valve timing and a Ti.sub.H map for high speed valve timing are stored in the memory means 5C.

K.sub.WOT represents a high load enriching coefficient for increasing the amount of fuel in a predetermined high load engine operating region.

K.sub.1 and K.sub.2 represent other correction coefficients and correction variables, respectively, which are calculated based on various engine parameter signals to such values as to optimize operating characteristics of the engine such as fuel consumption and accelerability, depending on operating conditions of the engine.

The CPU 5b decides the ignition timing .theta.ig based on the engine rotational speed Ne and the intake pipe absolute pressure P.sub.BA. As the .theta.ig map for determining the ignition timing, similarly to the Ti maps, a .theta.gi.sub.L map for the low speed valve timing and a .theta.ig.sub.H map for the high speed valve timing are stored in the memory means 5C.

Further, the CPU 5b controls the opening and closing of the electromagnetic valve 23 in a manner described hereinafter with reference to FIG. 9.

The CPU 5b supplies the output circuit 5d with driving signals for driving the fuel injection valves 6, the driving circuit 21, and the electromagnetic valve 23, based upon the results of the above calculations and decisions.

FIG. 2 shows a longitudinal cross-section of essential parts of the engine 1. Four cylinders 32, only one of which is shown, are arranged in a line within a cylinder block 31. Combustion chambers 35 are defined between a cylinder head 33 mounted on an upper end of the cylinder block 31 and pistons 34 slidably fitted within respective cylinders 32. The cylinder head 33 has a pair of inlet ports 36 and a pair of exhaust ports 37 formed in a portion thereof serving as a ceiling of each combustion chamber. Each inlet port 36 is connected to an inlet passage 38 which opens in one side wall of the cylinder head 33, while each exhaust port 37 is connected to an exhaust passage 39 which opens in another side wall of the cylinder head 33.

An inlet valve 40i is arranged in each inlet part 36 to open and close same, while an exhaust valve 40e is arranged in each exhaust port 37 to open and close same. The inlet valves 40i and exhaust valves 40e are guided by respective guide sleeves 41i' and 41e' which are fitted in respective guide holes 41i and 41e formed in the cylinder head 33. Valve springs 43i, 43e are interposed between respective valve seats formed at ends of the guide holes 41i, 41e and respective collars 42i, 42e secured on upper ends of each inlet valve 40i and each exhaust valve 40e projected from the respective guide holes 41i, 41e. The valve springs 43i, 43e urge the respective inlet and exhaust valves 40i, 40e in the upward or valve-closing direction.

The cylinder head 33 and a head cover 44 mounted on an upper end thereof define therebetween a working chamber 45 which accommodates an inlet valve-operating device 47i for opening and closing the inlet valve 40i in each cylinder 32, and an exhaust valve-operating device 47e for opening and closing the exhaust valve 40e in same. The valve-operating devices 47i, 47e are basically of the same construction. Therefore, only the component parts of the inlet valve-operating device 47i will be described below with reference numerals having a letter i affixed thereto, while those of the exhaust valve-operating device 47e are merely shown in the drawings with corresponding reference numerals having a letter e affixed thereto.

Referring also to FIG. 3, the inlet valve-operating device 47i comprises a camshaft 48i rotatatively driven by a crankshaft, not shown, at a speed ratio of 1/2, a high speed cam 51i and low speed cams 49i, 50i secured on the camshaft 48i, provided for each cylinder 32 (the low speed cam 50i has substantially the same configuration as the low speed cam 49i, both being arranged on opposite sides of the high speed cam 51i), a rocker shaft 52i extending parallel with the camshaft 48i, first and second driving rocker arms 53i and 54i, and a free rocker arm 55i pivotally mounted on the rocker shaft 52i, the three arms being provided for each cylinder 32, a connection-changeover mechanism 56i arranged in the corresponding rocker arms 53i, 54i, 55i for each cylinder.

As shown in FIG. 3, the connection-changeover mechanism 56i comprises a first changeover pin 81 capable of connecting the first driving rocker arm 53i with the free rocker arm 55i, a second changeover pin 82 capable of connecting the second driving rocker arm 54i with the free rocker arm 55i, a restriction pin 83 for restricting the movement of the first and second changeover pins 81, 82, and a spring 84 urging the pins 81, 82, 83 in the rocker arms-disconnecting direction.

The first driving rocker arm 53 i if formed therein with a first guide bore 85 extending parallel with the rocker shaft 52i with one end thereof closed and the other end opening in a side face thereof facing the free rocker arm 55i. The first changeover pin 81 is slidably fitted in the first guide bore 85, defining an oil hydraulic chamber 86 between one end thereof and the closed end of the first guide bore 85. Further, a passage 87 extends from the oil hydraulic chamber 86 and opens into an oil feeding passage 88i formed in the rocker shaft 52i so that the passage 88i permanently communicates via the passage 87 with the oil hydraulic chamber 86 irrespective of rocking motion of the first driving rocker arem 53i.

The free rocker 55i is formed therein with a guide through hole 89 at a location corresponding to the first guide bore 85, which extends through the free rocker arm 55i and parallel with the rocker shaft 52i. The second changeover pin 82 is slidably fitted in the guide through hole 89, with one end thereof abutting on an opposed end face of the first changeover pin 81.

The second driving rocker arm 54i is formed therein with a second guide bore 90 at a location corresponding to the guide through hole 89, which extends parallel with the rocker shaft 52i with one end thereof opening toward the free rocker arm 55i. The restriction pin 83 in the form of a disc is slidably fitted in the second guide bore 90, in a fashion abutting on the other end of the second changeover pin 82. Further, the second guide bore 90 has a guide sleeve 91 fitted therein, in which is slidably fitted an axial rod 92 which coaxially and integrally projects from the restriction pin 82. The spring 84 is interposed between the guide sleeve 91 and the restriction pin 83 and urges the pins 81, 82, 83 toward the oil hydraulic chamber 86.

In the connection-changeover mechanism 56i constructed as above, when the pressure in the oil hydraulic chamber 86 is increased, the first changeover pin 81 is forced to move into the guide through hole 89 and at the same time the second changeover pin 82 is urgedly moved into the second guide bore 90 to connect the rocker arms 53i, 55i, 54i together. When the pressure in the oil hydraulic chamber 86 is decreased, the first changeover pin 81 is moved back by the urging force of the spring 84 into a position in which the end face thereof abutting on the second changeover pin 82 corresponds in location to the space between the first driving rocker arm 53i and the free rocker arm 55i, and at the same time the second changeover pin 82 is moved back into a position in which the end face thereof abutting on the restriction pin 83 corresponds in location to the space between the free rocker arm 55i and the second driving rocker arm 54i, whereby the rocker arms 53i, 55i, 54i become disconnected from each other.

Next, the oil-feeding system for supplying oil to the valve-operating devices 47i, 47e will be described with reference to FIGS. 4A-4C. Oil galleries 98, 98' are connected to an oil pump, not shown, for pumping oil from an oil pan, not shown. From the oil galleries 98, 98', oil pressure is supplied to the connection-changeover mechanisms 56i, 56e, while lubricating oil is supplied to the lubricating parts of the valve-operating devices 47i, 47e.

Connected to the oil gallery 98 is a selector valve 99 for changing the oil pressure supplied between high and low levels. The oil feeding passages 88i, 88e in the respective rocker shafts 52i, 52e are connected via the selector valve 99 with the oil gallery 98.

Passage-forming members 102i, 102e respectively extend parallel with the camshafts 48i, 48e and are secured to upper surfaces of cam holders 59 by means of a plurality of bolts. The passage-forming members 102i, 102e are formed therein with respective low speed lubricating oil passages 104i, 104e, and high speed lubricating oil passages 105i, 105e, all of which have opposite closed ends and form pairs of parallel passages. The low speed lubricating oil passages 104i, 104e are connected to the oil gallery 98', and the high speed lubricating oil passages 105i, 105e are connected to the oil feeding passages 88i, 88e. Further, the low speed lubricating oil passages 104i, 104e are connected to the cam holder 59.

The selector valve 99 comprises an oil inlet port 119 connected to the oil gallery 98, an oil outlet port 120 connected to the oil feeding passages 88i, 88e, and a spool valve 122 slidably fitted within a housing 121 mounted on one end face of the cylinder head 33.

The housing 121 is formed therein with a cylinder hole 124 having an upper end thereof closed with a cap 123, and within which is fitted the spool valve 122 to define an operating oil hydraulic chamber 125 between an upper end thereof and the cap 123. Further, a spring 127 is accommodated within a spring chamber 126 defined between the spool valve 122 and a lower part of the housing 121 and urges the spool valve 122 in the upward or valve-closing direction. The spool valve 122 has an annular recess 128 formed therearound for communicating between the oil inlet port 119 and the oil outlet port 120. When the spool valve is in an upper position as shown in FIG. 4, it cuts off the communication between the oil inlet port 119 and the oil outlet port 120.

An oil filter 129 is held between the oil inlet port 119 and a high speed oil pressure feeding passage 116. Further, the housing 121 has a restriction passage 131 formed therein, which provides communication between the oil inlet port 119 and the oil outlet port 120. Therefore, even when the spool valve 122 is in a closed position, the oil inlet port 119 and the oil outlet port 120 are communicated with each other through the restriction passage 131 whereby oil pressure decreased through the restriction passage 131 is supplied via the oil outlet port 120 to the oil feeding passages 88i, 88e.

The housnig 121 also has a bypass port 132 formed therein, which is disposed to communicate via the annular recess 128 with the oil outlet port 120 only when the spool valve 22 is in the closed position. The bypass port 132 communicates with the interior of the cylinder heat 33 at an upper location thereof.

Connected to the housing 121 is a conduit line 135 which always communicates with the oil inlet port 119. The conduit line 135 is connected via the electromagnetic valve 23 with a conduit line 137 which is in turn connected with a communcation hole 138 formed in the cap 123. Therefore, when the electromagnetic valve 23 is opened, oil pressure is supplied to the operating oil hydraulic chamber 125 to thereby move the spool valve 122 in the valve-opening direction.

Further, the housing 121 has the oil pressure sensor 18 mounted thereon for detecting the oil pressure in the outlet port 120, i.e. the oil pressure in oil feeding passages 88i, 88e, to determine whether the selector valve 99 is normally functioning or not.

The operation of the valve-operating devices 47i, 47e having the above-described construction will be described below. Since the valve-operating devices 47i, 47e operate similarly to each other, the following description refers only to the operation of the inlet valve-operating device 47.

When the ECU 5 sends out a valve-opening instruction signal to the electromagnetic valve 23, the electromagnetic valve 23 is opened to thereby cause the selector valve 99 to open, so that the oil pressure in the oil feeding passage 88i is increased. This causes the connection-changeover mechanism 56i to operate to connect the rocker arms 53i, 54i, 55i together, whereby the high speed cam 51i operates the rocker arms 53i, 54i, 55i in unison to cause each pair of inlet valves 40i to open and close at high speed valve timing in which the valve-opening period and the valve lift amount are relatively greater.

On the other hand, when the ECU 5 supplies a valve-closing instruction signal to the electromagnetic valve 23, the electromagnetic valve 23 and in turn the selector valve 99 are closed to thereby decrease the oil pressure in the oil feeding passage 88i. This causes the connection-changeover mechanism 56i to operate to disconnect the rocker arms 53i, 54i, 55i from each other, whereby the low speed cams 49i, 50i operate the corresponding rocker arms 53i, 54i to cause the pair of inlet valves 40i to open and close at low speed valve timing in which the valve-opening period and the valve lift amount are relatively smaller.

Next, the valve timing-changeover control according to the invention will be described below.

In FIG. 5, the Ti value of the Ti.sub.L map for low speed valve timing and that of the Ti.sub.H map for high speed valve timing are respectively indicated by solid lines and dotted lines. As is clear from the figure, in the case of the low speed valve timing being selected, the rate of increase in the intake air amount becomes smaller with increase in the engine rotational speed Ne, while in the case of the high speed valve timing being selected, the charging efficiency becomes higher with increase in the engine rotational speed Ne whereby the intake air amount becomes greater than in the case of the low speed valve timing being selected. Therefore, there is a point of engine rotational speed Ne where the Ti value for low speed valve timing and the Ti value for high speed valve timing are identical to each other. In this point of Ne, in both cases of the high and low speed valve timings being selected, the intake air amounts is identical, and at the same time the air-fuel ratio is also identical, so that the engine output becomes substantially identical.

The charging efficiency finely varies with the engine rotational speed Ne, and particularly in the vicinity of the maximum throttle valve opening (.theta.th), the variation becomes markedly great. FIG. 6 shows on an enlarged scale part of FIG. 5 for explaining this variation. At a plurality of points, the Ti value for low speed valve timing and that for high speed valve timing become identical to each other. As described hereinafter, when the valve timing is changed at a point where the Ti value for low speed valve timing and that for high speed valve timing are identical to each other, hunting in changeover the valve timing, i.e. frequent changeover of valve timing, is liable to occur in the region of wide throttle valve opening (WOT), which adversely affects the durablity of the connection-changeover mechanisms 56i, 56e.

In this connection, when the engine is in a high load operating region (maximum .theta.th (WOT) region), the air-fuel ratio is enriched by the high load enriching coefficient K.sub.WOT to increase the engine output. In such a high load operating region, the engine output can be more effectively increased if the valve timing is changed to the high speed valve timing. Therefore, when the engine is in the high load operating region (maximum .theta.th (WOT) region), from a T.sub.VT table in which a high load determination value T.sub.VT experimetally obtained based on the fuel injection amount T.sub.OUT is set in relation to the engine rotational speed Ne as shown in FIG. 7, a T.sub.VT value is obtained in accordance with the engine rotational speed Ne, and when the fuel injection amount T.sub.OUT is equal to or higher than the T.sub.VT value, the valve timing is changed to the high speed valve timing. In this case, if it is arranged that the region defined by T.sub.OUT .gtoreq.T.sub.VT includes the aforesaid points in the wide throttle valve opening region where the Ti value for low speed valve timing and that for high speed valve timing are identical to each other, the hunting in the valve timing can be prevented. In addition, the T.sub.VT table used for vehicles with automatic transmissions is different from that used for vehicles with manual transmissions.

Further, generally, in order to prevent overspeed of the engine, fuel cut is carried out when the engine rotational speed Ne exceeds a predetermined value (so-called revolution limitter value) N.sub.HFC. Load acting on a timing belt connecting between the crankshaft and the camshaft increases with decrease in the opening period of the valve because the acceleration of opening movement of the valve increases with decrease in the valve opening period. Further, as the acceleration increases, a critical value of the engine rotation speed above which there can occur jumping of the valve decreases. Therefore, the maximum allowable engine rotational speed should be different between when the valve timing is set to the low speed valve timing in which the valve opening period is shorter and when the valve timing is set to the high speed valve timing in which the valve opening period is longer. Accordingly, in this embodiment, the revolution limitter value is set at a relatively low value N.sub.HFC1 (e.g. 7500 rpm) for the low speed valve timing, and at a relatively high value N.sub.HFC2 (e.g. 8100 rpm) for the high speed valve timing.

Next, reference is made to FIG. 8 showing the valve timing regions. In the figure, the solid line indicates a boundary line between the low speed valve timing region and the high speed valve timing region, which is selected when the valve timing is changed from the low speed valve timing to the high speed valve timing, and the broken line indicates one which is selected when the valve timing is changed from the high speed valve timing to the low speed valve timing.

The changeover of the valve timing is carried out in a region between a value Ne.sub.1 of the engine rotational speed below which the engine output obtained by the low speed valve timing always exceeds the engine output obtained by the high speed valve timing and a value Ne.sub.2 of the engine rotational speed above which the engine output obtained by the high speed valve timing always exceeds the engine output obtained by the low speed valve timing. In this embodiment, hysteresis is imparted to the engine rotational speed values Ne.sub.1 and Ne.sub.2 between changeover of the valve timing from the low speed valve timing to the high speed valve timing and vice versa such that Ne.sub.1 is set to e.g. 4800 rpm/4600 rpm and Ne.sub.2 set to e.g. 5900 rpm/5700 rpm.

In FIG. 8, X indicates a region in which the engine is in a high load operating region (WOT region) and the changeover of valve timing is carried out by comparison between T.sub.OUT and T.sub.VT, and Y indicates a region in which the changeover of the valve timing is carried out by comparison between a T.sub.IL value for the low speed valve timing and a T.sub.IH value for the high speed valve timing. Incidentally, since the changeover characteristic in the region X is also under the influence of parameters other than the engine rotational speed Ne and the intake pipe absolute pressure P.sub.BA used for calculation of T.sub.OUT, the changeover characteristic cannot be accurately plotted in FIG. 8 in which the engine rotational speed Ne is indicated by the abscissa and the intake pipe absolute pressure P.sub.BA is indicated by the ordinate. Therefore, the changeover characteristic in the region X of FIG. 8 should be taken as one for mere understanding of the concept of the invention.

Next, reference is made to FIGS. 9A-9C to explain a program for controlling the changeover of the valve timing executed by the ECU 5, i.e. a program for output control of signals supplied to the electromagnetic valve 23. This program is executed upon generation of each pulse of the TDC signal and in synchronism therewith.

At a step 901, it is determined whether or not a failsafe operation should be carried out, e.g. by determining whether or not any engine operating parameter sensor is normally functioning or whether or not abnormality has occurred in the control system other than such sensor.

Specifically, it is determined that the engine is in an operating condition in which a failsafe operation should be carried out, if, for example, there is detected an abnormality in any of the outputs from the intake pipe absolute pressure (P.sub.BA) sensor 8, the cylinder-discriminating (CYL) sensor 12, the engine rotational speed (TDC) sensor 11, the engine coolant temperature sensor 10, and the vehicle speed sensor 16, an abnormality in outputting of a control signal for ignition timing or in outputting of driving signals for the fuel injection valves, an abnormality in the amount of electric current supplied to the electromagnetic valve 23 for the valve timing control, or an abnormality that a normal change has not been detected in oil pressure at the oil outlet port 120 responsive to opening and closing of the electromagnetic valve 23 for the valve timing control by an oil pressure switch of the oil pressure sensor 18, over a predetermined time period. In addition, when one of the CYL sensor and the TDC sensor is abnormal, the other is used in place thereof.

At a step 902, it is determined whether or not the engine is being started, from the engine rotational speed Ne, etc., and at a step 903, it is determined whether or not a delay timer has counted up a predetermined time period (e.g. 5 seconds) t.sub.ST. If the answer to the question of the step 902 is Yes, the program proceeds to a step 904, where the timer is set to the predetermined time period t.sub.ST during failsafe operation or engine starting, for starting to count same after the engine starting has been completed. At a step 905, it is determined whether or not the engine coolant temperature T.sub.W is lower than a predetermined value T.sub.W1 (e.g. 60.degree.C.), i.e. whether or not the engine is in warming-up operation. At a step 906, it is determined whether or not the vehicle speed V is lower than a very low predetermined value V.sub.1 (with hysteresis, e.g. 8 km/5 km). At a step 907, it is determined whether or not the vehicle on which the engine is installed is provided with a manual transmission (MT). At a step 908, it is determined, when the vehicle is an automatic transmission type (AT), whether or not the shift lever is positioned in the parking range (P) or the neutral range (N). At a step 909, it is determined whether or not the engine rotational speed Ne is not lower than the predetermined lower limit value Ne.sub.1 (e.g. 4800 rpm/4600 rpm). If as a result of the above determinations, the failsafe operation is being carried out (the answer to the question of the step 901 is Yes), or if the engine is being started (the answer to the question of the step 902 is Yes), or if the predetermined time period t.sub. ST has not elapsed after the enginer has completed starting (the answer to the question of the step 903 is No), or if the engine is still in warming up operation (the answer to the question of the step 905 is Yes), or if the vehicle is standing or moving slowly (the answer to the question of the step 906 is Yes), or if the shift lever is in the P or N range (the answer to the question of the step 908 is Yes), or if Ne<Ne.sub.1 (the answer to the question of the step 909 is No), the electromagnetic valve 23 is closed to maintain the low speed valve timing.

If it is determined at the step 909 that Ne.gtoreq.Ne.sub.1 is satisfied, at a step 910, from the Ti.sub.L map and the Ti.sub.H map, there are obtained a Ti value (hereinafter referred to as "Ti.sub.L ") of the Ti.sub.L map and a Ti value (hereinafter referred to as "Ti.sub.H ") of the Ti.sub.H map each corresponding to the engine rotational speed Ne and the intake air absolute pressure P.sub.BA. Then, at a step 911, from the T.sub.VT table set depending on whether the vehicle is AT or MT is obtained a high load determination value T.sub.VT corresponding to the engine rotational speed Ne. At a step 912, the T.sub.VT is compared with the T.sub.OUT in the immediately preceding loop to determine whether T.sub.OUT .gtoreq.T.sub.VT is satisfied, i.e. whether the engine is in the high load operating condition in which the air-fuel ratio should be enriched. If the step 912 is No, i.e. if T.sub.OUT <T.sub.VT is satisfied, the program proceeds to a step 913, where it is determined whether or not the engine rotational speed Ne is not lower than the predetermined upper limit value Ne.sub.2. If the answer to the question of the step 913 is No, i.e. if Ne<Ne.sub.2 is satisfied, the program proceeds to a step 914, where the Ti.sub.L and the Ti.sub.H obtained at the step 910 are compared with each other. If Ti.sub.L >Ti.sub.H is satisfied, it is determined at a step 916 whether or not a timer value t.sub.VTOFF of a delay timer set at a step 915, referred to hereinafter, has been counted up. If the answer to the question of the step 916 is Yes, an instruction signal for closing the electromagnetic valve 23, i.e. an instruction for changing the valve timing to the low speed valve timing, is generated at a step 917. On the other hand, if any of T.sub.OUT .gtoreq.T.sub.VT, Ne.gtoreq.Ne.sub.2, and Ti.sub.L .ltoreq.Ti.sub.H is satisfied, the delay timer for closing the electromagnetic valve is set to the predetermined value t.sub.VTOFF (e.g. 3 seconds) and started at the step 915. Then at a step 918, an instruction signal for opening the electromagnetic valve 23, i.e. an instruction for changing the valve timing to the high speed valve timing is generated.

If the valve-closing signal is generated at the step 917, it is determined at a step 919 whether or not the oil pressure switch within the oil pressure sensor 18 has been turned on, i.e. if the oil pressure in the oil feeding passages 88i, 88e has become low. If the answer to the question of the step 919 is Yes, i.e. if the oil pressure switch has been turned on, it is determined at a step 921 whether or not a changeover delay timer has counted up a predetermined time period t.sub.LVT for the low speed valve timing. If the answer to the question of the step 921 is Yes, i.e. if t.sub.LVT= 0, another changeover delay timer for the high speed valve timing is set to a predetermined time period t.sub.HVT (e.g. 0.1 second) and started at a step 923. Then at a step 925, the Ti.sub.L map and an ignition timing map (.theta.ig.sub.L map) for the low speed valve timing are selected as the Ti map and the ignition timing map to be used in a routine for fuel injection control. At the following step 927, the revolution limitter value N.sub.HFC is set to a predetermined value N.sub.HFC1 for the low speed valve timing.

On the other hand, if the valve-opening signal is generated at the step 918, it is determined at a step 920 whether or not the oil pressure switch within the oil pressure sensor 18 has been turned off, i.e. if the oil pressure in the oil feeding passages 88i, 88e has become high. If the answer to the question of the step 920 is Yes, i.e. if the oil pressure switch has been turned off, it is determined at a step 922 whether or not the changeover delay timer for the high speed valve timing has counted up the value t.sub.HVT. If the answer to the question of the step 922 is Yes, i.e. if t.sub.HVT= 0, the the changeover delay timer for the low speed valve timing is set to a predetermined time period t.sub.LVT (e.g. 0.2 seconds) at a step 924, and then at a step 926, the Ti.sub.H map and an ignition timing map (.theta.ig.sub.H map) for the high speed valve timing are selected as the Ti map and the ignition timing map to be used in the routine for fuel injection control. At the following step 928, the revolution limitter value N.sub.HFC is set to a predetermined value N.sub.HFC2 for the high speed valve timing, which is higher than N.sub.HFC1.

The predetermined delay time periods t.sub.HVT and t.sub.LVT are set at such values as correspond to the respective time lags, i.e. periods of time to be elapsed from opening and closing of the electromagnetic valve 23, through switching of the selector valve 99, and changing of the oil pressure in the oil feeding passages 88i, 88e, until completion of changeover operations by the connection-changeover mechanisms 56i, 56e of all the cylinders. When the electromagnetic switch 23 is closed, the program proceeds in the order of 919-922-924-926-928 until the oil pressure switch within the oil pressure sensor 18 is turned on. After the oil pressure switch has been turned on, the program proceeds in the order of 919-921-926-928 until the connection-changeover mechanisms 56i, 56e of all the cylinders have been changed over to the low speed valve timing position. Further, if the selector valve 99 is not closed due to failure of the electromagnetic valve 23 or the selector valve 99 etc. so that the oil pressure switch within the oil pressure sensor 18 remains open or off, the program also proceeds in the above-mentioned order of 919-922-924-926-928. Thus, until the connection-changeover mechanisms 56i, 56e of all the cylinders have been changed to the low speed valve timing position, the fuel injection is controlled in a manner suitable for the high speed valve timing. Also, when the electromagnetic switch 23 is opened, the fuel injection is controlled in a manner suitable for the low speed valve timing until the connection-changeover mechanisms 56i, 56e of all the cylinders have been changed to the high speed valve timing position.

In the meanwhile, if failsafe operation is being carried out (the answer to the question of the step 901 is Yes), or if the engine is being started (the answer to the question of the step 902 is Yes), or if the time period t.sub.ST has not elapsed after completion of the engine starting (the answer to the question of the step 903 is No), or if the engine has not yet been warmed up (the answer to the question of the step 905 is Yes), or if the vehicle is standing or moving slowly (the answer to the question of the step 906 is Yes), the program proceeds to the step 929, where the instruction signal for closing the electromagnetic valve 23 is generated, followed by the program proceeding in the order of 923-925-927. If it is determined at the step 908 that the shift lever position is in the N or P range, the program proceeds to a step 930, where it is determined whether or not the Ti.sub.H map has been selected in the immediately preceding loop. Also, if it is determined at the step 909 that Ne<Ne.sub.1 is satisfied, the program proceeds to the step 930. If the answer to the question of the step 930 is Yes, i.e. if the Ti.sub.H map has been selected in the immediately preceding loop, the time period t.sub.VTOFF of the delay timer over which the electromagnetic valve is to be opened is set to 0 at a step 931, and then the program proceeds to a step 917. If the answer to the question of the step 930 is No, i.e. if the Ti.sub.H map has not been used in the immediately preceding loop, in other words, if the connection-changeover mechanisms 56i, 56e of all the cylinders have not been changed over to the high speed valve timing position, the program proceeds, as described above, in the order of 929-923-925-927, whereby the fuel injection is controlled in a manner suitable for the low speed valve timing irrespective of the state of the oil pressure switch within the oil pressure sensor 18. This is a countermeasure for the case in which the oil pressure switch within the oil pressure sensor 18 continues to be off due to disconnection in the wiring etc.

The aforesaid revolution limitter Ne value N.sub.HFC1 is set at a value higher than Ne.sub.2, and normally the valve timing is switched to the high speed valve timing and accordingly the revolution limitter N.sub.HFC is set to the higher vlaue N.sub.HFC2 before the engine rotational speed Ne rises to N.sub.HFC1, so that fuel cut is not carried out even at N.sub.HFC1. On the other hand, when the engine is in an operating condition in which the program proceeds from any of the steps 902-906, and 908 to the step 929, the fuel cut can be carried out at N.sub.HFC1, since the low speed valve timing is maintained even after the engine rotational speed Ne exceeds Ne.sub.2 due to racing of the engine etc. Further, even after the valve timing is switched from the low speed valve timing to the high speed valve timing, fuel cut is carried out at N.sub.HFC1 before t.sub.HVT becomes 0, i.e. before the connection-changeover mechanisms 56i, 56e are actually changed over to the high speed valve timing position.

FIG. 10 shows the subroutine used at the step 910 for obtaining Ti.sub.L and Ti.sub.H from the respective Ti.sub.L and Ti.sub.H maps. It is determined whether or not the instruction signal for opening the electromagnetic switch 23 has been generated in the immediately preceding loop. If the instruction signal has not been generated, the Ti.sub.L to be used at the step 914 is set to a value Ti.sub.L obtained from the Ti.sub.L map, whereas if the instruction signal has been generated, the Ti.sub.L to be used at the step 914 is set to a value obtained by subtracting a predetermined hysteresis amount of .DELTA.Ti from a valve Ti.sub.L obtained from the Ti.sub.L map. Thus, hysteresis is imparted to the changeover characteristic in the region Y in FIG. 8.

FIG. 11 shows the subroutine used at the step 911 for obtaining the high load determination value T.sub.VT from the T.sub.VT table. It is determined whether or not the instruction signal for opening the electromagnetic valve 23 has been generated in the immediately preceding loop. If the signal has not been generated, the T.sub.VT to be used at the step 912 is set to a value T.sub.VT obtained from the T.sub.VT table, whereas if the signal has been generated, the T.sub.VT to be used at the step 912 is set to a value obtained by subtracting a predetermined hysteresis amount .DELTA.T.sub.VT from a value T.sub.VT obtained from the T.sub.VT table. Thus, hysteresis is imparted to the changeover characteristic in the region X in FIG. 8.

Referring again to FIG. 9, if the answer to the question of the step 901 is Yes, i.e. if the failsafe operation is being carried out, the instruction signal for closing the electromagnetic valve 23 is generated at the step 932, and then at a step 933 it is determined whether or not the engine rotational speed Ne is higher than a predetermined value Ne.sub.FS for fail-safe (e.g. 3,000 rpm). If the answer at the step 933 is Yes, i.e. if Ne>Ne.sub.FS, the Ti.sub.H map and the .theta.ig.sub.H map for the high speed valve timing are selected at a step 934, followed by the program proceeding to the step 927, whereas if Ne.ltoreq.Ne.sub.FS, the Ti.sub.L map and the .theta.ig.sub.L map for the low speed valve timing are selected at a step 935, followed by the program proceeding to the step 927.

As described above, at the steps 933-935 one of the Ti.sub.H Ti.sub.L maps is selected depending on the engine rotational speed Ne during the fail-safe operation. Therefore, even when the inlet and exhaust valves 40i, 40e are actually operated at the high speed valve timing due to failure of any of the selector valve 99, the connection-changeover mechanisms 56i, 56e, etc. in spite of the fact that the instruction signal for closing the electromagnetic valve 23 is generated during the failsafe operation, it is possible to prevent overleaning of the air-fuel ratio and hence an excessive rise in the buring temperature of the mixture or the catalyst temperature of exhaust gas purifying means, and accordingly also prevent melting of ignition plugs due to preignition of the mixture, knocking at a high engine rotational speed, and shortened life of the catalyst.

The manner of ignition timing control which is carried out by the knocking control system of the invention will now be explained with reference to FIG. 12.

At a step 1201, the detected and stored engine rotational speed Ne and intake pipe absolute pressure P.sub.BA are read out, and at a step 1202 a basic ignition timing advancing amount .theta.ig.sub.-BASE is retrieved from the .theta.ig.sub.L or .theta.ig.sub.H map selected at the step 925 or 926 in FIG. 9, in response to the engine rotational speed Ne and the intake pipe absolute pressure P.sub.BA read out at the step 1201. Then, the engine coolant temperature T.sub.W and the intake air temperature T.sub.A are read out at a step 1203, and at a step 1204 a correction variable .theta.ig.sub.-COR for correcting the basic ignition timing advancing amount .theta.ig.sub.-BASE is calculated based on the engine coolant temperature T.sub.W and the intake air temperature T.sub.A read out at the step 1203. At a step 1205, a correction variable .theta.ig.sub.KN for preventing knocking is read out, which is determined depending on the valve timing and the result of a determination as to occurrence of knocking, hereinafter referred to, and then stored. A final ignition timing advancing amount .theta.ig (namely a crank angle before the top-dead-center of each cylinder at the start of compression stroke) is calculated at a step 1206 by summing up the value .theta.ig.sub.-BASE, .theta.ig.sub.-COR , and .theta.ig.sub.KN obtained at the steps 1202, 1204, and 1205, and at a step 1207 an ignition signal based on the calculated value .theta.ig is supplied to the driving circuit 21 to effect ignition by the ignition plug 22.

The determination as to whether knocking has occurred, which is used for determining the correction variable .theta.ig.sub.KN read out at the step 1025, is carried out by the use of the knocking sensor 24 and means 1301-1305 shown in (b) of FIG. 13. Specifically, an output from the knocking sensor 24 which is indicative of engine vibration is applied to gate means 1301, where the sensor output is subjected to detection by a knocking gate and a noise gate of the gate means 1301, as shown in (a) of FIG. 13, at both the low speed valve timing (hereinafter referred to as "Lo V/T") and the high speed valve timing (hereinafter referred to as "Hi V/T"). Vibration detected by the noise gate is levelled through a low-pass filter having a predetermined time constant, of noise level-calculation means 1302, to obtain mean noise levels NL.sub.L, NL.sub.H at Lo V/T and Hi V/T, respectively. The mean noise levels NL.sub.L, NL.sub.H are multiplied by multiplier terms G.sub.L, G.sub.H and increased by addened terms OS.sub.L, OS.sub.H by the following equations (2) and (3) at knocking discrimination calculating means 1303, to obtain respective knocking discrimination levels aLo, aHi at Lo V/T and Hi V/T.

where the multiplier terms G.sub.L, G.sub.H and addend terms OS.sub.L, OS.sub.H are set at Lo V/T and Hi V/T, respectively, and obtained from graphs shown in FIG. 17, respectively. The multiplier terms and addend terms are both set to larger values as the engine rotational speed Ne or the engine load increases. Further, the multiplier term G.sub.H and addend term OS.sub.H at Hi V/T are set to larger values than then the multiplier term G.sub.L and addend term OS.sub.L at Lo V/T. Also, the increasing rates of the multiplier term G.sub.L and addend term OS.sub.L at Lo V/T become reduced when the engine rotational speed Ne or engine load exceeds a chageover point CH, e.g. 4800 rpm. By thus setting the multiplier terms G.sub.L, G.sub.H and addend terms OS.sub.L, OS.sub.H, the knocking discrimination levels aLo, aHi at Lo V/T and Hi V/T are substantially equal to each other and at the same time set in the vicinity of a lower limit value of a knocking level region, in spite of the fact that the noise level NL.sub.L at Lo V/T is different from the noise level NL.sub.H at Hi V/T, as shown in FIG. 18.

In practice, the knocking discrimination level calculating means is composed of amplifiers adapted such that the multiplier terms G.sub.L, G.sub.H, and the addend terms OS.sub.L, OS.sub.H are determined by the gains and offset of the amplifiers, respectively, in dependence on the engine rotational speed Ne or engine load, as well as on the selected valve timing.

Then, the vibration levels detected by the knocking gate at both Lo V/T and Hi V/T are compared with the respective knocking discrimination levels aLo and aHi by means of knocking discriminating means 1304. If the former is higher than the later, the knocking discriminating means 1304 judges that knocking has occurred, and then supplies an electric signal indicative of occurrence of knocking to knocking control means 1305.

The manner or knocking control will now be described with reference to FIG. 14.

At a step 1401, it is determined whether or not there exists a failure in the value timing control means, that is, it is determined similarly to the step 901 in FIG. 9 whether or not the failsafe operation should be carried out. If the answer at the step 1401 is No, that is, if there is no failure in the valve timing control means, the program proceeds to a step 1402, where it is determined whether or not the actual valve timing is Lo V/T. If at the step 925 in FIG. 9 the Lo V/T map has been selected, the valve timing is judged to be Lo V/T, whereas if at the step 926 the Hi V/T map has been selected, the valve timing is judged to be Hi V/T.

If the answer at the step 1402 is Yes, that is, if Lo V/T is being applied, various control data for Lo V/T are selected at a step 1403, whereas if the answer is No, that is, if Hi V/T is being applied, various control data for Hi V/T are selected at a step 1404.

The various control data include: the multiplier terms G.sub.L, G.sub.H, and the addend terms OS.sub.L, OS.sub.H to be used at a step 1405, hereinafter referred to, for calculation of knocking discrimination levels; discrimination parameters to be used in the knocking discriminating means 1304 such as crank angles at which the knocking gate and noise gate are operated, the time constant of the low-pass filter for levelling noise levels, and a reference discrimination level used for detecting abnormality in the knocking sensor 24; and control parameters to be used in the knocking control means 1305 such as a predetermined retarding limit value .theta.ig.sub.KNRD in FIG. 15, a predetermined advancing limit value .theta.ig.sub.KNAV in FIG. 16, the frequency of occurrence of knocking used for determining the octane number of fuel (the octane number of fuel used is determined based on the knocking frequency, and the retarding amount of the ignition timing is set depending on the octane number), a predetermined retarding correction amount .DELTA..theta..sub.RD in FIG. 15, and a predetermined advancing correction amount .DELTA..theta..sub.AV in FIG. 16.

At the step 1405, the knocking discrimination levels aLo and aHi are calculated by means of the knocking discrimination level-calculating means 1303 (FIG. 13) on the basis of the multiplier terms G.sub.L, G.sub.H and addend terms OS.sub.L, OS.sub.H for Lo V/T and Hi V/T, respectively, from the various data selected at the steps 1403, 1404. At a step 1406, it is determined whether or not the output of the knocking sensor 24 detected by the knocking gate is higher than corresponding one of the levels aLo and aHi calculated at the step 1405. If the answer is Yes, which indicates that there exists knocking, the program proceeds to a step 1407 to calculate an ignition timing retarding value or ignition timing correction variable .theta.ig.sub.KN in accordance with a subroutine of FIG. 15, whereas if the answer is No, which indicates that there is no knocking, the program proceeds to a step 1408 to calculate an ignition timing advancing value (.theta.ig.sub.KN) in accordance with a subroutine of FIG. 16.

Referring again to the step 1401, if the answer is Yes, that is, if there exists a failure in the valve timing control means and hence the failsafe operation should be carried out, the program proceeds to a step 1409 where the failsafe operation for knocking is carried out.

The failsafe operation is carried out in accordance with a subroutine shown in FIG. 20. That is, similarly to the step 1403 in FIG. 14, various control data for Lo V/T are selected at a step 2001, followed by the program proceeding to the step 1405 in FIG. 14.

According to a first aspect of the invention, if a failure has occurred in the value timing control means, various control data for Lo V/T are selected, the knocking discrimination levels are set on the basis of the selected control data (step 1405), and it is determined in response to the output of the knocking sensor 24 whether or not knocking has occurred (step 1406).

That is, the control data for Lo V/T are selected irrespective of the actual valve timing, in the event of a failure in the valve timing control means. Therefore, when Lo V/T is correctly selected by the valve timing control means in a normal condition, the knocking discrimination level aLo is obtained to thereby effect normal knocking control, whereas if Hi V/T is wrongly selected due to a failure in the valve timing control means, the knocking discrimination level is obtained by multiplying the noise level NL.sub.H for Hi V/T by the multiplier term G.sub.L for Lo V/T and adding thereto the addend term OS.sub.L for Lo V/T. As described above with reference to FIGS. 17 and 18, the noise level at Hi V/T is lower than that at Lo V/T, and accordingly the multiplier term G.sub.L and the addend term OS.sub.L for Lo V/T are smaller than those for Hi V/T, respectively, so that the knocking discrimination level becomes a considerably smaller value when the valve timing control means is abnormal than when it is normal. Consequently, the answer at the step 1406 is liable to become Yes more often when the valve timing control means is abnormal, compared with the case where it is normal, so that the ignition timing is retarded in the former case. Thus, when there exists a failure in the valve timing control means, the ignition timing is retarded at the time of occurrence of knocking at a lower level to thereby eliminate same.

According to a second aspect of the invention, the failsafe operation is carried out in accordance with a subroutine in FIG. 21. At a step 2001, a predetermined retarding value .theta..sub.RETF/S for retarding the ignition timing is determined from a table shown in FIG. 22, and the value .theta..sub.RETF/S is set as the ignition timing correction variable .theta.ig.sub.KN at a step 2002, followed by termination of the program. Although, as shown in FIG. 22, the predetermined retarding value .theta..sub.RETF/S increases as the engine rotational speed Ne increases, the value .theta..sub.RETF/S may increase as the engine load increases.

Therefore, according to the second aspect of the invention, if there exists a failure in the valve timing control means, the program is terminated immediately after execution of the step 1409' (the failsafe operation in FIG. 21), as shown in FIG. 14A. That is, the knocking correction value .theta.ig.sub.KN is set to the predetermined retarding value .theta..sub.RETF/S to obtain the ignition timing .theta.ig and then effect ignition, without using the knocking control means responsive to the knocking sensor output for controlling knocking. By thus inhibiting the operation of the knocking control means, knocking can be prevented irrespective of the actual valve timing.

If it is determined at the step 1406 in FIG. 14 that there exists knocking, the ignition timing correction variable .theta.ig.sub.KN is calculated in accordance with a subroutine shown in FIG. 15. The variable .theta.ig.sub.KN is set e.g. to a value of 0 as an initial value in respective first loops of the subroutine in FIG. 15 and a subroutine in FIG. 16, hereinafter referred to.

At a step 1501, a count value CNO counted by a counter over which no knocking has taken place is set to a value of 0. Then at a step 1502, a value of the correction variable .theta.ig.sub.KN is calculated by substracting the predetermined correction amount .DELTA..theta..sub.RD for retarding the ignition timing from a value of the correction variable .theta.ig.sub.KN obtained in the last loop, followed by the program proceeding to a step 1503 where it is determined whether or not the value .theta.ig.sub.KN obtained at the step 1502 is larger than the predetermined retarding limit value .theta.ig.sub.KNRD. The predetermined retarding limit value .theta.ig.sub.KNRD is determined depending on the engine rotational speed Ne and the engine load. If the answer at the step 1503 is Yes, that is, if .theta.ig.sub.KN >.theta.ig.sub.KNRD, the correction variable .theta.ig.sub.KN is set to the predetermined retarding limit value .theta.ig.sub.KNRD at a step 1504, whereas if the answer is No, the correction variable .theta.ig.sub.KN is set to the value .theta.ig.sub.KN calculated at the step 1502.

On the other hand, when it is determined at the step 1406 that there is no knocking, the ignition timing correction variable .theta.ig.sub.KN is calculated in accordance with the subroutine of FIG. 16, in order to return or advance the ignition timing which has been retarded in accordance with the subroutine in FIG. 15.

At a step 1601, a predetermined count value CAV to be counted is retrieved from a map, not shown, depending on the engine rotational speed Ne and engine load. The predetermined count value CAV corresponds to a predetermined number or TDC signal pulses over which no knocking has taken place. At a step 1602, the count value CNO of the counter is increased by a value of 1, and then it is determined at a step 1603 whether or not the value obtained by adding the value of 1 to the count value CNO is larger than the predetermined count value CAV retrieved at the step 1601. If the answer at the step 1603 is Yes, that is, if the count value CNO has reached the predetermined count value CAV, the count value CNO is set to the value of 0 at a step 1604, followed by the program proceeding to a step 1605 to calculate the correction variable .theta.ig.sub.KN by adding the predetermined correction amount .DELTA..theta..sub.AV for advancing the ignition timing to the value of the correction variable .theta.ig.sub.KN obtained in the last loop. If the answer at the step 1603 is No, that is, if the count value CNO has not reached the predetermined count value CAV, the program skips over the steps 1604 and 1605 to a step 1606.

At the step 1606, it is determined whether or not the value of .theta.ig.sub.KN obtained at the step 1605, or, if the answer at the step 1603 is No, the value of .theta.ig.sub.KN obtained in the last loop is larger than the predetermined advancing limit value .theta.ig.sub.KNAV. The predetermined advancing limit value .theta.ig.sub.KNAV is determined depending on the engine rotational speed Ne and engine load. If the answer at the step 1606 is Yes, that is, if .theta.ig.sub.KN >.theta.ig.sub.KNAV, the correction variable .theta.ig.sub.KN is set to the value .theta.ig.sub.KNAV, whereas if the answer is No, the correction variable .theta.ig.sub.KN is set to the value .theta.ig.sub.KN.

The control parameters such as the correction amount .DELTA..theta..sub.RD for retarding the ignition timing, the correction amount .DELTA..theta..sub.AV for advancing the ignition timing, the predetermined retarding limit value .theta.ig.sub.KNRD, the predetermined advancing limit value .theta.ig.sub.KNAV, and the predetermined count value CAV are set as follows, depending on the valve timing:

This is, the charging efficiency at Hi V/T is higher than that at Lo V/T so that the compression ratio at Hi V/T is actually larger than that at Lo V/T, whereby knocking is more liable to take place at Hi V/T. In order to prevent knocking at Hi V/T, the control parameters are set to larger values at Hi V/T than at Lo V/T, as shown in the above table.

Although in the embodiment of the invention described above, the ignition timing is controlled in order to prevent knocking at the time of changeover of the valve timing, this is not limitative to the invention, but fuel supply control to enrich the air-fuel ratio or supercharging pressure control to reduce the supercharging pressure may be employed instead of the ignition timing control, in order to prevent knocking.

Further, although in the embodiment, the valve timing is changed between Lo V/T and Hi V/T, the valve timing may be varied in a continuous manner, whereby the knocking discrimination parameters or knocking control parameters are varied in a continuous manner in accordance with the continuously varied valve timing.