JP2008045475A - Adapting method of internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To calculate an optimal parameter value by a smaller calculation load. <P>SOLUTION: This adapting method selects a plurality of whole characteristic values provided or generated from operation of an internal combustion engine, from parameter values capable of falling within respective corresponding predetermined allowable ranges, as a target value of a plurality of respective parameters imparted to operation of the internal combustion engine. A discrete value is imparted to respective values of the respective parameters in response to whether or not a continuous value of the characteristic values falls within the allowable ranges when making the respective parameters take a plurality of values. The target value of the respective parameters capable of making the plurality of whole characteristic values fall within the respective corresponding predetermined allowable ranges, is selected by using a support vector machine, based on the discrete value. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method for adapting an internal combustion engine.

  In the field of internal combustion engines, for example, throttle valve opening, ignition timing by spark plugs, on / off valve characteristics of intake valves or exhaust valves, fuel, so as to satisfy the requirements for the characteristics of the internal combustion engine such as generated torque, exhaust emission and fuel consumption Parameters such as the fuel injection amount from the injection valve are controlled. Such parameter values (hereinafter referred to as “parameter values”) have optimum values for the state of the internal combustion engine at that time, for example, the load on the internal combustion engine and the engine speed. For this reason, in order to satisfy the requirements for the characteristics of the internal combustion engine sufficiently and accurately, an optimal value as a parameter value is obtained in advance as a target value for each state of the internal combustion engine by an experiment or the like. It is preferable to control the operation of, for example, a throttle valve, a spark plug, an intake valve or an exhaust valve, and a fuel injection valve so that the control parameter value becomes the target value thus determined.

  However, if an optimum value for the target value of the parameter value is obtained in advance for each state of the internal combustion engine by experiments or the like, a large number of experimental data must be acquired. Therefore, in the invention described in Patent Document 1, for example, characteristic values of the internal combustion engine such as fuel consumption and NOx (nitrogen oxide) emission from the combustion chamber, the opening of the throttle valve, the ignition timing by the spark plug, and the fuel injection A model formula indicating a relationship with a parameter value such as a fuel injection timing from the valve is calculated, and based on this model formula, an optimal parameter value for satisfying the requirements for the characteristics of the internal combustion engine is calculated as a target value.

JP 2002-206456 A JP 2004-68729 A JP 2005-903353 A Japanese Patent Laid-Open No. 2004-279111 JP 2005-180196 A JP 2005-226544 A

  By the way, when calculating the optimal parameter value using the model formula as in the invention described in Patent Document 1, if the model formula is higher order, the model formula is the characteristic value and parameter value of the internal combustion engine. It can be said that a more optimal parameter value is calculated. However, in order to calculate a higher-order model formula, calculation with a larger load is required. On the other hand, if the model formula is low order, the calculation load for calculating the model formula becomes smaller. However, the model formula does not accurately represent the relationship between the characteristic value of the internal combustion engine and the parameter value, and it cannot be said that a more optimal parameter value is calculated.

  In any case, when an optimal parameter value is calculated using a model equation as in the invention described in Patent Document 1, it is difficult to calculate a model equation that calculates a more optimal parameter value. I can say that.

  Accordingly, an object of the present invention is to calculate a more optimal parameter value with a smaller calculation load.

  In order to solve the above-described problem, in the first invention, as target values for each of a plurality of parameters that affect the operation of the internal combustion engine, all of a plurality of characteristic values obtained or generated from the operation of the internal combustion engine are respectively corresponded in advance. In an adaptation method that is obtained by selecting from among parameter values that can fall within a predetermined allowable range, the continuous value of the characteristic value when a plurality of values are taken for each parameter is the allowable range. A discrete value is assigned to each value of each parameter depending on whether or not it falls within the range, and a plurality of characteristic values are respectively determined in advance using a support vector machine based on the discrete value. The range of each parameter that can be within the allowable range is determined, and the target value of each parameter is selected from the parameter values within the determined range. To.

  In the second invention, in the first invention, the present invention is applied to an object having strong nonlinearity.

  According to the present invention, a more optimal parameter value is calculated with a smaller calculation load.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows an internal combustion engine to which the present invention is applied. In FIG. 1, 1 is an engine body, 2 is a cylinder block, 3 is a piston, 4 is a cylinder head, 5 is a combustion chamber, 6 is an intake valve, 7 is an intake port, 8 is an exhaust valve, 9 is an exhaust port, Spark plugs 11 indicate fuel injection valves. The fuel injection valve 11 is attached to the cylinder head 4 so as to inject fuel directly into the combustion chamber 5. A cavity 12 is formed on the upper wall surface of the piston 3.

  The intake port 7 is connected to a surge tank 14 via a corresponding intake branch pipe 13. The surge tank 14 is connected to the intake pipe 15. A throttle valve 18 driven by a step motor 17 is disposed in the intake pipe 15. On the other hand, each exhaust port 9 is connected to a corresponding exhaust branch pipe 19. Also, the intake valve 6 is provided with a variable valve mechanism 20 that changes the on-off valve characteristics of the intake valve 6, that is, the timing at which the intake valve 6 opens and closes and the period during which the intake valve 6 is open. .

  A throttle opening sensor 31 for measuring the opening of the throttle valve 18 is attached to the throttle valve 18. An air flow meter 32 for measuring the flow rate of air flowing through the intake pipe 15 is attached to the intake pipe 15 upstream of the throttle valve 18. A temperature sensor 33 for measuring the temperature of exhaust gas discharged from the engine body 1 and an air-fuel ratio sensor 34 for measuring the air-fuel ratio of exhaust gas discharged from the engine body 1 are attached to the exhaust branch pipe 19. ing. A torque sensor (not shown) for detecting torque generated from the internal combustion engine is attached to the crankshaft (not shown) of the engine body 1.

  In FIG. 1, reference numeral 40 denotes an electronic control unit. The electronic control device 40 receives outputs from a throttle opening sensor 31, an air flow meter 32, a temperature sensor 33, an air-fuel ratio sensor 34, and a torque sensor. Further, the electronic control unit 40 controls the operation of the spark plug 10, the fuel injection valve 11, the throttle valve 18, and the variable valve mechanism 20.

  By the way, in the field of internal combustion engines, for example, the degree of opening of the throttle valve 18 and the mixture in the combustion chamber 5 by the spark plug 10 are satisfied so that the requirements for the characteristics of the internal combustion engine such as generated torque, exhaust emission, and fuel consumption are satisfied. Parameters such as the ignition timing, the on-off valve characteristics of the intake valve 6 or the exhaust valve 8, and the fuel injection timing from the fuel injection valve 11 are controlled. Here, in order to ensure that the requirements for the characteristics of the internal combustion engine are satisfied, for each parameter, values that satisfy the requirements for the characteristics of the internal combustion engine (hereinafter referred to as “optimum values”) are obtained in advance by experiments. In addition, during operation of the internal combustion engine, it is preferable to control the values of the parameters to optimum values. Therefore, in the embodiment of the present invention, the optimum value of the parameter is obtained by experiment as follows.

  For example, a torque generated from the internal combustion engine (hereinafter referred to as “engine torque”) is adopted as a characteristic of the internal combustion engine, and a timing at which the intake valve 6 is opened / closed as a parameter that affects the engine torque (hereinafter referred to as “open / close valve timing”). )), The timing at which the air-fuel mixture in the combustion chamber 5 is ignited by the spark plug 10 (hereinafter referred to as “ignition timing”), and the timing at which fuel is injected from the fuel injection valve 11 (hereinafter referred to as “fuel injection timing”). As a condition (hereinafter referred to as “constraint condition”) that is adopted and limits the range of allowable values for these parameters, engine torque fluctuation (hereinafter referred to as “torque fluctuation”) and NOx generated in the combustion chamber 5 are used. The amount of (nitrogen oxide) (hereinafter referred to as “NOx generation amount”) and the amount of unburned HC (unburned hydrocarbon) discharged from the combustion chamber 5 (hereinafter referred to as “HC emission amount”) are taken. Description will be given of a case where the.

  First, for each engine speed and torque required for the internal combustion engine (hereinafter referred to as “required torque”), engine torque corresponding to a plurality of predetermined values for each parameter is measured. Next, based on the measured values of the engine torque, a statistical approximation model of engine torque with each parameter as a function is created.

  Next, or at the same time, for each parameter, depending on whether or not the torque fluctuation corresponding to each of the plurality of predetermined values is smaller than a certain fluctuation amount, that is, within an allowable range. A discrete value is assigned to each value of each parameter. Specifically, when the torque fluctuation when the parameter value is set to a certain value is within an allowable range, a discrete value “1” is assigned to the parameter value. On the other hand, if the torque fluctuation when the parameter value is a certain value is not within the allowable range, a discrete value “0” is assigned to the parameter value.

  Next, or at the same time, for each parameter, the amount of NOx generated corresponding to each of the plurality of predetermined values is smaller than a certain amount, that is, depending on whether or not it falls within an allowable range. A discrete value is assigned to each value of each parameter. Specifically, when the amount of NOx generated when the parameter value is a certain value is within an allowable range, a discrete value “1” is assigned to the parameter value. On the other hand, if the amount of NOx generated when the parameter value is a certain value is not within the allowable range, a discrete value “0” is assigned to the parameter value.

  Next, or at the same time, for each parameter, depending on whether the HC emission amount corresponding to each of the plurality of predetermined values is smaller than a certain amount, that is, within an allowable range. A discrete value is assigned to each value of each parameter. Specifically, when the HC generation amount when the parameter value is a certain value is within an allowable range, a discrete value “1” is assigned to the parameter value. On the other hand, if the HC emission amount when the parameter value is a certain value is not within the allowable range, a discrete value “0” is assigned to the parameter value.

  Then, for all of the constraint conditions, that is, for all of the torque fluctuation, NOx generation amount, and HC emission amount, depending on whether or not the discrete value “1” is given, for each value of each parameter A discrete value is given again. Specifically, the discrete value “1” is newly assigned to the parameter values to which the discrete value “1” is assigned for all the constraint conditions. On the other hand, the discrete value “0” is newly assigned to the parameter value to which the discrete value “0” is assigned with respect to at least one constraint condition.

  Then, using the support vector machine, based on the value of the parameter to which the discrete value “1” is assigned, the boundary line of the parameter value that falls within the allowable range of the constraint condition, that is, the range is determined.

  Finally, using a statistical approximation model of engine torque, the engine speed and required torque are optimized with the parameter value that gives the highest engine torque within the parameter value range within which the constraints are within the allowable range. It is calculated every time and is stored in the form of a map. Thus, the optimum value of the parameter as a function of the engine speed and the required torque with respect to the engine torque is obtained in the form of a map.

  FIG. 2 shows an example of a routine for executing the method of the above-described embodiment for creating an optimum value map for each parameter. In the routine shown in FIG. 2, first, in step 10, for each engine speed and required torque, engine torque corresponding to a plurality of values determined in advance for each parameter is measured. Next, at step 11, based on the engine torque measured at step 10, a statistical approximation model of engine torque with each parameter as a function is created.

  Next, in step 12, for each parameter, a discrete value is obtained for each value of each parameter depending on whether or not the torque fluctuation corresponding to each of the plurality of predetermined values is within an allowable range. “1” or “0” is assigned. Next, in step 13, for each parameter, the value of each parameter is discrete depending on whether the NOx generation amount corresponding to each of the plurality of predetermined values is within an allowable range. The value “1” or “0” is assigned. Next, at step 14, for each parameter, the discrete values for each parameter are determined depending on whether or not the HC emission amount corresponding to each of the plurality of predetermined values is within an allowable range. Assign a value.

  Next, in step 15, the discrete value “1” or “0” is newly given to each value of each parameter depending on whether or not the discrete value “1” is given for all the constraint conditions. . Next, in step 16, using the support vector machine, based on the parameter value to which the discrete value “1” is newly given, the parameter value range within which the constraint condition is acceptable is determined.

  Next, at step 17, using the statistical approximate model of engine torque created at step 11, the parameter value at which the engine torque is highest within the range of parameter values within which the constraint conditions are acceptable is set as the optimum value. calculate. Finally, in step 18, based on the optimum value calculated in step 17, an optimum value map using the engine speed and the required torque as a function is created.

  In the above-described example, the case where the engine torque is employed as the characteristic of the internal combustion engine and the torque fluctuation, the NOx generation amount, and the HC emission amount are employed as the constraint conditions has been described. When the fuel consumption is adopted as the characteristics of the engine, and the fluctuation conditions are combustion fluctuations (combustion fluctuations in the combustion chamber), the temperature of the exhaust gas discharged from the combustion chambers, and the exhaust purification catalyst is disposed in the exhaust passage Temperature of exhaust purification catalyst, temperature of air sucked into combustion chamber, pressure in intake passage, amount of CO (carbon monoxide) generated in combustion chamber, amount of smoke generated in combustion chamber, generated in combustion chamber The present invention is equally applicable when the amount of particulates is employed.

1 is an overall view of an internal combustion engine to which the present invention is applied. It is the figure which showed an example of the routine which performs the method of embodiment of this invention which produces the map of the optimal value of each parameter.

Explanation of symbols

6 Intake valve 8 Exhaust valve 10 Spark plug 11 Fuel injection valve 18 Throttle valve 20 Variable valve mechanism

Claims (2)

  As a target value for each of a plurality of parameters given to the operation of the internal combustion engine, a parameter value that can be set within a predetermined allowable range corresponding to all of a plurality of characteristic values obtained or generated from the operation of the internal combustion engine. In the calibration method to be selected and selected from the above, the respective values of the respective parameters are determined depending on whether or not the continuous values of the above characteristic values are within the allowable range when each parameter has a plurality of values. A target value of each parameter that can be set within a predetermined allowable range corresponding to each of the plurality of characteristic values by using a support vector machine based on the discrete value. A fitting method characterized by selecting.   A fitting method characterized by being applied to an object having strong nonlinearity.

Images