EP0170891B1 - Method and apparatus for specifically controlling each cylinder group in a multicylinder engine - Google Patents

Description

State of the art

The invention is based on a method and a device for controlling a multi-cylinder internal combustion engine and a device for carrying out the method according to the preamble of the main claim and the secondary claim. Such a device has already been described in DE-A-2941 977 and in parallel application US-A-43 42 097. To optimize the delivered torque of an internal combustion engine or the specific fuel consumption, a test signal generator for varying the metered fuel quantity and a sensor for detecting the size to be optimized are used and, based on a torque signal, the maximum power or the minimum specific fuel consumption is determined depending on the load area of the internal combustion engine. From US-A-43 66 793 a multi-cylinder internal combustion engine is also known, in which, however, a misfire in a cylinder is first waited for before the fuel supply to this cylinder is increased and the fuel supply to the regularly operating cylinders is reduced.

Although such devices have proven themselves very well in practical operation, further developments and improvements are still possible, which come into play in particular with regard to the stricter exhaust gas legislation and the efforts to reduce the gasoline consumption of the internal combustion engines.

For example, studies have shown that the individual cylinders of an internal combustion engine are normally operated with different air-fuel ratios. The reasons for this are a. to look for in a different intake manifold guide and in not completely identical injection valves.

It is therefore an object of the invention to precisely measure the control variables that are required for each individual cylinder of the internal combustion engine in order to work at the optimum efficiency for the relevant operating point. This object is achieved with the features of the independent method or device claim.

Advantages of the invention

The main advantage of the invention is a reduced fuel consumption of the internal combustion engine while maintaining good exhaust gas values despite larger permissible tolerances in the injection valves and in the filling of the individual cylinders. Furthermore, it proves to be advantageous that, according to the invention, that lambda is set for each cylinder at which the cylinder in question operates at its optimum efficiency. For a given engine design and for given operating conditions, the engine can thus be operated in the range of the theoretically minimal fuel consumption.

Further advantages of the invention result in connection with the claims, from the following description and the drawing.

drawing

Embodiments of the invention are shown in the drawing and are described and explained in more detail below. 1 and 2 show diagrams of an arbitrarily assumed torque curve of the cylinders of an internal combustion engine to explain the method according to the invention, FIG. 1 shows an embodiment of the device for carrying out the method, FIG. 4 shows a flow diagram to explain the mode of operation of the embodiment of FIG. 3. FIG. 5 shows a time diagram FIG. 6 shows a time diagram to explain the application of the method to a multi-cylinder internal combustion engine with only a single injection valve.

Description of the embodiments

The devices for optimizing the operating parameters of an internal combustion engine that do not have a cylinder-specific effect will not be discussed further below, since their mode of operation is described, for example, in DE-OS 28 47 021 (UK patent application 20 34 930A), SAE paper 72 02 54 or also the US-PS 4064846 is sufficiently explained. In general, these methods are based on extreme value control, in which an input variable of the internal combustion engine is varied, for example, periodically. The reaction of the internal combustion engine to this periodic variation is monitored via an output variable of the internal combustion engine, for example the torque. In accordance with this monitoring result, an input variable of the internal combustion engine is adjusted until the variation in the output variable has dropped to a minimum. In all known methods, however, it is not taken into account that a different operating mixture is generally made available to each individual cylinder of the internal combustion engine. The variations of the company mix for the individual For example, cylinders can be attributed to different fillings or different injection quantities.

mit

• a gleich Drosselklappenstellung bzw. angesaugte Luftmenge und
• T,, T₂ gleich Einzelzylindereinspritzzeiten angesetzt.

The essence of the invention will be explained in more detail using the example of a 2-cylinder internal combustion engine. In Figure 1a, the torque curve different adopted for the two cylinders M1 and M ₂ is for this purpose the individual cylinders in dependence upon the throttle position α and plotted as a function of the amount of intake air. In order to simplify the numerical treatment of the problem, a parabolic curve according to the torque was arbitrarily used

With

• a equals throttle valve position or intake air quantity and
• T ,, T ₂ equal to single cylinder injection times.

For the sake of clarity, the total torque as the sum of the individual torques is not plotted in all the other figures, but rather the total torque divided by the number of cylinders. The injection time is included in these curves as a parameter. The special choice of the course of the single-cylinder torques simulates that cylinder 1 has a larger filling than cylinder 2. This follows from the fact that T ₁ = T ₂ = 7 [w. E.] the torque curve of the first cylinder already with a throttle valve position α = 4 [w. E.] compared to a = 6 [w. E.] reached its maximum torque in the second cylinder. Due to this different filling of the individual cylinders, the total torque based on the number of cylinders (1/2 ΣM) with a throttle valve position a = 5 [w. E.] do not reach the values of the single cylinder torques. In order to optimize the single-cylinder torque curve or the cylinder-specific efficiency, it is now proposed according to the invention to wobble the injection quantity for the two cylinders of the internal combustion engine in opposite directions with a constant intake air quantity in such a way that the total injection time or quantity of all cylinders is kept constant. A comparison of the phase position of the wobble signal for the injection times with the signal of a sensor for the torque of the internal combustion engine provides the cylinder-specific injection times to achieve the maximum torque of the internal combustion engine. On the basis of the results of the phase comparison, the cylinder-specific injection quantities are varied in opposite directions until the torque variations assume a minimum due to the wobbling of the injection times. The essential boundary condition of this method is to keep the sum of the individual injection times constant so that the operating point of the internal combustion engine as well as the average exhaust gas composition are preserved. The results of such a cylinder-specific optimization process are plotted in FIG. 1b. According to the optimization method, cylinder 1, which has a higher degree of filling than cylinder 2 with the same throttle valve position α, is given a higher fuel quantity corresponding to an injection time T = 8 [w. E.] supplied, while the cylinder 2 with injection times T ₂ = 6 [w. E.] is operated. The sum of the injection times and thus the amount of fuel supplied thus remained unchanged, while the total cylinder-related torque (1/2 ΣM) increased by 25% from 4 [w. E.] to 5 [w. E.] has increased. This is equivalent to the fact that the efficiency of the internal combustion engine would be increased by 25%. To illustrate the relationships, the course of the cylinder-weighted total torque as a function of the injection time T is plotted in FIG. The throttle valve position α serves as a parameter, with α 5 [w. E.] assumes. The injection time T ₂ is implicitly included in the total torque function on the condition that the sum of the injection time T and T _{2 should form} a constant (here constant = 14). From this Figure 1c it can be seen that the two-cylinder internal combustion engine then delivers an optimal torque and is thus operated at the maximum efficiency when the injection time T, the value 8 [w. E.] with a total injection time T. and T ₂ of 14 [w. E.] with a throttle valve position α = 5 [w. E.] assumes. This process is now repeated for each throttle valve position.

The method for a four-cylinder internal combustion engine will be explained with reference to FIG. In FIG. 2a, the single-cylinder torque curves and the cylinder-related total torque curve are plotted analogously to FIG. 1a. It was assumed that cylinders 1, 2 and 3 have the same filling and, accordingly, the same torque curves M ₁₂₃ . By contrast, cylinder 4 works with a low degree of filling; so that the maximum torque is only reached with larger throttle valve positions a or air volumes. The arbitrarily assumed single-cylinder torque curves should satisfy the following equations:

Anm. : w. E. ⁼ willkürliche Einheiten

Note: w. E. ⁼ arbitrary units

The optimization process now proceeds in such a way that the injection times or quantities (T, + T ₂ ) for cylinders 1 and 2 are wobbled in opposite directions to the injection times (T ₃ + T ₄ ) for cylinders 3 and 4. The boundary condition that the sum of all four injection times should remain unchanged must also be observed here. The wobbling of the injection quantity in connection with a phase analysis of the output signal for the torque or the speed of the internal combustion engine serves to determine the direction of the required adjustment of the mean values of (T, + T ₂ ) and (T ₃ + T ₄ ) such that there is a maximum torque, ie the torque modulation goes to zero. The determined ratio values of the injection quantities T ₁ , T ₂ and T ₃ , T ₄ are first stored. The process described is now repeated in the same way for two further cylinder groups or cylinders. By alternately combining the cylinders or cylinder groups and repeating the optimization process, the absolute torque maximum or the absolute minimum specific fuel consumption is set for the relevant operating point of the internal combustion engine after a few steps. The result can, for example, be recorded in a learning map. It is therefore necessary to change the cylinder groups or individual cylinders since only the ratio of two fuel injection quantities can be determined by each individual optimization process. In the case of a four-cylinder internal combustion engine, four unknowns, namely four injection times, have to be determined. It is therefore necessary to repeat the optimization process three times, so that three different injection time ratios are obtained for different cylinders or groups of cylinders. The fourth condition is that the sum of all injection times must have a constant value. To determine the four unknowns, the four injection times for each individual cylinder, four equations (three injection time ratios, total T = constant) are available so that the calculation of the single cylinder injection times can be carried out without any problems. If it turns out in the respective special case that there is a coupling between the variables, i.e. there are not four independent variables, an alternative determination of the cylinder-specific injection times is appropriate. Repeating the described optimization process several times then delivers the same result after a few passes. Such iterative methods for solving coupled systems of equations are well known per se, so that the person skilled in the art can also easily carry out the method according to the invention iteratively.

FIG. 2b shows the result of the optimization process, namely injection times T, = T ₂ = T ₃ = 7.5 [w. E.] and T ₄ = 5.5 [w. E.] for a throttle valve position a = 4.5 [w. E.] Also in this example there is an approximately 20% increase in the average total torque per cylinder. In FIG. 2c, the dependence of the mean total torque per cylinder on the injection time T, for a specific throttle valve position a = 4.5 [w. E.] applied. The injection times T _z , T ₃ and T ₄ are over the conditions T, = T ₂ = T ₃ and

T = constant implicitly included. The extreme value of this curve is an injection time T. = 7.5 [w. E.], so that the optimal injection time values of FIG. 2b, as is to be expected, are confirmed.

For an internal combustion engine with a number of cylinders not considered here, the individual method steps are to be applied analogously, the only change being the number of steps and the change of oppositely wobbled cylinders or cylinder groups.

FIG. 3 shows the circuit structure of a device for carrying out the described optimization process. In a microcomputer 50, the components CPU 51, RAM 52, ROM 53, timer 54, first input / output unit 55 and second input / output unit 56 are connected to one another via an address and a data bus 57. An oscillator 58 is used to time the program sequence in the microcomputer 50, which is connected on the one hand directly to the CPU 51 and on the other hand via a divider 59 to the timer 54. The signals of an exhaust gas probe 63, a speed sensor 64 and a reference mark sensor 65 are fed to the first input / output unit 55 via conditioning circuits 60, 61 and 62, for example. The battery voltage 66, the throttle valve position 67, the cooling water temperature 68 and the output signal of the torque transmitter 69 are provided as further input variables. If the torque of the internal combustion engine is obtained directly from the speed, the speed sensor 64 could also be used to detect the torque.

These input variables are connected to a series circuit comprising a multiplexer 74 and an analog-to-digital converter 75 via associated conditioning circuits 70, 71, 72 and 73. The function of the multiplexer 74 and the analog-digital converter 75 can be implemented, for example, by the 0809 module from National Semiconductors. The multiplexer 74 is controlled via a line 76 starting from the first input / output unit 55. The second input / output unit 56 controls injection valves 78 of the internal combustion engine via power output stages 77. For the application of the method according to the invention, it is immaterial whether the fuel is an injection system with one injection valve per cylinder or an injection system with a single injection valve arranged in the air intake duct of the internal combustion engine.

The functioning of the device described naturally depends very considerably on the programming of the microcomputer. In the German patent application P 34 03 394.7, the program sequence for the fuel metering in an internal combustion engine with pilot control, extreme value control and map learning method is described quite extensively. For this reason, only the narrow procedural steps that are typical of a cylinder-specific optimization are to be explained in the block diagram below with reference to FIG. After the ignition is switched on, the injection _quantity or times _dependent on the operating parameters are calculated in the main program or read out from a characteristic _diagram , the same injection _times T _m0 being initially _assumed for each cylinder n of the internal combustion engine. Ignition times and other variables are also calculated in the main program. The cylinder-specific optimization of the fuel metering or the efficiency takes place by means of the subroutine T _m . First, the injection times T _{, 10} , T _{, 30,} for example of cylinders 1 and 3 of the internal combustion engine, are swept in opposite directions by the amount AT. After a phase comparison between torque change or speed change and wobble signal of, for example, cylinder 1, the individual cylinder injection times are changed in accordance with the result of the comparison under the boundary condition of a constant total injection time. Subsequently, there is a query as to whether the torque or speed change caused by the wobbling of the injection time approximately assumes the value zero or has fallen below a certain lower threshold value. If this is the case, the ratio of the injection times for the first and third cylinders is stored. If the change in torque is still above a predetermined threshold value, the cylinder-specific injection times are modified accordingly after a new phase comparison. As a boundary condition when varying the cylinder-specific injection times, it must always be taken into account that the sum of the injection times, here for example T ,, and T, _{3 takes on} a constant value.

• 1. Optimierung : T_i1/T_i3 = Konstante 1
• 2. Optimierung : T,₂/T,₄ = Konstante 2
• 3. Optimierung : T_i1/T_i4 = Konstante 3
• (die 3. Optimierung könnte auch mit den Einspritzzeiten T_i2, T,₃ alternativ durchgeführt werden).
  

In the next step, for example, the injection times of cylinders 2 and 4 are optimized in accordance with the subroutine T _m and the injection times are stored as a ratio in a memory. After further optimization of a third combination of individual cylinders or individual cylinder groups in the present exemplary embodiment, either cylinders 1 and 4 or 2 and 3, sufficient information is available to calculate the cylinder-specific injection times. The dotted connection line marked "iteration steps" is intended to indicate that the optimization can be carried out more frequently than indicated here for the iterative approximation of the cylinder-specific injection times. Ideally, an n-cylinder internal combustion engine (n-1) optimization processes for different cylinders or cylinder groups are required. This can be seen from a short example of a four-cylinder internal combustion engine from the following list:

• 1. Optimization: T _i1 / T _i3 = constant 1
• 2. Optimization: T, ₂ / T, ₄ = constant 2
• 3. Optimization: T _i1 / T _i4 = constant 3
• (The third optimization could alternatively be carried out with the injection _times T _i2 , T, ₃ ).
  

This means that four independent equations are available for the four unknown single-cylinder injection times based on three optimization processes and the sum condition, which can be easily solved.

In order to ensure that the operating conditions of the internal combustion engine have approximately constant values during the optimization process, corresponding, known interrogation devices are provided which interrupt or restart the optimization process if the changes are too great.

In FIG. 5, the wobble signals are plotted using the example of an optimization process for the injection times T-, T, ₃ and the associated torque or speed signals. The injection time T ,. increased by the amount ΔT and the injection time T, ₃ decreased by the amount AT. The reaction of the internal combustion engine to these modified injection times can manifest itself in an increase or decrease in torque. Depending on whether the increase in the injection time of the cylinder 1 leads to an increase in torque (in phase) or a decrease in torque (against phase), the injection time T. (T ₃ ) is increased (decreased) or decreased (increased) under the boundary condition a constant total injection time (T ,. + T ₃ ). After the first time period _T , the optimization process continues in this manner. that the injection time T .. is reduced by the amount AT and the injection time T _{3 is} increased by ΔT. The phase of the torque change of the internal combustion engine also changes accordingly. To evaluate the phase position between the wobble signal of the injection time and the resulting change in torque or speed, digital filters, as described in the German application P 3 403 304.7, can advantageously be used.

While the applications discussed so far are always single cylinder engines Concerning injection, the application of the invention to an internal combustion engine with a single central injection valve will be briefly explained with reference to FIG. In the diagram in FIG. 6, the ignition times, the opening times of the intake valves and the injection pulses for the central injection valve are plotted against the crankshaft angle. An ignition sequence 1-3-4-2 was assumed for cylinders 1 to 4. The injection process must now be synchronized in this way. _' That an injection pulse can be assigned to each cylinder or that the amount of fuel supplied per injection pulse for the most part reaches a single cylinder.

In the example, the first injection pulse is sprayed off at a time selected in such a way that after the running time (injection valve - intake valve) it reaches the intake valve of the 4th cylinder exactly at the time of opening. Accordingly, the 2nd injection pulse appears in the second cylinder. In practice, it may prove necessary to shift the start of the injection period depending on the operating parameters in order to take into account the running times from the injection valve to the intake valve. Given the total injection quantity per two revolutions, the injection quantity assigned to the individual cylinder can now be varied. Again, the injection pulses belonging to two cylinders or cylinder groups are swept in opposite directions and varied in opposite directions on average so that, as already described, a maximum torque results.

The proposed cylinder optimization can be used in every operating point of the internal combustion engine, of course also in the be _{n "} or P max operating point. It is also possible to use a higher-level control loop, for example using a lambda probe, to determine the air ratio lambda averaged over all cylinders Then, as already described above, the maximum efficiency of the internal combustion engine is found for this operating point with the aid of the single cylinder optimization. With regard to future exhaust gas legislation, the operating points at Lambda = are particularly interesting 1. The higher-order control circuit then maintains the mean air ratio at the value Lambda = 1 in a manner known per se by means of a (lambda = 1) probe. By means of a single-cylinder optimization, the lambda in which the cylinder in question can be set can now be set for each cylinder in his we efficiency degree works. Since, without optimization, the tolerances A Lambda in the lambda value from cylinder to cylinder can easily be A Lambda - 0.1, a significantly smaller fluctuation range can be expected after optimization. A smaller fluctuation range of the lambda value from cylinder to cylinder would also bring advantages in the dimensioning of the catalytic converter, since today's catalytic converters are constructed quite voluminously because of this fluctuation range in order to average over several combustion cycles of the internal combustion engine.

Claims (14)

1. Procedure for managing the operating characteristics of a multi-cylinder internal combustion engine in order to optimize its efficiency, in which the fuel-air ratio of groups of cylinders of the internal combustion engine is varied, characterized by the following procedural steps :

a) a first step, in which time-dependant signals for influencing the fuel-air ratio lambda of the operating mixture fed in each case to at least two arbitrary groups of cylinders, comprising in each case at least one cylinder of the internal combustion engine, are produced in such a way that the fuel-air ratio is modified in a fashion specific to the groups of cylinders, and the average fuel-air ratio of the operating mixture fed to all cylinders is kept at least approximately constant ;

b) a second step, with detection of the reaction of the internal combustion engine to the signals of the first step expressed in a variation of an output magnitude linked with the efficiency, the output magnitude being the torque or the speed of the internal combustion engine ; and

c) a third step, with influence being exerted on the efficiency of the individual groups of cylinders of the internal combustion engine in order to optimize the efficiency according to the results of the second step.

2. Procedure according to Claim 1, characterized in that the third step the fuel-air ratio specific to the groups of cylinders is varied for the groups of cylinders involved.

3. Procedure according to Claim 2, characterized in that the fuel-air ratio specific to the groups of cylinders is varied inversely.

4. Procedure according to one of Claims 1 to 3, characterized in that the variation of an output magnitude as reaction of the internal combustion engine to the first step is compared with a threshold value.

5. Procedure according to one of Claims 1 to 4, characterized by repeated application to different groups of cylinders, the number of repetitions being determined at least by the number of cylinders.

6. Procedure according to one of Claims 1 to 5, characterized in that the groups of cylinders are combined from different cylinders, the number of the combinations being determined by at least the number of the cylinders.

7. Procedure according to one of Claims 1 to 6, characterized in that in the first step the fuel-air ratio lambda is influenced in a fashion specific to the groups of cylinders by variation of the fuel quantity fed to the groups of cylinders, the air feed being approximately constant.

8. Procedure according to Claim 7, characterized in that the fuel feed is actuated by at least one injection valve, and varied over the time duration of injection, or over the time duration of injection and the point of injection.

9. Procedure according to one of Claims 4 to 8, characterized in that after the variation in output magnitude of the internal combustion engine has fallen below the threshold value, the amplitude of the time dependant signals, or the lambda values specific to the groups of cylinders, or the injection period are stored.

10. Procedure according to Claim 8, characterized in that the time durations of injection and the points of injection specific to the groups of cylinders are modified inversely, so that as the total of the individual injection periods of the individual cylinders the overall injection periods (sic) assumes a constant value.

11. Procedure according to one of Claims 1 to 10, characterized in that the fuel-air ratio lambda the (sic) operating mixture fed to the internal combustion engine is precontrolled by an identification field.

12. Procedure according to Claim 11, characterized in that the identification field values may be adapted in a fashion specific to the groups of cylinders.

13. Procedure according to one of Claims 1 to 12, characterized in that the average fuel-air ratio of the operating mixture fed to all cylinders is controlled to assume a value that may be set as a function of the operating parameters.

14. Apparatus for carrying out the procedure according to at least one of the preceding claims, having a micro-computer and peripheral devices for optimizing the efficiency of an internal combustion engine, characterized in that there is (sic) provided a first function for producing time-dependant signals for influencing the fuel-air ratio lambda of the operating mixture, fed in each case to at least two arbitrary groups of cylinders comprising in each case at least one cylinder of the internal combustion engine, in such a way that the fuel-air ratio is modified in a fashion specific to the groups of cylinders, and the average fuel-air ratio of the operating mixture fed to all cylinders is kept at least approximately constant, a second function for detection of the reaction of the internal combustion engine to the signals of the first function expressed in a variation of an output magnitude linked with the efficiency, the output magnitude being the speed or the torque of the internal combustion engine, and a third function influencing the efficiency of the individual groups of cylinders of the internal combustion engine according to the results of the second function in order to optimize the efficiency.

Images