A method of controlling fuel supply and a fuel injection apparatus

This invention relates to a method of controlling fuel supply to a diesel engine and a fuel injection apparatus, and more particularly to an arrangement for effecting a modification of steady state derived injection timing control during transient operation of the diesel engine according to the first part of claim 1 or 5.

In diesel engine control, the fuel injection timing is generally tailored to minimize noise and regulated exhaust emissions while maximizing fuel economy and power. The optimum timing for a particular family of diesel engines is empirically determined by operating representative diesel engines under steady state conditions in a dynamometer test cell and compiling a schedule of timing values for various combinations of engine speed and engine load which best satisfy the noise, emissions, economy and power criteria.

In passenger car applications, diesel engines experience transient as well as steady state operating conditions. In this respect, it is generally known that the optimum timing values for transient operation of the diesel engine are somewhat different than for steady state operation. If the timing control is operated in accordance with an empirically derived steady state schedule, the exhaust emissions, noise and power levels in transient operation will be different from those which occur under steady state operation. The primary reason for the difference in timing value requirement under transient operation is that the combustion chamber conditions (cylinder wall temperature and charge composition, in particular) have not stabilized. During an acceleration, for example, the cylinder wall temperature is cooler for a given speed and load condition than it would be under steady state operation; the reverse is true for deceleration. Since the cylinder wall temperature and charge composition directly affect the evaporation and precombustion reactions of the fuel, the timing value empirically determined under steady state operation may be inappropriate.

Notwithstanding the above, it has been difficult in the past to implement a different timing schedule during transient operation because the timing controls have not had sufficiently fast speed of response. The cylinder wall temperature instability has its main impact on the engine cylinder firings at the beginning of a transient. By the time the typical timing control mechanism could be adjusted to a suitable transient value, steady state operation would have resumed. However, increasingly stringent exhaust emission regulations have spurred the development of electrical timing controls, and currently the control technology has advanced to the point where the start and stop of each injection of fuel can be precisely controlled. In other words, it is now possible to accurately control both the fuel quantity and time of injection for each engine cylinder firing event based on current engine operating conditions. Known fuel control systems are shown in US-A 4 368 705; US-A 4 503 830; and JP-A 6 060 238.

The present invention is directed toward an arrangement for modifying the empirically derived steady state timing schedule on a cylinder-to-cylinder basis by a correction amount determined by the extent to which the demanded engine operation deviates from steady state operation along the lines indicated in US-A 4 503 830 and JP-A 6 060 238. It is broadly recognized herein that the rate of change in combustion chamber conditions experienced during transient operation (and hence the different timing requirement) is related to the cylinder firing frequency, and that the modification of the steady state derived timing values should thus be updated on a cylinder-by-cylinder basis.

It is therefore the general object of this invention to provide an improved method of controlling fuel supply to a diesel engine, and a fuel injection apparatus having an empirically determined steady state timing schedule wherein the timing values obtained from the timing schedule are modified by a correction amount updated on a cylinder-by-cylinder basis in relation to an indication of the extent to which the demanded engine operation deviates from steady state operation, wherein the correction amount is removed at a rate determined in relation to the cylinder firing frequency when the engine returns to substantially steady state operation, and wherein a timing correction amount is maintained after return to steady state which returns to zero at a rate determined by the cylinder firing frequency.

A more specific object of the invention is to provide a method and apparatus as set forth above wherein the extent to which the demanded engine operation deviates from steady state operation is determined in relation to the cylinder-to-cylinder change in requested fuel quantity.

A further object of an embodiment of this invention is to provide a method and apparatus as set forth above wherein the correction amount under relatively heavy engine acceleration serves to advance the timing value relative to the timing value from the steady state schedule to compensate for the ignition delays normally experienced under such conditions, thereby increasing the engine power output, and wherein the correction amount under light to moderate engine accelerations serves to retard the injection timing relative to the timing value from the steady state schedule, thereby decreasing the levels of engine noise and certain exhaust emissions normally experienced under such conditions. The retard correction may be eliminated or appropriately modified under engine operating conditions where drivability and/or performance take precedence.

To this end, a method of controlling fuel supply to a diesel engine, and a fuel injection apparatus, are characterised over US-A 4 503 830 by the features specified in the characterising portions of Claims 1 and 5 respectively.

The above objects are carried forward with a solenoid spill fuel injection pump of the type referred to in US-A 4 351 283. In such a fuel injection pump, the plungers which deliver fuel to the individual cylinders are normally supplied with an excess quantity of fuel from the fuel reservoir, and the pressurized fuel delivered during a pump stroke of the plunger is either injected into a cylinder or spilled via a return passage to the fuel reservoir. A solenoid valve controls the opening and closing of the spill passage such that fuel injection is initiated in a pump stroke when the spill passage is closed and terminated in the same pump stroke when the spill-pas--sage is subsequently opened.

Electronically, the fuel injection apparatus in- dudes a control unit effective for each cylinder firing event to output a fuel value indicative of the desired fuel quantity for the respective cylinder based on operator demand, and a timing value indicative of the point in the engine cycle at which supply of the fuel should be initiated. The timing value is obtained from a timing schedule comprising a compilation of timing data empirically derived during steady state operation. In addition, the electronic portion includes solenoid timer and driver means responsive to the fuel value and timing value outputted by the control unit for electrically controlling the spill valve of the fuel pump so as to initiate the supply of fuel at the point in the engine cycle indicated by the timing value and to terminate the supply of fuel at a later point determined in relation to the fuel quantity. Timing values obtained from the empirically derived steady state timing schedule are modified by a correction amount which accounts, at least in part, for the fuel timing requirement differences occasioned by transient operation of the diesel engine so that the timing value sent by the control unit yields an improvement in the resulting engine exhaust emissions, noise, power or economy. The correction amount is determined in relation to the cylinder-to-cylinder change in fuel quantity determined by the control unit and the correction amount is updated on a cylinder-by-cylinder basis so long as the transient operation persists. When the cylinder-to-cylinder change in fuel quantity indicates a return to substantially steady state operation of the diesel engine, the correction amount is reduced or phased out at a rate determined in relation to the magnitude and direction of the transient and the cylinder firing frequency. The control unit detects the return of the fuel injection amount to a substantially steady state; and maintains a timing correction amount after said return has been detected which is gradually reduced to zero at a rate determined in relation to the cylinder firing frequency.

In the following illustrated embodiment, the correction amount retards the start of fuel injection when the change in requested fuel quantity indicates light to moderate acceleration demand, and advances the start of fuel injection when the change in requested fuel quantity indicates heavy acceleration demand. As a result, noise and certain exhaust emissions are reduced during light to moderate acceleration, and engine performance is improved during heavy acceleration. In light to moderate acceleration, the timing retard permits reduced usage of exhaust gas recirculation (EGR) for emission control, and thereby reduces exhaust gas soot and engine oil contamination. At the initiation of a light to moderate acceleration, the timing retard may be delayed until EGR is at least partially cut off so that the emission reduction is attained when it is needed the most.

in the Drawings: <ul id="ul0001" list-style="none"><li>Figure 1 is a block diagram of a computerized diesel engine fuel control system for carrying out one embodiment of the method of this invention.</li><li>Figure 2 is a graph depicting the initial timing correction amount as a function of the cylinder-to-cylinder change in requested fuel quantity.</li><li>Figures 3-5 depict flow diagrams representative of program instructions executed by the computerized fuel control system of Figure 1 in carrying out the method of this invention.</li><li>Figures 6 and 7 are graphs describing the operation of the fuel control system of Figure 1 according to this invention.</li></ul>

Referring now more particularly to Figure 1, reference numeral 10 generally designates a portion of a motor vehicle power plant comprising a diesel engine 12 and a fuel pump 14 which is solenoid operated. The diesel engine 12 draws combustion air at atmospheric pressure for each cylinder through an intake manifold 16, and the fuel for the various engine cylinders is supplied by the fuel pump 14 via the fuel lines 18 and a plurality of individual fuel injectors 20. Although only two fuel lines 18 are shown, it will be understood that the fuel pump 14 supplies fuel to each fuel injector 20 via a separate fuel line 18. Following combustion, exhaust gases from the various engine cylinders are conducted away from the diesel engine 12 by the exhaust manifold and pipe assembly 22. For the purpose of controlling exhaust gas emission, an exhaust gas recirculation (EGR) valve 24 which is electrically operated is controllable via a line 26 to return a portion of the exhaust gases from the exhaust manifold and pipe assembly 22 to the intake manifold 16 via lines 28 and 30. Also included (but not shown) within the EGR valve 24, is a switch for providing an electrical signal on line 32 when the EGR valve is substantially closed -- that is, when substantially no exhaust gas is being returned to the intake manifold 16. The fuel pump 14 is driven by the diesel engine 12 and continuously circulates fuel from a fuel reservoir 34 through individual passages within the fuel pump 14 via the supply and return lines 36 and 38 respectively. Thus, the plungers (not shown) in fuel pump 14 which deliver fuel to the individual engine cylinders are supplied with an excess quantity of fuel from the fuel reservoir 34 via the supply line 36, and the pressurized fuel delivered during a pump stroke of a respective plunger is either injected into an engine cylinder via a fuel line 18 or spilled via the return line 38 to the fuel reservoir 34. A plunger-type solenoid 40 energizable via a line 42 controls a valve for opening and closing the return line 38 such that fuel injection is initiated in a pump stroke when the return line 38 is closed, and terminated in the same stroke when the return line 38 is subsequently opened. As indicated above, a pump and solenoid valve assembly of this type is disclosed in more detail in the above mentioned U.S. Patent 4,351,283.

A control unit 44, which is computerized, controls the operation of the EGR valve 24 and the plunger-type solenoid 40 in response to a number of sensed inputs according to a predetermined control algorithm. The control unit 44 thus includes an input/output circuitry for receiving and outputting the various input and control signals, and a microcomputer for processing the input signals and carrying out the control algorithm. Design details of such circuits and devices are well known to those skilled in the art of electronic controls, and are therefore not presented herein. The inputs for control unit 44 include an accelerator pedal position signal (AP) on line 46, an air temperature signal (AT) on line 48, an engine coolant temperature signal (CT) on line 50, an engine speed/crank position signal (N_e/POS) on line 52, a mass air flow signal (AF) on line 54, and an EGR valve closure signal (CS) on line 32. The above input signals are each obtained with conventional transducers which are only schematically illustrated in Figure 1. Thus, the accelerator pedal position signal (AP) on line 46 is obtained with a position transducer (PT) 56 such as a rotary potentiometer mechanically connected to an operator manipulated accelerator pedal 58; the air temperature and engine coolant temperature signals (AT, CT) on lines 48 and 50 are obtained with conventional temperature transducers 60 and 62; the engine speed/crank position signal (Ne/POS) on line 52 is obtained with a variable reluctance or similar transducer 64 responsive to the passage of teeth formed on an engine pulley or flywheel 66 that rotates with the engine crankshaft; and the mass air flow signal (AF) on line 54 is obtained from a hot wire or other conventional air flow transducer (MAFT) 68. As indicated above, the EGR valve closure signal (CS) on line 32 is obtained from a switch (not shown) within the EGR valve 24, which senses substantial closure of the EGR valve between the lines 28 and 30. A further input signal (SV) for the control unit 44 is provided on line 70 to indicate closure of the spill valve (solenoid operated) of fuel pump 14 so that the beginning of supply of fuel to the respective engine cylinders is precisely known. The input signal (SV) may be obtained, as is generally known in the art of solenoid control, by monitoring the solenoid voltage and detecting the abrupt voltage rise due to the inductance change upon closure of the spill valve.

The primary outputs of the control unit 44 comprise the EGR signal on line 26 for controlling the operation of the EGR valve 24, the TDC signal on line 72, the TIMING signal on line 74, and the FUEL signal on line 76. Also provided is a closure signal SC from the spill valve on line 77 and an engine speed related CLOCK signal on line 78. The closure signal (SC) on line 77 is derived from the input signal (SV) on line 70 as described above and becomes active whenever closure of the spill valve of fuel pump 14 is detected. The CLOCK signal on line 78 may be developed directly by the control unit 44 or by an external hardware circuit such as a phase- locked loop, and essentially comprises a series of relatively high frequency pulses developed in synchronism with the engine speed/crank position signal (N_e/POS) on line 52, but at a multiple frequency thereof. The development of the EGR signal on line 26 is described in more detail hereinafter in reference to Figure 4 and the development of the FUEL, TIMING and TDC signals are described in more detail hereinafter in reference to Figures 3 and 5.

As explained below, the TDC, CLOCK and closure (SC) signals are used in conjunction with the FUEL and TIMING signals to develop an energization pulse for the plunger-type solenoid 40 on line 42 for effecting the injection of fuel into the various cylinders of diesel engine 12. The initiation of the fuel supply is determined from the TIMING signal and the duration of the fuel supply is determined from the FUEL signal. More specifically, the above signals are applied to the programmable timers 80 and 82. Each of the programmable timers 80 and 82 comprises a latch, or register, and a downcounter. Digital information stored in the latch may be gated or transferred into the downcounter and decremented at a desired clock frequency. When the count in the downcounter is decremented to zero, the counter issues an output signal which is used to gate or trigger another device. In the illustrated embodiment, the TIMING signal on line 74 is applied to the latch of programmable timer 80 and the FUEL signal on line 76 is applied to the latch of programmable timer 82. The CLOCK signal on line 78 is applied to the clock inputs of both programmable timers 80 and 82. The TDC signal on line 72 is applied to the latch gate input of both programmable timers 80 and 82 and to the count enable input of the programmable timer 80.

In operation, TIMING and FUEL signals outputted from the control unit 44 on lines 74 and 76 are stored in the latch of the programmable timers 80 and 82. When the control unit 44 issues a TDC signal on line 72, the TIMING and FUEL signals stored in the latches are transferred to the respective downcounters, and the count in the counter of programmable timer 80 is decremented at the frequency of the CLOCK signal on line 78. When the count is decremented to zero an output pulse is issued on line 84 which signals a solenoid driver 86 to begin energization of the plunger-type solenoid 40 for initiating the supply of fuel to diesel engine 12. However, a finite period of time is required for the spill valve to close, and decrementing of the count stored in programmable timer 82 is only enabled when the closure signal (SC) on line 77 becomes active. When the count stored in the counter of programmable timer 82 is decremented to zero, an output pulse is issued on line 88 which signals the solenoid driver 86 to terminate the energization of the plunger-type solenoid 40 for terminating the supply of fuel to diesel engine 12. In such mechanization, it will be recognized that the Information contained in the TIMING signal on line 74 is actually representative of the time delay from top dead centre (TDC) to start of injection for a respective engine cylinder.

The control unit 44 develops the FUEL signal primarily as a function of operator demand as determined by the accelerator pedal position signal (AP), and the TIMING signal, in turn, is determined as a function of the desired fuel quantity and engine speed (N_e). More particularly, the TIMING signal is formed as a combined function of an empirically derived steady state timing value SSTIME and a correction amount CURMOD which accounts, at least in part, for timing requirement differences occasioned by transient operation of the diesel engine. The correction amount is determined in relation to the cylinder-to-cylinder change in requested fuel quantity and is updated on a cylinder-by-cylinder basis so long as the transient operation persists. When the change in requrested fuel quantity indicates light to moderate acceleration demand, the correction amount serves to retard the start of fuel injection to decrease the levels of certain exhaust emissions and noise. When the change in requested fuel quantity indicates heavy acceleration demand, the correction amount serves to advance the start of fuel injection to improve engine performance. When the cylinder-to-cylinder change in fuel quantity indicates a return to substantially steady state operation of the diesel engine, the correction amount is reduced or phased out at a rate determined in relation to the cylinder firing frequency, and the magnitude and direction of the transient.

The initial timing correction as a function of the cylinder-to-cylinder change in requested fuel quantity is graphically depicted in Figure 2 by the trace 90. In such representation, the timing correction (TT MOD) is given in degrees, a positive correction indicating an advance of the start of fuel injection and a negative correction indicating a retard of the start of fuel injection. The cylinder-to-cylinder change in fuel quantity ( A FUEUCYL) is given in mm³ and is defined such that a positive change indicates acceleration demand and a negative change indicates deceleration demand. The trace 90 represents the illustrated embodiment wherein light to moderate acceleration demand is indicated by a cylinder-to-cylinder change in requested fuel quantity of about 1.2 - 8.8 mm³, while heavy acceleration demand is indicated by a cylinder-to-cylinder change in requested fuel quantity of about 8.8 mm³ or larger. During heavy acceleration demand, the correction amount can effect an advance of the injection timing by up to approximately 3.0 degrees of crankshaft rotation; during light to moderate acceleration the correction amount can effect a retard of the injection timing by up to approximately 2.6 degrees of crankshaft rotation.

Figures 3-5 depict flow diagrams representative of program instructions executed by the control unit 44 for controlling the supply of fuel in accordance with the teachings of this invention. Figure 3 shows a main program periodically executed to read the various input values, control the operation of the EGR valve 24 and determine a steady state timing value (SSTIME) from an empirically derived steady state timing schedule. Figure 4 sets forth the EGR control of the main program in more detail. Figure 5 shows an interrupt routine which is executed in response to the passage of each engine piston through its top dead centre (TDC) position (as determined from the engine speed/crank position signal Ne/POS) and which operates to output FUEL and TIMING signals for upcoming engine cylinder firing events.

Referring now more particularly to Figures 3 and 4, the reference numeral 100 designates a series of program instructions for initializing the various input values, registers, timers, etc., within the control unit 44. Such program instructions are executed at the initiation of each period of vehicle operation and serve to ensure repeatable and reliable operation of the control unit 44. Following such initialization procedures, the instruction block 102 is executed to read and process information associated with the various input signals identified in Figure 1. While most of the information is in the required form and is simply stored for later use, certain information, including the engine speed/crank position signal (Ne/POS) on line 52 and the input signal SV on line 70 is processed to provide the desired information. Thus, the engine speed/crank position signal (N_e/POS) is processed to identify both the engine speed N_e and the occurrence of each TDC, and the input signal (SV) is processed as indicated above to identify substantial closure of the spill valve of the fuel pump 14. Thereafter, the instruction block 104 is executed to determine a steady state timing value (SSTIME) based on empirically derived timing data compiled during engine testing under steady state engine speed and load conditions. As indicated above, such data is compiled in a dynamometer test cell and timing values are chosen which best satisfy noise, emission, economy and power criteria for the particular steady state conditions under test. Thus, the steady state timing value (SSTIME) determined at instruction block 104 is addressed as a function of engine speed signal Ne and an indication of engine load such as the accelerator pedal position signal AP. Then, the instruction block 106 is executed to call the EGR subroutine depicted in Figure 4 for determining and outputting an EGR signal on line 26 for controlling the operation of the EGR valve 24. The instruction block 108 represents other engine control functions performed by the control unit 44, such as temperature and altitude compensation, idle control, and the development of closure signal (SC) on line 77 if appropriate. Upon executing such other engine control functions, the instruction block 102 is reexecuted to read fresh values of the various inputs as indicated by the flow diagram return line 110.

Referring now to the EGR subroutine depicted in Figure 4, the instruction block 112 represents program instructions for determining the desired air/fuel (A/F) ratio as a function of engine speed N_e and the requested fuel amount. Exhaust gas recirculation is used primarily for exhaust emission control and the desired air/fuel ratio values are chosen for different combinations of engine speed and load to result in acceptable emission levels. Such data is compiled in much the same way as the fuel injection timing values described above. As also indicated at the instruction block 112, the scheduled air/fuel ratio may be modified by various parameters to account for changes in altitude and temperature. Based on the desired air/fuel ratio and the requested fuel quantity (determined as a function of the accelerator pedal position), the desired mass air flow value DESAIR is then calculated as indicated at the instruction block 114. Then, as indicated at the instruction block 116, the desired mass air flow value DESAIR and the actual mass air flow value ACTAIR determined from the mass air flow signal AF on line 54 are differenced to determine the air flow error. Based on the error, a pulse-width-modulation signal for the EGR valve 24 is then determined as indicated at instruction block 118 for causing the actual air flow through the intake manifold 16 to correspond with the desired mass air flow value DESAIR. If the error determined at instruction block 116 indicates that the actual mass air flow value ACTAIR is too high, for example, the pulse-width-modulation signal for the EGR valve 24 is altered to increase the amount of exhaust gases returned to the intake manifold 16, thereby effectively displacing some of the atmospheric air therein. Conversely, if the actual mass air flow value ACTAIR is too low, the pulse-width-modulation signal for the EGR valve 24 is adjusted to lessen the amount of exhaust gas returned to the intake manifold 16. In steady state engine operation, the pulse-width-modulation signal adjustments are made in the manner described above to maintain the actual air/fuel ratio at or about the desired air/fuel ratio for emission controls. However, during transient operation of the diesel engine 12, the actual air/fuel ratio typically changes by a large amount, thereby causing the control unit 44 to fully open or fully close the EGR valve 24. In an acceleration, for example, the actual air/fuel ratio suddenly drops causing the desired mass air flow value DESAIR to greatly exceed the actual mass air flow value ACTAIR. At such point, the pulse-width-modulation signal is adjusted to fully close the EGR valve 24 and the EGR valve 24 remains closed until substantially steady state operation of the diesel engine is resumed. As will be described later, the control functions of this invention interact with the operation of the EGR valve 24 according to the illustrated embodiment to provide a reduction of the exhaust emissions when such reduction is most needed -- upon substantial closure of the EGR valve 24 in the course of an acceleration.

Figures 5A and 5B depict the flow diagram of an interrupt routine executed by the control unit 44 at each occurrence of top dead centre in diesel engine 12. Thus, if the diesel engine 12 has eight cylinders, the routine depicted in Figures 5A and 5B will be executed by the control unit 44 -four times per engine revolution. As indicated above, the information concerning the occurrence of each top dead centre in diesel engine 12 is obtained from the engine speed/crank position signal N_e/POS on line 52. Essentially, the routine depicted in Figures 5A and 5B performs two functions: updating the value of the timing correction amount based on the computed change in requested fuel amount, and sending new TIMING and FUEL signals to the programmable timers 80 and 82.

Initially, the instruction block 120 is executed to output a TDC signal on line 72 for initializing the programmable timers 80 and 82 for the next injection of fuel. Then as indicated at instruction block 122, the requested fuel quantity is determined as a function of the accelerator pedal position AP and engine speed N_e. Then as indicated at instruction block 124, the cylinder-to-cylinder change in fuel quantity ( λ fuel) is calculated according to the difference between the fuel quantity determined at instruction block 122 (NEW FUEL) and the previously determined fuel quantity (OLD FUEL). In addition, the term OLD FUEL is set equal to the term NEW FUEL so that in the next execution of the routine, the cylinder-to-cylinder change in fuel quantity may be accurately computed.

The decision blocks 126 - 130 are then executed to determine if the timing correction amount should be updated. If so the flow diagram branch 132 is executed; if not, the flow diagram branch 134 is executed. At decision block 126, the accelerator pedal position and engine speed/crank position signals are tested to determine if the diesel engine 12 is at idle. If so, the diesel engine 12 is at or near steady state operation regardless of the cylinder-to-cylinder change in requested fuel quantity, and the flow diagram branch 134 is executed as indicated by the flow lines 136 and 138. If not, the decision block 128 is executed to determine if the change in fuel quantity (A fuel) is greater than a reference change (_Aref) above which transient operation of the diesel engine 12 is indicated. In the illustrated embodiment, the reference change Aref is set at approximately 1.2 mm³ as indicated in Figure 2. If the computed change in fuel quantity is less than or equal to the reference change, the diesel engine 12 is at or near steady state operation, and the flow diagram branch 134 is executed as indicated by the flow line 138. If the computed change in requested fuel quantity is greater than the reference change, the decision block 130 is executed to determine if the engine speed N_e is in the control range. As indicated above, this invention recognizes that the changes in the cylinder combustion chamber conditions, and hence the duration of transient operation of the diesel engine 12, decreases with increasing engine firing frequency. As a result, there is an engine speed Ne above which transient corrections to the steady state timing value SSTIME cannot be effectively made and may not be beneficial to engine operation. In the illustrated embodiment, it has been found that the effective control range extends from engine idle speeds to approximately 2200 rpm. If the engine speed N_e is not within the control range, the flow diagram branch 134 is executed; if the engine speed Ne is within the control range, the flow diagram branch 132 is executed to update the correction amount CURMOD. Thus, the flow diagram branch 132 is executed at the initiation of a period of transient operation of the diesel engine 12 and continues to be executed so long as the transient condition persists. When further adjustment of the correction amount CURMOD is inappropriate, or the change in requested fuel amount indicates a return to substantially steady state operation of the diesel engine 12, the flow diagram branch 134 is executed to reduce the previously developed correction amount at a controlled rate.

Three program variables, TTMOD, NEWMOD and BIGMOD are used in the updating of the correction amount CURMOD. TTMOD represents the initial correction amount as determined from a look-up table such as represented by the trace 90 in Figure 2. The terms NEWMOD and BIGMOD are used in the case of a timing retard as will be explained below to identify the proper retard value and to delay the initiation of the retard until the EGR valve 24 is substantially closed.

Referring firstly to the flow diagram branch 132, the instruction block 146 is executed to set the NEWMOD flag, indicating that a modification of the correction amount CURMOD is being made. Then, the instruction block 148 is executed to look-up a transient timing modification TTMOD as a function of the change in requested change in fuel quantity (A fuel) as indicated by the trace 90 in Figure 2. The decision block 150 is then executed to determine if the transient timing modification TTMOD is positive in sign indicating an injection timing advance, or negative in sign indicating an injection timing retard. If the transient timing modification TTMOD is positive, the instruction block 152 is executed as indicated by the circled numeral 1 to set the program variables CURMOD and NEWMOD equal to the new timing correction value of transient timing modification TTMOD so that a modification of the steady state timing value SSTIME can be made without delay. If the timing correction amount of TTMOD is negative in sign, the decision block 154 is executed as indicated by the circled numeral 2 to determine if the magnitude of TTMOD is greater than the term BIGMOD. If so, the terms NEWMOD and BIGMOD are set equal to TTMOD as indicated at the instruction block 156. If the magnitude of TTMOD is less than or equal to the term BIGMOD, the execution of instruction block 156 is skipped, as indicated by the flow diagram line 157. As mentioned above, retarding of the injection timing due to a sensed transient operation of the diesel engine 12 is delayed until it is determined that the EGR valve 24 is substantially closed; until such time, the program variable NEWMOD saves the desired correction amount as determined at decision block 154.

Following the execution of the flow diagram branch 132, the instruction block 158 is executed to determine a TIMING signal according to the sum of the steady state timing value SSTIME and the correction amount CURMOD. In addition, the TIMING and FUEL signals are outputted to the latches of programmable timers 80 and 82. At such point, execution of the interrupt routine is complete and the control unit 44 is returned to continue execution of the main program depicted in Figure 3.

Referring now to the interrupt routine of flow diagram branch 134, the decision block 162 is executed to determine if the NEWMOD flag is set. If so, this is the first execution of the flow diagram branch (i.e., a new modification NEWMOD of the correction amount has just been made), and the decision block 164 is executed to determine if the current transient timing modification TTMOD is positive in sign. If not, a retardation of the injection timing is in order and the decision block 166 is executed to determine if the EGR valve 24 is substantially closed. As indicated above, this decision.is made in the illustrated embodiment by a switch internal to the EGR valve 24 which closes upon substantial closure of the EGR valve 24. If the EGR valve 24 is closed, the instruction block 168 is executed to set the correction amount CURMOD equal to the term NEWMOD so as to effect the timing correction without further delay, and to reset the term BIGMOD to zero. If it is determined at the decision block 164 that the transient timing modification TTMOD is positive in sign, an advance of the injection timing has already been initiated, and the execution of the decision block 166 and the instruction block 168 is skipped as indicated by the flow line 170. Then, as indicated by the circled numeral 3, the instruction blocks 172 and 174 are executed to calculate a reduction factor FACTOR for the correction amount CURMOD and to reset the NEWMOD flag. As indicated at the instruction block 172, the reduction factor FACTOR is determined as a function of the correction amount CURMOD and a constant K. If it is determined at the decision block 166 that the EGR valve has not yet closed, the execution of the instruction blocks 168, 172 and 174 is skipped as indicated by the flow line 176. If it is determined at the decision block 162 that the NEWMOD flag is not set, the execution of the blocks 164-174 is skipped as indicated by the flow lines 178 and 176 and the circled numeral 4.

According to the illustrated embodiment, a reduction of the correction amount CURMOD is made on a periodic basis in synchronism with the engine rotation by reducing the correction amount CURMOD by the reduction factor FACTOR determined at instruction block 172 at every Nth execution of the TDC interrupt routine. The number N may be any integer greater than or equal to 1; in the preferred embodiment, N has a value of 8 in the case of an injection advance and 16 in the case of an injection retard. Accordingly, reductions of the correction amount CURMOD are made every second engine revolution in the case of an advance and every fourth engine revolution in the case of a retard. The rationale for the difference is that the beneficial effect of injection timing retardation in a light to moderate engine acceleration extends over a relatively large number of engine cylinder firings, while the beneficial effect of injection timing advance in heavy engine acceleration extends over a relatively short duration. As set forth in Figure 5B, the reduction of the correction amount CURMOD is implemented with an Nth cylinder counter (Nth CYL CTR) which is used to count executions of the TDC interrupt routine. Initially, the cylinder counter has a value of zero and the decision block 180 is answered in the affirmative. At such point, the instruction block 182 is executed to effect a reduction of the correction amount CURMOD by subtracting the reduction factor FACTOR from the correction amount CURMOD. Then, the decision block 184 and the instruction blocks 186 and 188 are executed to reset the Nth cylinder counter to a predetermined setting. If the sign of current transient timing modification TTMOD indicates an injection timing advance, the decision block 184 is answered in the affirmative, and the instruction block 186 is executed to reset the counter to 8 as indicated by the term RATE 1. If the sign of TTMOD indicates an injection timing retard, decision block 184 is answered in the negative, and the instruction block 188 is executed to reset the Nth cylinder counter to 16 as indicated by the term RATE 2. If it is determined at the decision block 180 that the Nth cylinder counter is not zero, the instruction block 190 is executed to decrement the counter.

The operation of the control unit 44 in executing the program instructions depicted by the flow diagrams of Figures 3-5 will now be described in reference to the graphs of Figures 6 and 7. In each case, the various graphs are depicted on a common time base graduated in terms of engine cylinder top dead centres. Thus, the term Co represents the cylinder top dead centre (TDC) which occurs at the beginning of a change in operator demand; the term C₁ represents the next engine cylinder TDC, and so on. In Figure 6, the change in operator demand is representative of heavy engine acceleration demand; in Figure 7, the change in operator demand is representative of light to moderate engine acceleration demand. In each case, the graphs depict requested fuel quantity, change in requested fuel quantity (A fuel), engine speed, SSTIME, CURMOD, and the TIMING signal. In Figure 7, an additional graph (Graph D) depicts the EGR valve closure signal (CS).

Referring now more particularly to Figure 6, the trace 200 of Graph A represents the requested fuel quantity with respect to time for a wide-open throttle engine acceleration beginning at the cylinder TDC designated by C_o. Although the rate of increase of fuel actually delivered to the diesel engine 12 may be modified somewhat by a fuel limiting device, as designated by the dashed line 202, the timing correction amount of this invention is based, as explained above, on changes in the requested fuel quantity. Changes in the requested fuel quantity determined by the TDC interrupt routine of Figures 5A and 5B at the times of cylinder TDCs C₁ and C₂ are graphically depicted by the heavy traces 204 and 206, respectively. At cylinder TDC C₃ and thereafter, the change in requested fuel quantity is zero or negligibly small. As seen by the trace 208 of Graph B, the eventual delivery of the requested fuel quantity to the engine cylinders results in a sharp engine speed increase from the idle value Nj.

The Graphs C, D, and E of Figure 6 depict injection timing values in degrees of crank angle advance with respect to cylinder TDC position. In Graph C, the trace 210 depicts the injection timing values obtained from the empirically derived steady state timing schedule as a function of the engine speed Ne and load AP as described above in reference to the main program of Figure 3. In Graph D, the trace 212 depicts the values of the timing correction amount CURMOD derived in the course of the acceleration. In Graph E, the trace 214 depicts the sum of the steady state timing (trace 210) and the timing correction amount (trace 212), and therefore represents the TIMING signal output from the control unit 44. The trace 210 is repeated in broken lines in Graph E to show the convergence of the steady state timing value SSTIME and the TIMING signal as-the correction amount CURMOD is reduced to zero following a return of the diesel engine 12 to substantially steady state operating conditions. The value of the steady state timing value SSTIME is determined at a relatively high rate by the main program of Figure 3, and is therefore depicted by a smooth trace in Graph C. On the other hand, the correction amount CURMOD and the TIMING signal are updated only during execution of the TDC interrupt routine of Figures 5A and 5B; hence the discontinuous form of the traces 212 and 214 in Graphs D and E.

When the engine crankshaft passes through the TDC position Co, the control unit 44 executes the TDC interrupt routine of Figures 5A and 5B to issue a TDC signal on line 72 to load the counters of programmable timers 80 and 82 with TIMING and FUEL signals for the next firing event (C₁). Since the change in fuel quantity is less than the reference change (Aref), the timing correction amount CURMOD remains at its initialized value of zero (as seen in Graph D), and new TIMING and FUEL signals for the following firing event (C₂) are sent to the latches of the programmable timers 80 and 82. At the TDC position C₁, the control unit 44 reexe- cutes the TDC interrupt routine to issue another TDC signal on line 72 to load the counters of programmable timers 80 and 82 with TIMING and FUEL signals for the next firing event (C₂), to update the value of correction amount CURMOD (trace 212) in accordance with the sensed change in requested fuel quantity (trace 204), and to output new TIMING (trace 214) and FUEL signals for the following firing event (C₃). Thus, at each TDC position (C_n) the counters of programmable timers 80 and 82 are loaded with TIMING and FUEL signals for the next firing event (C_n 1) the value of correction amount CURMOD is updated in accordance with the sensed change in requested fuel quantity, if any, and new TIMING-and FUEL signals are sent for the following firing event (C_{n + 2}). However, at TDC position (Cs) and thereafter, the change in requested fuel quantity is less than the reference change Aref, and the flow diagram branch 134 is executed to progressively reduce the timing correction amount CURMOD to zero. As set forth in Figures 5B, the reduction factor FACTOR is determined in the first execution of the flow diagram branch 134 - at the cylinder TDC position (C₃) for the example illustrated in Figure 6. After the timing correction amount CURMOD is reduced to zero (as depicted by the joining of the traces 210 and 214 in graph E), at or about the cylinder TDC position (C₄₆), the TIMING signal sent to the latch of programmable timer 80 is once again determined solely in accordance with the steady state timing values SSTIME obtained from the empirically derived steady state timing schedule.

Referring now more particularly to Figure 7, the trace 220 of Graph A depicts the requested fuel quantity with respect to time for a light to moderate acceleration of the diesel engine 12. The trace 222 of Graph B depicts the corresponding increase in engine speed N_e from an idle value N_i. The changes in requested fuel quantity ( A fuel) determined by successive executions of the TDC interrupt routine at the TDC positions 0₁, C₂ and C₃ are given by the heavy traces 224, 226 and 228, respectively. As shown in Graph D, the EGR valve 24 is open for returning exhaust gases from the exhaust manifold and pipe assembly 22 to the intake manifold 16 at the initiation of the engine transient at the time of TDC position Co. Due to the sudden change in fuel quantity, the control unit 44 operates to close the EGR valve 24 as described above in reference to Figure 4 in order to maintain the desired air/fuel ratio. As indicated in Graph D, such closure requires an interval of approximately 200 milliseconds. Retardation of the steady state timing value SSTIME due to a sensed engine transient is delayed until substantial closure of the EGR valve 24 so as to effect an improvement in the engine exhaust emissions when it is most needed. While the retardation could be effected immediately (as with injection advances) to effect a more immediate reduction of the engine exhaust emissions, such retard would cause a corresponding decrease in fuel economy as well. The correction amount CURMOD applied upon closure of the EGR valve 24 is based on the largest value of TTMOD determined during the delay period prior to the return to substantially steady state operation of the diesel engine 12. The largest transient timing modification TTMOD is determined with the term BIGMOD at the decision block 154 and the instruction block 156. The largest modification is used because it provides a better indication of the extent of the light to moderate acceleration transient than smaller modifications which occur a few cylinders later. For the period of operation depicted by the traces of Figure 7, the correction amount CURMOD would be based on the change in fuel quantity ( A fuel) depicted by the trace 224 and determined at the time of TDC position C₁. Prior to the closure of the EGR valve 24 just prior to the TDC position C₈, the TIMING signals sent by the control unit 44 are in accordance with the empirically derived steady state timing schedule as depicted by the trace 238 of Graph F.

As described above in reference to Figures 5A and 5B, the timing correction amount CURMOD to be introduced upon substantial closure of the EGR valve 24 is stored by the term NEWMOD. When the TDC interrupt routine is executed at the TDC position C₈, the correction amount stored in the term NEWMOD is transferred to the term CURMOD (Graph E), the reduction factor FACTOR is computed as a function of the correction amount magnitude, and updated TIMING (Graph F) and FUEL signals are sent to the latches of programmable timers 80 and 82 for the upcoming firing event (C₁₀). Also, as set forth in reference to the example of Figure 6, the execution of the TDC interrupt at each TDC position Cn operates to load the counters of programmable timers 80 and 82 with TIMING and FUEL signals for the next firing event (Cn _{+ 1}). Further executions of the TDC interrupt routine at the TDC positions C₉ and thereafter, operate to progressively reduce the timing correction amount CURMOD as seen in the Graphs E and F. After the correction -amount CURMOD is reduced to zero at or about the cylinder TDC position C₆₅ (not shown), the TIMING signal outputted to the latch of programmable timer 80 is determined solely in accordance with the steady state timing values SSTIME obtained from the empirically derived steady state timing schedule.

The above examples of Figures 6 and 7 illustrate the diverse operation of the timing control of this invention in periods of transient operation of the diesel engine 12. In the usual situation, the transient operation will be characterized as light to moderate and the transient timing correction of this invention serves to retard the timing in respect to that obtained from the steady state timing schedule (at least following closure of the EGR valve 24) to effect a reduction in the level of engine noise and exhaust gas emissions. Such operation generally occurs at relatively low engine speed and the correction amount CURMOD is phased out or reduced over a relatively long period of time. However, when a period of heavy acceleration is demanded by the operator of the vehicle, engine performance considerations take precedence and the injection timing is immediately advanced to offset the injection retard which is normally experienced, to thereby improve the engine performance. Such operation is normally followed by relatively high engine speed and the timing correction is phased out or reduced over a relatively short time interval. In implementing the present invention in a diesel engine fuel control system in a motor vehicle, the engine performance under heavy demanded engine acceleration was increased, and the engine noise and certain exhaust emission constituents under light to moderate demanded engine acceleration were reduced.

Various changes can be made to the above described embodiment. For example, the timing modifications depicted by the trace 90 in Figure 2 may be varied to suit the particular engine or family of engines for which the timing control functions of this invention are being implemented. In addition, and as mentioned above, the retardation of injection timing with light to moderate engine acceleration may be implemented immediately without regard to closure of the EGR valve. Alternately, the implementation of the retard may be delayed for a predetermined period of time sufficient to ensure at least partial closure of the EGR valve 24, thereby eliminating the need for separately sensing closure of the valve. Likewise, the various constants and timing modifications referred to in the description of the illustrated embodiment may be altered when the control is applied to a different family of engines.