JP3598521B2 - Data table processing device - Google Patents

Description

[0001]
[Industrial applications]
The present invention relates to a data table processing apparatus suitable for table interpolation of non-linear output characteristics such as an output characteristic of an air flow meter for measuring an intake air amount of an engine.
[0002]
[Prior art]
In a non-linear real-time control system such as an engine control, for example, as described in Japanese Patent Publication No. 2-14985, a control amount for each input variable value is obtained by interpolating adjacent data values in a data table. Many interpolation calculations are used.
[0003]
[Problems to be solved by the invention]
As described above, many interpolation calculations are used in engine control and the like, and each of them requires a relatively long calculation time, so that the calculation load on the CPU increases. Further, for example, in the case where the interpolation accuracy is to be improved for an output characteristic having a large change such as an output characteristic of an air flow meter with respect to an intake air amount of an engine, linear interpolation, i.e. In the conventional method of interpolating, since a large amount of data must be input by reducing the grid interval, a large-capacity memory is required, and the number of development steps (setting steps) also increases. In addition, the number of grid points themselves can be reduced by performing quadratic or higher order interpolation.However, in a system in which quadratic or higher order interpolation is performed every time, a large amount of time is required for interpolation calculation, and in general, it is not practical. No. Therefore, an interpolation operation expression such as a quadratic function for interpolating between the grid points of the data table is set, and the coefficients of the interpolation operation expression are defined for each of the lattice points and stored in the ROM area in advance. It is conceivable to use an interpolation method in which a coefficient is obtained by a search and substituted into an interpolation operation expression. However, although this method can solve the problems of the memory capacity, accuracy, and calculation time, the setting at the time of development becomes complicated and troublesome, and the coefficient cannot be updated by learning.
[0004]
The present invention has been made in view of the above problems, and has as its object to realize highly accurate interpolation calculation using an interpolation calculation expression without complicating settings during development, and to reduce ROM capacity. I do.
[0005]
[Means for Solving the Problems]
According to the present invention, if the grid interval of the data table is reduced, the number of data to be stored increases, the memory capacity increases, and the number of development steps increases. If interpolation is performed based on the calculation formula based on the data values, the calculation time will be too long and it will not be practical.Therefore, the grid point data will be converted into coefficients and stored, and the setting during development will be complicated. The setting of the coefficients is automatically performed so as not to make the conversion, and FIG. 1 is an overall configuration diagram thereof. That is, the data table processing device according to the present invention uses a data table that stores control data corresponding to the size of a predetermined parameter, and exhibits a non-linear output characteristic corresponding to an input variable value related to the parameter. Linear interpolation calculating means for calculating control data by linear interpolation based on data values of two adjacent grid points on the data table corresponding to each input variable value; using a data table for storing control data Te, the data provided with the secondary interpolation means for calculating by interpolation equation of the secondary function control data corresponding to the input variable value showing the output characteristics of the non-linearly related to the parameter a processing device table, and an ignition detecting means for detecting the ignition-on at the start of the engine, the ignition O The newly computed coefficients of the interpolation calculation equation for each interstitial by respectively substituting the data value for each three lattice points neighboring the data table the interpolation calculation equation of the secondary function but each time it is detected A coefficient calculating means for setting, and a predetermined routine is started at a predetermined cycle every time the ignition is detected, and the linear interpolation calculating means is set until the coefficient setting of the interpolation formula by the coefficient calculating means is completed. When the coefficient setting of the interpolation calculation expression by the coefficient calculation means is completed, the interpolation execution means for executing the secondary interpolation by the secondary interpolation calculation means instead of the linear interpolation by the linear interpolation calculation means It is characterized by having been provided.
[0006]
In this way, the coefficients of the interpolation arithmetic expression are automatically set at a predetermined time when the processing function of the CPU has room .
[0007]
Further, the data table processing device of the present invention calculates control data for engine control, and the control data is an output characteristic of an air flow meter for measuring an intake air amount, and the output characteristic is non-linear. Things.
[0008]
[Action]
According to the present invention, for example, interpolation between grid points of a data table is performed in an interpolation calculation of control data indicating a nonlinear output characteristic that greatly changes, such as an output characteristic of an air flow meter that measures an intake air amount in engine control. Coefficients of the interpolation formula are automatically set based on data values of a plurality of adjacent grid points, and control data corresponding to an input variable value is interpolated between grid points based on the set coefficients. The automatic setting of the coefficient is performed at the time of starting the engine , which has sufficient processing functions of the CPU.
[0009]
【Example】
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0010]
FIG. 2 is an input / output block diagram of a CPU in the engine control system according to one embodiment of the present invention. In FIG. 2, reference numeral 1 denotes a computer, and 2 denotes a CPU. As shown in the figure, the inputs of the computer 1 are a crank angle signal, an intake air amount signal, a water temperature (cooling water temperature) signal, an intake air temperature signal, and the like, and outputs are an ignition signal defining a control amount of an ignition timing, and a fuel injection amount. (Pulse width of injection pulse) is a fuel injection signal or the like that defines a control amount. Here, the control amounts such as the ignition timing and the fuel injection amount are obtained by adding various corrections to a basic control amount calculated by interpolation from a data table using the engine speed and the intake charge amount as parameters. At this time, the intake air charge is calculated based on the amount of intake air measured by the air flow meter. However, since the output characteristics of the air flow meter vary greatly and are non-linear, the amount of intake air is calculated based on the output voltage. The calculation of the air amount is performed by secondary interpolation. The interpolation calculation of the ignition timing and the fuel injection amount is also performed by secondary interpolation.
[0011]
FIG. 3 is an explanatory diagram schematically illustrating the interpolation operation of this embodiment, with the input variable X on the horizontal axis and the output variable Y on the vertical axis. In the calculation of the airflow meter measurement value, the input variable X is the output voltage of the airflow meter, and the output variable Y is the intake air amount. In the calculation of the ignition timing and the fuel injection amount, the input variable X is the engine speed or the intake charge, and the output variable Y is the pulse width of the fuel injection pulse or the ignition advance amount. FIGS. 4, 5 and 6 are flow charts for executing table interpolation of these air flow meter measurement values and the like. FIG. 4 shows a main routine executed when the ignition is turned on when the engine is started. FIG. 6 shows a table conversion routine for converting grid point data into coefficients of an interpolation operation expression, and FIG. 6 shows a table interpolation routine.
[0012]
The main routine of FIG. 4 is comprised of the steps of M1 to M 4, and started at the same time as the ignition is turned on, first, initializes the system M1. Then, the table conversion end flag is cleared in M2, and the conversion process is started in M3. When the conversion process is completed, a table conversion end flag is set in M4.
[0013]
The conversion process is performed by a table conversion routine shown in FIG. This routine is composed of two steps S1 and S2, starts, and reads data values (X1, Y1), (X2, Y2), (X3, Y3) for every three adjacent grid points in S1. Then, in S2, the simultaneous equations in which the data of the respective grid points are substituted into the quadratic interpolation formula (Y = aX ² + bX + c) are solved to obtain coefficients a, b and c.
[0014]
The table interpolation routine of FIG. 6 is executed at a period of 5 msec to 10 msec. This routine includes steps P1 to P9. When the routine starts, first, an input variable value X is read in P1, and it is checked in P2 whether a table conversion end flag is set. Then, when the table conversion end flag is not set, since the coefficients of the interpolation calculation expression have not been set yet, linear interpolation using normal grid point data is performed as shown by a dotted line in FIG. That is, the process proceeds to P3, in which the input variable value X is guarded against the maximum value and the minimum value. At P4, an address offset value, that is, an absolute address defining a grid point corresponding to the input variable value X is calculated. The offset width of the X from the address offset value (X-axis difference from the corresponding grid point: XX ₁ ) is calculated. Then, the interpolation defining the ratio of the X-axis spacing of the offset width _{(X-X 1)} and longitudinal lattice point _{(X 2 -X 1) ((} X-X 1) / (X 2 -X 1)) at P6 The coefficient is calculated, and at P7, Y is calculated by linear interpolation using an interpolation coefficient based on the data values (Y ₁ , Y ₂ ) of the preceding and following lattice points. When the table conversion end flag is set and the setting of the coefficients of the interpolation equation is completed, the process proceeds to P8, where the coefficients a, b, and c of the interpolation equation are read, and the coefficients a, b, and c are substituted in P9. To obtain Y.
[0015]
In the above embodiment, the coefficients of the interpolation formula are automatically set at the time of starting the engine, which has sufficient processing functions of the CPU, and the coefficients are fixed in the ROM area. However, in an embodiment of the present invention, the coefficient may be updated by learning when the processing function of the CPU has a margin without fixing the coefficient.
[0016]
Further, the present invention is the amount of intake air, ignition timing, Ru can be applied to an engine control other than the fuel injection amount.
[0017]
【The invention's effect】
Since the present invention is configured as described above, it is possible to realize a highly accurate interpolation operation by an interpolation operation expression without complicating the setting at the time of development, and to reduce the ROM capacity of the data table.
[Brief description of the drawings]
FIG. 1 is an overall configuration diagram of the present invention; FIG. 2 is a block diagram showing an embodiment of the present invention; FIG. 3 is an explanatory diagram for explaining an interpolation operation according to an embodiment of the present invention; FIG. Flow chart for executing control of the embodiment (main routine)
FIG. 5 is a flowchart for executing control according to an embodiment of the present invention (table conversion routine);
FIG. 6 is a flowchart for executing control according to an embodiment of the present invention (table interpolation routine);
[Explanation of symbols]
1 computer 2 CPU

Claims (2)

Using a data table that stores control data corresponding to the size of a predetermined parameter, the control data indicating a non-linear output characteristic corresponding to the input variable value related to the parameter, the data corresponding to each input variable value using a linear interpolation calculation means for calculating by linear interpolation based on the data value of the phase adjacent two grid points on the table, the data table also stores control data in response to the magnitude of a predetermined parameter, said parameter A data table processing device provided with secondary interpolation calculation means for calculating control data indicating a non-linear output characteristic corresponding to the input variable value according to the quadratic interpolation formula ,
Ignition detection means for detecting ignition-on at the time of engine start, and each time the ignition-on is detected, substituting data values for each of three adjacent grid points of the data table into the interpolation function of the quadratic function. a coefficient calculating means for newly calculated set the coefficients of the interpolation calculation equation for each interstitial by,
A predetermined routine is started at a predetermined cycle every time the ignition is detected, and linear interpolation is performed by the linear interpolation calculating means until the coefficient setting of the interpolation calculation expression by the coefficient calculating means is completed. Interpolation executing means for executing secondary interpolation by the secondary interpolation calculating means in place of linear interpolation by the linear interpolation calculating means when coefficient setting of the interpolation arithmetic expression by the coefficient calculating means is completed. Data table processing device. 2. A data table processing device for calculating control data for engine control, wherein the control data is an output characteristic of an air flow meter for measuring an intake air amount, and the output characteristic is non-linear. Data table processing device.

Images