METHOD FOR MODEL-BASED CONTROL AND REGULATION OF AN INTERNAL COMBUSTION ENGINE

Description

[0001] The invention relates to a method for model-based control and regulation of an internal combustion engine, in which a quality measure is calculated by an optimizer and set as decisive for the operating point of the internal combustion engine.

[0002] The behavior of an internal combustion engine is largely determined by an engine control unit (ECU) depending on the desired power output. For this purpose, corresponding characteristic curves and maps are typically applied in the ECU's software. These are used to calculate the engine's manipulated variables, such as the start of injection and the required rail pressure, from the desired power output. These characteristic curves / maps are populated with data on a test bench at the engine manufacturer's plant. However, the multitude of these characteristic curves / maps and the correlation between them requires a significant amount of tuning effort.

[0003] In practice, attempts are therefore being made to reduce the tuning effort by using mathematical models. For example, DE 10 2006 004 516 B3 describes a Bayesian network with probability tables for determining the injection quantity, and US 2011 / 0172897 A1 describes a method for adapting the start of injection and the injection quantity using combustion models with neural networks. Since trained data is used in this case, it must first be learned during a test bench run.

[0004] DE 10 2017 005 783 A1 discloses a method for model-based control and regulation of an internal combustion engine. In this method, the setpoints for the injection system actuators are calculated using a combustion model, and the setpoints for the gas path actuators are calculated using a gas path model. Both the combustion model and the gas path model are based on Gaussian process models. From the setpoints, an optimizer determines a quality measure and predicts how the quality measure would develop if the setpoints were changed within a prediction horizon. Once the best possible quality measure has been calculated, the optimizer sets the injection system setpoints and the gas path setpoints as the determining factors for the internal combustion engine's operating point.

[0005] From WO 2018 / 234093 A1 a method for model-based control and regulation of an internal combustion engine is known, which comprises an optimization.

[0006] Test bench experiments have shown that the integration of control variables with discrete switching states into the previously described model-based method is not yet satisfactory. Control variables with discrete switching states include, for example, the activation of the second exhaust gas turbocharger during turbocharging, cylinder bank deactivation, the activation of pre- or post-injection, and the open or closed position of various valves. So-called branch and bound methods for finding optimal solutions for discrete control variables are very computationally intensive, since in the worst case, all possible combinations of the discrete control variables must be investigated. Their application in an internal combustion engine quickly leads to very complex structures that cannot be represented on an engine control unit.

[0007] The invention is based on the object of improving the previously described model-based method with regard to the integration of manipulated variables.

[0008] This object is achieved by the features of claim 1. The embodiments are presented in the subclaims.

[0009] The process is carried out in three steps. In the first step, the optimizer calculates a pre-optimized quality measure depending on the operating situation, whereby the discrete manipulated variables with discrete setting values ​​are interpreted as continuous manipulated variables with a continuous setting range. The pre-optimized quality measure is a calculated value, i.e., it is not applied to the internal combustion engine. In the second step, these continuous manipulated variables are quantized and set as new discrete manipulated variables with discrete setting values. Quantization is performed using switching thresholds and hysteresis. Finally, in the third step, the optimizer calculates a post-optimized quality measure depending on the new discrete manipulated variables and the operating situation of the internal combustion engine and sets this as the decisive factor for the operating point of the internal combustion engine.However, when calculating the re-optimized quality measure, the new discrete manipulated variables are assumed to be fixed. They therefore no longer represent a degree of freedom for optimization within the predicted horizon. The remaining continuous manipulated variables are re-optimized so that the solution is the best possible with respect to the fixed new manipulated variables.

[0010] The operating situation of the internal combustion engine includes both the external conditions, in particular the emission limits or the desired power output, as well as the current operating point. Both the pre-optimized performance measure and the post-optimized performance measure are determined by calculating the injection system setpoints for controlling the injection system actuators, such as the target rail pressure, using the combustion model. The gas path setpoints for controlling the gas path actuators are calculated using a gas path model. These setpoints are then modified by the optimizer with the goal of finding the minimum.

[0011] The invention allows the solution of optimization tasks with partially continuous-value and partially discrete-value input variables, even with limited computing capacity for the optimization method used. Instead of parallel calculation of the manipulated variables, as required for the implementation of branch-and-bound methods, the invention uses a serial methodology. Only in this way can the quality measure and the resulting values ​​for the manipulated variables be fully calculated on an engine control unit.

[0012] The figures show a preferred embodiment. Fig. 1 a system diagram, Fig. 2 a model-based system diagram, Fig. 3 a block diagram, Fig. 4 a program flow chart, Fig. 5 a subroutine, Fig. 6 a subroutine, Fig. 7 a subroutine, Fig. 8 timing diagrams and Fig. 9 timing diagrams.

[0013] The Figure 1shows a system diagram of an electronically controlled internal combustion engine 1 with a common rail system. The common rail system comprises the following mechanical components: a low-pressure pump 3 for pumping fuel from a fuel tank 2, a variable intake throttle 4 for influencing the fuel volume flow, a high-pressure pump 5 for pumping the fuel under increased pressure, a rail 6 for storing the fuel, and injectors 7 for injecting the fuel into the combustion chambers of the internal combustion engine 1. Optionally, the common rail system can also be designed with individual accumulators, in which case, for example, an individual accumulator 8 is integrated into the injector 7 as an additional buffer volume. The other functionality of the common rail system is assumed to be known. The gas path shown includes both the air supply and the exhaust gas discharge.The following are arranged in the air supply: the compressor of an exhaust gas turbocharger 11, a charge air cooler 12, a throttle valve 13, an inlet point 14 for combining the charge air with the recirculated exhaust gas, and a variably controllable inlet valve 15. The following are arranged in the exhaust gas duct: a variably controllable outlet valve 16, an EGR actuator 17, the turbine of the exhaust gas turbocharger 11, and a turbine bypass valve 18.

[0014] The operation of the internal combustion engine 1 is determined by an electronic control unit 10 (ECU). The electronic control unit 10 contains the usual components of a microcomputer system, such as a microprocessor, I / O modules, buffers, and memory modules (EEPROM, RAM). The operating data relevant to the operation of the internal combustion engine 1 are applied as models in the memory modules. The electronic control unit 10 uses these models to calculate the output variables from the input variables. Figure 1 The following input variables are shown as examples: A target torque M(TARGET), which is specified by an operator, the actual rail pressure pCR, which is measured by a rail pressure sensor 9, the engine speed nIST, the charge air pressure pLL, the charge air temperature TLL, the humidity phi of the charge air, the exhaust gas temperature TExhaust gas, the air-fuel ratio lambda, the NOx actual value, optionally the pressure pES of the individual accumulator 8 and an input variable EIN. The other sensor signals not shown are summarized under the input variable EIN, for example the coolant temperatures. In Figure 1are shown as output variables of the electronic control unit 10: a PWM signal for controlling the intake throttle 4, a ve signal for controlling the injectors 7 (start of injection / end of injection), a DK control signal for controlling the throttle valve 13, a VVT control signal for controlling the intake or exhaust valve, an EGR control signal for controlling the EGR actuator 17, a TBP control signal for controlling the turbine bypass valve 18 and an output variable OFF. The output variable OFF represents the other control signals for controlling and regulating the internal combustion engine 1, for example, a control signal for activating a second exhaust gas turbocharger during turbocharging. When displaying the Figure 1For example, the throttle valve 13, the EGR actuator 17, the turbine bypass valve 18, or the throttle valve 4 can be controlled with a continuous control signal and are therefore adjustable within a continuous range of values. A discrete control variable, on the other hand, would be the control signal for activating a second exhaust gas turbocharger, since this control signal can only assume individual discrete values; intermediate values ​​therefore do not exist.

[0015] The Figure 2shows a model-based system diagram. In this representation, a combustion model 19, a gas path model 20, and an optimizer 21 are listed within the electronic control unit 10. Both the combustion model 19 and the gas path model 20 represent the system behavior of the internal combustion engine as mathematical equations, for example in the form of Gaussian process models. The combustion model 19 statically represents the combustion processes. In contrast, the gas path model 20 represents the dynamic behavior of the air flow and exhaust gas flow. The combustion model 19 contains individual models, for example, for NOx and soot formation, for the exhaust gas temperature, for the exhaust gas mass flow, and for the peak pressure. These individual models, in turn, depend on the boundary conditions in the cylinder and the injection parameters.The combustion model 19 is determined for a reference internal combustion engine in a test bench run, the so-called DoE test bench run (DoE: Design of Experiments). During the DoE test bench run, operating parameters and manipulated variables are systematically varied with the goal of mapping the overall behavior of the internal combustion engine as a function of engine variables and environmental boundary conditions. The optimizer 21 evaluates the combustion model 19 with regard to the target torque M(TARGET), the emission limits, the environmental boundary conditions, for example, the humidity phi of the charge air, and the operating situation of the internal combustion engine. The operating situation is defined by the engine speed nIST, the charge air temperature TLL, the charge air pressure pLL, etc. The function of the optimizer 21 is to evaluate the injection system target values ​​for controlling the injection system actuators and the gas path target values ​​for controlling the gas path actuators.Here, optimizer 21 selects the solution that minimizes a quality measure. The quality measure is calculated as the integral of the squared target-actual deviations within the prediction horizon. For example, in the form: . J = ∫ w 1 N 0 x SOLL − N 0 x IST 2 + w 2 M SOLL − M IST 2 + w 3 … . +

[0016] Here, w1, w2, and w3 represent a corresponding weighting factor. As is well known, nitrogen oxide emissions result from the charge air humidity phi, the charge air temperature TLL, the start of injection SB, and the rail pressure pCR.

[0017] The optimizer 21 determines the best possible quality measure via minimum finding by calculating a first quality measure at a first point in time, varying the injection system setpoints and the gas path setpoints, and using this to predict a second quality measure within the prediction horizon. Based on the deviation between the two quality measures, the optimizer 21 then determines a minimum quality measure and sets this as the decisive factor for the internal combustion engine. For the example shown in the figure, this is the target rail pressure pCR(SL) for the injection system. The target rail pressure pCR(SL) is the reference variable for the underlying rail pressure control loop 22. The manipulated variable of the rail pressure control loop 22 corresponds to the PWM signal for actuating the intake throttle. For the gas path, the optimizer 21 indirectly determines the gas path setpoints.In the example shown, these are a lambda setpoint LAM(SL) and an EGR setpoint EGR(SL) for the specification of the two lower-level control loops 23 and 24. The fed-back measured variables MESS are read in by the electronic control unit 10. The measured variables MESS include both directly measured physical variables and auxiliary variables calculated from them. In the example shown, the lambda actual value LAM(IST) and the EGR actual value AGR(IST) are read in. The control variables of the internal combustion engine are summarized with the reference symbol SG. This includes both the continuous control variables with a continuous setting range and the discrete control variables with discrete setting values. Continuous control variables can be continuously adjusted between a minimum and maximum value, for example the start and end of injection with which the injector (. Fig. 1: 7) is applied directly. Discrete manipulated variables with discrete setting values ​​can only be set in stages as fixed values, for example, cylinder deactivation.

[0018] The Figure 3 shows a block diagram with the operating situation BS of the internal combustion engine as the input variable and the quality measure as the output variable, referred to here as the post-optimized quality measure J(NA). The block diagram shows a pre-optimization 25, a quantization 26, and a post-optimization 27. In a first step, a pre-optimized quality measure J(VO) is calculated using the pre-optimization 25, in which the discrete manipulated variables with discrete setting values ​​are interpreted as continuous manipulated variables with a continuous setting range.

[0019] An example of a discrete control variable is pre-injection, which can only be activated or deactivated. By using pre-injection, the peak pressure of combustion can be significantly reduced. In addition, all other combustion variables, such as NOx emissions or particle count, also change when pre-injection is activated. The internal combustion engine is measured once with pre-injection activated and once with pre-injection deactivated. This results in two separate combustion models. When calculating the pre-optimized quality measure J(VO), the optimizer interpolates intermediate values. This means that by interpolating between these two combustion models, the state of pre-injection activated or deactivated is artificially converted into a continuous input variable. This variable is then used continuously in the pre-optimization 25. Figure 3These continuous manipulated variables are referred to as SG(k). The pre-optimized quality measure J(VO) is a purely internal calculation variable that has no access to the actuators of the internal combustion engine. In other words: the pre-optimized quality measure J(VO) is access-free and is not applied to the internal combustion engine. In a second step, new discrete manipulated variables SG(new) are calculated from the continuous manipulated variables SG(k) using quantization 26. For the pre-injection, a fixed assignment to pre-injection activated or pre-injection deactivated is therefore made again in the quantization. Quantization 26 offers the advantage that, for example, the variable valve timing can be set to three discrete values, namely minimum, average, and maximum, for example 450°, 495°, and 540° crankshaft angle. This significantly reduces the computational effort required for the subsequent determination of the re-optimized quality measure.During quantization 26, the calculated values ​​are also stabilized using optional hysteresis bands. In a third step, the new discrete manipulated variables SG(new) and the operating situation are combined, and the optimizer calculates a re-optimized quality measure J(NA). When calculating the re-optimized quality measure J(NA), the new discrete manipulated variables SG(new) are not changed. As such, they do not represent a degree of freedom when calculating the re-optimized quality measure J(NA). During re-optimization, the actually continuous manipulated variables are adapted to the curve specified from the quantization, for example, the pre-injection. In other words: During re-optimization, the manipulated variables that are actually described by continuous manipulated variables are varied.The post-optimized quality measure J(NA) corresponds to the minimum quality measure J(min), which is set by the optimizer as decisive for the operating point of the internal combustion engine (1), i.e. is applied to the internal combustion engine.

[0020] In the Figure 4The process is shown in a program flow chart. After initialization at S1, a check is made at S2 to determine whether the start-up process has been completed. If the start-up process is still running (query result S2: no), the program branches back to point A. Once the start-up process has been completed, the operating status of the internal combustion engine is recorded at S3. The operating status is defined by the engine speed nIST, the charge air temperature TLL, the charge air pressure pLL, etc. At S4, the optimizer subroutine is called, and the initial values, for example, the start of injection, are generated at S5. In steps S6 to S8, the pre-optimization, quantization, and post-optimization subroutines are called one after the other. These subroutines are used in conjunction with the Figures 5 to 7The optimized quality measure calculated in the re-optimization subroutine is set as the minimized quality measure J(min), which determines the operating point of the internal combustion engine. Subsequently, a check is performed at S10 to determine whether an engine stop has been initiated. If this is not the case, query result S10: no, and the program branches back to point B. Otherwise, the program flowchart is terminated.

[0021] In the Figure 5the pre-optimization subroutine is shown as a program flow chart. At S1, a first quality measure J1(VO) of the pre-optimization is calculated using equation (1). A key feature here is that when calculating the first quality measure J1(VO), in addition to the continuous manipulated variables with a continuous control range, the discrete manipulated variables with discrete values ​​are interpreted as continuous manipulated variables via interpolation. At S2, a control variable i is set to zero. Then, at S3, the initial values ​​are changed and calculated as new setpoints for the manipulated variables. At S4, the control variable i is increased by one. Based on the new setpoints, a second quality measure J2(VO) of the pre-optimization is then forecast at S5 within the prediction horizon, for example for the next 8 seconds. At S6, the second quality measure J2(VO) is subtracted from the first quality measure J1(VO) and compared with a limit value GW.The further progress of the quality measure is checked by calculating the difference between the two quality measures. Alternatively, the number of times an optimization has already been run is checked by comparing the control variable i with a limit value iGW. The two limit value considerations therefore serve as a termination criterion for further optimization. If further optimization is possible (query result S6: no), the program branches back to point A. Otherwise, at S7 the optimizer outputs the second quality measure J2(VO) as a pre-optimized quality measure J(VO) together with the manipulated variables calculated in the process and feeds it into the main program of the . Figure 4 returned. The pre-optimized quality measure J(VO) is purely a computational value, meaning that the optimizer does not apply the calculated injection system setpoints, the calculated gas path setpoints, and the calculated manipulated variables to the internal combustion engine.

[0022] In the Figure 6The quantization subroutine is shown. At S1, the pre-optimized quality measure J(VO) is read in with the corresponding manipulated variables. Subsequently, those manipulated variables with original discrete setting values ​​are discretized. This is done at S2 using corresponding threshold values ​​with a hysteresis band. The hysteresis band prevents oscillating calculated values. Instead of a hysteresis band, other logic can be used which prevents rapid switching, for example, a time control. Subsequently, at S3, the new discrete manipulated variables SG(new) are output and fed into the main program of the Figure 4 returned.

[0023] In the Figure 7the re-optimisation sub-routine is shown as a program flow chart. Using the re-optimisation sub-routine, a re-optimised quality measure is determined from the operating situation of the internal combustion engine and the new discrete manipulated variables SG(new). When calculating the re-optimised quality measure, the new discrete manipulated variables are not adjusted. At S1, a first quality measure J1(NA) of the re-optimisation is calculated using equation (1). At S2, a control variable i is set to zero. Then, at S3, the initial values ​​are changed and calculated as new setpoints for the manipulated variables. At S4, the control variable i is increased by one. Based on the new setpoints, a second quality measure J2(NA) of the re-optimisation is then forecast at S5 within the prediction horizon, for example for the next 8 seconds. At S6, the second quality measure J2(VO) is subtracted from the first quality measure J1(VO) and compared with a limit value GW.The further progress of the quality measure is checked by calculating the difference between the two quality measures. Alternatively, the number of times an optimization has already been run is checked by comparing the control variable i with a limit value iGW. The two limit value considerations therefore serve as a termination criterion for further optimization. If further optimization is possible (query result S6: no), the program branches back to point A. Otherwise, at S7, the optimizer outputs the second quality measure J2(VO) as the minimum quality measure J(min) and feeds it into the main program of the . Figure 4 returned.

[0024] The two Figures 8 and 9show a comparison of the course of selected variables over time in seconds. The variables shown are: the variable valve timing VVT in degrees crankshaft angle, the start of injection SB in degrees before top dead center (TDC), the combustion pressure pCYL in the cylinder, and the engine speed nMOT. For the combustion pressure pCYL, the maximum permissible combustion pressure pMAX is also shown as a dashed line. On the left half of the drawing, these variables are shown using the previous optimization, while on the right half of the drawing, they are shown using the invention. The representation of the Figure 8 and the Figure 9 a gradually increasing target torque is used as the input variable. First, the variables are calculated according to the Figure 8described. In a first step, the optimizer calculates a pre-optimized quality measure based on the operating situation via pre-optimization. In this calculation, the discrete manipulated variables with discrete setting values ​​are interpreted as continuous manipulated variables with a continuous setting range. For the variable valve control VVT, this results in a continuous curve with arbitrary intermediate values ​​over the entire time range. However, such a curve is not representable for the VVT ​​actuator for controlling the variable valve with three defined actuator positions. The pre-optimized quality measure corresponds to a calculated start of injection SB and the corresponding cylinder pressure pZYL. The maximum value pMAX is maintained for the cylinder pressure pZYL. The manipulated variables result in an increasing engine speed nMOT during the observation period. The following Figure 9The VVT ​​curve shown corresponds to the curve after quantization. This clearly shows that, in contrast to the representation of the Figure 8 , the VVT ​​curve shows only three discrete values, namely 450°, 495°, and 540° crankshaft angle. The advantage is that the VVT ​​actuator can be controlled with just three values, which significantly reduces computational effort. The optimized quality measure is calculated from the VVT ​​curve based on the operating situation of the internal combustion engine. This corresponds to the curve of the start of injection (SB) and the cylinder pressure (pCYL), which also remains below the maximum value (pMAX) in this case. Reference symbol

[0025] 1 Internal combustion engine 2 Fuel tank 3 Low-pressure pump 4 Intake throttle 5 High-pressure pump 6 Rail 7 Injector 8 Individual accumulator 9 Rail pressure sensor 10 Electronic control unit 11 Exhaust turbocharger 12 Intercooler 13 Throttle valve 14 Junction point 15 Intake valve, variably controllable 16 Exhaust valve, variably controllable 17 EGR actuator (EGR: exhaust gas recirculation) 18 Turbine bypass valve 19 Combustion model 20 Gas path model 21 Optimizer 22 Rail pressure control loop 23 Lambda control loop 24 EGR control loop 25 Pre-optimization 26 Quantization 27 Post-optimization

Claims

1. Method for the model-based open-loop and closed-loop control of an internal combustion engine (1), in which a pre-optimized quality measure (J(VO)) is calculated by an optimizer (21) in a first step in a manner dependent on the operating situation (BS), wherein, during the calculation of the pre-optimized quality measure (J(VO)), discrete control variables with discrete setting values are interpreted as continuous control variables (SG(k)) with a continuous range of setting in addition to further continuous control variables with a continuous range of adjustment, in which method these continuous control variables (SG(k)) are quantized and set as new discrete control variables (SG(new)) with discrete setting values in a second step, in which method a post-optimized quality measure (J(NA)) is calculated by the optimizer (21) in a third step in a manner dependent on the new discrete control variables (SG(new)) and the operating situation (BS) of the internal combustion engine (1), wherein the new discrete control variables (SG(new)) are assumed to be fixed and the further continuous control variables are re-optimized, and the post-optimized quality measure (J(NA)) is set by the optimizer (21) as being definitive for the operating point of the internal combustion engine (1).

2. Method according to Claim 1, characterized in that the pre-optimized quality measure (J(VO)) is determined by virtue of injection system setpoint values for the activation of injection system control elements being calculated by means of a combustion model (19), by virtue of gas path setpoint values for the activation of gas path control elements being calculated by means of a gas path model (20), and by virtue of the continuous control variables (SG(k)) being calculated from the discrete setting values of the discrete control variables by means of interpolation.

3. Method according to Claim 2, characterized in that the pre-optimized quality measure (J(VO)) of the internal combustion engine (1) is not offered.

4. Method according to Claim 3, characterized in that the continuous control variables (SG(k)) are quantized in the second step by means of switching thresholds together with hysteresis.

5. Method according to Claim 1, characterized in that the post-optimized quality measure (J(NA)) is determined in the third step by virtue of injection system setpoint values for the activation of injection system control elements being calculated by means of the combustion model (19), gas path setpoint values for the activation of gas path control elements being calculated by means of the gas path model (20), and by virtue of the injection system setpoint values and the gas path setpoint values for constant new discrete control variables (SG(new)) being varied by the optimizer (21) with the aim of finding a minimum within a prediction horizon.