JP3644350B2 - Automotive electronic control device with floating point arithmetic function - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention has a floating point arithmetic functionFor automobileThe present invention relates to an electronic control device.
[0002]
[Prior art]
In general, an electronic control unit (ECU) applied to automobile engine control or the like has conventionally performed various calculations using integer type data (fixed point type data), but in recent years, a floating point arithmetic processor (FPU) has been used. : Floating-Point Unit), it is now possible to perform operations with floating-point data. According to the floating-point type data, the operation result can be obtained with very fine accuracy compared to the fixed-point type data.
[0003]
The floating-point type data is configured in accordance with, for example, the IEEE 754 standard, and includes a 1-bit sign part, an 8-bit exponent part, and a 23-bit mantissa part as shown in FIG. Thus, in the case of 4-byte (single-precision storage format) floating-point data having a mantissa part of 23 bits, it has a resolution of 7 digits (0.0000001 resolution).
[0004]
FIG. 10B shows a bit pattern in a single-precision storage format. Floating-point type data is a normalized number, a denormalized number, an infinite number, zero, and a combination of an exponent part and a mantissa part. A distinction is made between numbers. Here, a non-numeric value represents a numerical value, and a non-numeric value represents a non-numeric value. For example, when a calculation result that cannot be expressed as a numerical value such as 0/0 or + ∞−∞ is represented, a non-number is used.
[0005]
[Problems to be solved by the invention]
As described above, floating point type data has a data format of non-numeric, and for example, the results of arithmetic operations including non-numerics are all non-numeric and invalid. For example, in the comparison operation, when comparing whether the non-number is 1 or more, the result is false. Therefore, when a non-number occurs in the electronic control unit, there is a problem that the calculation result (output value) cannot be guaranteed at all.
[0006]
There are mainly two conditions for generating non-numbers in engine control. The first condition is when the floating point type RAM value is rewritten due to noise during operation of the electronic control unit or during battery backup, and the RAM value itself changes to an innumerable number. For example, there is a case where the floating-point RAM value is rewritten to FFFFFFFFh (all bits “1”) due to noise. The second condition is a case where an argument used for floating-point arithmetic is rewritten due to noise or the like, and an arithmetic operation such as 0/0 is performed to generate a secondary number.
[0007]
Here, the case will be described in the calculation processing of the engine speed shown in FIG. That is, in the process of FIG. 11, the time T360 required for the engine to rotate at 360 ° CA (crank angle) is calculated and stored as the FR0 value (step 900). 1 sec "is set in FR1, and" 60 "is set in FR2 (steps 910 to 930).
[0008]
Then, formula
FR0 = FR1 / FR0 × FR2
Thus, the FR0 value is calculated (step 940), and the FR0 value is set as the engine speed Ne (step 950).
[0009]
In the process of FIG. 11, it is assumed that the FR0 and FR1 values change to 0 due to noise or the like immediately after step 930. In this case, the calculation of “0/0” is performed in step 940, and the FR0 value that is the calculation result of step 940 becomes a non-number. As a result, the engine speed cannot be calculated correctly.
[0010]
Further, for example, the calculation formula of the fuel injection amount f is expressed as the following formula.
f = Fbase × fHL
Here, Fbase is a basic injection amount calculated as integer type data, and fHL is a load correction amount calculated as floating point type data. In this calculation, when the load correction amount fHL is non-numeric, the value of the calculated fuel injection amount f is non-numeric, and normal fuel injection cannot be performed.
[0011]
  The present invention has been made paying attention to the above-mentioned problem, and the object of the present invention is to prevent a control failure caused by the occurrence of a non-number in an electronic control device having a floating-point arithmetic function. CanFor automobileAn electronic control device is provided.
[0012]
[Means for Solving the Problems]
  According to the first aspect of the present invention, it is determined whether or not the floating-point type data is non-numeric by the non-numeric determination means. Instead of calculationFurthermore, the map data is interpolated using the value calculated as integer type data as a parameter, and the interpolation value obtained as a result is used as a backup value.Backup processing is performed. That is, in floating point arithmetic, all the arithmetic results performed on data including a non-number are all non-numbers, so that correct control data cannot be obtained. However, in the present invention, floating-point type data is non-numbers. In this case, a backup process is performed in place of the floating point arithmetic. As a result, it is possible to prevent a control failure accompanying the occurrence of a non-number.In particular, as backup processing executed by the backup means, instead of floating point type data, map data is interpolated using a value calculated as integer type data as a parameter, and the interpolation value obtained thereby is used as a backup value. Therefore, the accuracy is inferior to that of control data obtained by floating-point arithmetic, but control can be continued without interruption when a non-number occurs. That is, in this case, the accuracy of the calculation result is inferior to that of the floating point calculation using the floating point type data, but the control data (interpolation value) corresponding to the integer type data can be calculated as a backup value, so the controllability is deteriorated. Can be suppressed.
[0013]
According to the second aspect of the present invention, whether or not the floating-point data used at that time is a non-numeric value is determined for each floating-point operation, so that the floating-point operation at that time is executed without error. The That is, it is possible to reliably determine a floating-point operation that becomes invalid due to an operation including a non-number.
[0014]
According to the third aspect of the present invention, it is determined whether or not the floating-point data that affects each floating-point operation is non-numeric. In this case, the floating-point data that affects the floating-point arithmetic is, for example, data that is used as a precondition for the floating-point arithmetic. By performing the backup process, it is possible to avoid the inconvenience that the floating point arithmetic is erroneously performed and the controllability is deteriorated.
[0015]
According to the fourth aspect of the present invention, it is determined whether any one of all the floating point type data is non-numeric. That is, when any one of the floating-point type data is a non-numeric value, it can be determined that the microcomputer is operating in an environment where the generation factor (specifically, noise) is present. Therefore, at that time, since it is not wrong when a non-number occurs, all floating point operations are prohibited, and a backup process in place of the floating point operations is performed.
[0016]
According to the fifth aspect of the present invention, if the floating-point type data obtained by the floating-point operation is a non-number, a non-number determination flag indicating that is set, and the non-number determination means sets the non-number Whether the number is non-numeric is determined by referring to the determination flag. Then, backup processing is performed based on the determination result of the non-number determination flag. In this case, it is determined whether the same floating point type data is used for a plurality of floating point operations or whether any one of all the floating point type data is non-numeric as described in claim 4. If it is applied to the case, the processing can be simplified, which is practically preferable.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an outline of an engine control system. In this system, the vehicle-mounted engine 1 is configured as, for example, a gasoline injection type multi-cylinder internal combustion engine.
[0020]
The engine control ECU 10 includes a microcomputer (hereinafter referred to as a microcomputer) 11, which includes a central processing unit (CPU) 12, a readable / writable storage device (RAM) 13, and a read-only storage device (ROM) 14. A floating point arithmetic processor (FPU) 15 and an input / output device (I / O) 16 are provided. Here, the FPU 15 performs a calculation in the floating point format, and the CPU 12 performs a calculation other than the floating point format. The I / O 16 includes a well-known A / D converter.
[0021]
Various types of information representing the engine operating state are input to the ECU 10 from a sensor group provided in the engine 1. The sensor group includes, for example, a rotational speed sensor 2 for detecting the engine rotational speed, an intake pressure sensor 3 for detecting the intake pipe pressure, a water temperature sensor 4 for detecting the cooling water temperature, and an air-fuel ratio from the oxygen concentration in the exhaust gas. It comprises an A / F sensor 5 for detecting (A / F). Then, the ECU 10 performs fuel injection control by an injector (not shown), ignition timing control by an ignition device, and the like based on the input sensor signal.
[0022]
As described above, the floating-point data calculated by the FPU 15 is configured according to, for example, the IEEE 754 standard, and has a data format shown in FIG. Further, as shown in FIG. 10B, the floating point type data is classified into a normalized number, a denormalized number, an infinite number, zero, and a non-number according to the combination of the exponent part and the mantissa part.
[0023]
In the present embodiment, when the result of the floating point calculation (floating point type data) by the FPU 15 becomes a non-number, it is determined so and the floating point calculation is not performed. Specifically, the process for performing the floating point calculation using the FPU 15 includes a fuel injection amount calculation process and an ignition timing calculation process. For example, the fuel injection quantity f is calculated as follows.
[0024]
f = Fbase × fHL
Here, Fbase is a basic injection amount calculated as integer type data, and fHL is a load correction amount calculated as floating point type data. Therefore, when calculating the fuel injection amount f using the FPU 15, if the load correction amount fHL is a non-number, the calculation result (the value of the injection amount f) will also be a non-number. For this reason, in the present embodiment, when the load correction amount fHL is non-numeric, the calculation of the fuel injection amount f using the FPU 15 is prohibited, and the fuel injection amount f is obtained by backup processing instead. ing.
[0025]
Here, among the processes executed by the CPU 12, the calculation process of the load correction amount fHL will be described with reference to FIG. 2, and the fuel injection amount calculation process will be described with reference to FIG. 2 is a time synchronization process performed every predetermined time, and the process of FIG. 3 is a crank synchronization process performed every predetermined crank angle based on the pulse signal of the rotation speed sensor 2. .
[0026]
As shown in FIG. 2, the CPU 12 determines whether or not the engine speed is 3000 rpm or more at step 100, and proceeds to step 110 if it is 3000 rpm or more. In step 110, the CPU 12 determines whether or not the engine water temperature is 60 ° C. or higher. If it is 60 ° C. or higher, the CPU 12 proceeds to step 120. In step 120, the FPU 15 is used to determine the load correction amount fHL as in the following equation.
[0027]
fHL = fNe × MAF × fGain
Here, fNe is the engine speed calculated as floating point type data based on the pulse signal of the speed sensor 2, and MAF is the detected value of the intake pressure sensor 3, and is the intake pipe pressure (intake pressure). Is a voltage value of integer type data based on FGain is a constant set with floating point type data.
[0028]
In this way, as a precondition for load correction, when the condition that the rotational speed is 3000 rpm or higher and the water temperature is 60 ° C. or higher is satisfied, the FPU 15 performs the floating point calculation and calculates the load correction amount fHL of the floating point type data. Is done.
[0029]
On the other hand, if the rotation speed is less than 3000 rpm or the water temperature is less than 60 ° C., a negative determination is made in step 100 or step 110, and in step 130, the CPU 12 is floating point type data as a fixed value for the load correction amount fHL. After substituting “1.0”, this process is terminated.
[0030]
Next, the injection amount calculation process will be described with reference to FIG.
First, in step 200, the CPU 12 calculates a basic injection amount Fbase of integer type data based on the engine speed and the intake pressure, and then in step 210, determines whether or not the load correction amount fHL is non-numeric. judge. That is, as described above, the load correction amount fHL is obtained as floating point type data, and the bit 30 to 23 of the value of the correction amount fHL is “11111111” and the bit 22 to of the value of the correction amount fHL is It is determined whether any one of 0 is “1”. If a negative determination is made in step 210, the CPU 12 proceeds to step 220 and calculates the fuel injection amount f by using the FPU 15 by the floating point calculation of the following equation.
[0031]
f = Fbase × fHL
That is, the fuel injection amount f is calculated by multiplying the basic injection amount Fbase by the load correction amount fHL, and the value is calculated as floating point type data. Thereafter, in step 230, the CPU 12 determines whether or not the calculated value is non-numeric, and if it is determined that the calculated value is not non-numeric, the value of the fuel injection amount f obtained by the floating point calculation is correct. This processing is terminated.
[0032]
On the other hand, if an affirmative determination is made in step 210 or step 230, the CPU 12 proceeds to step 240, and after substituting the basic injection amount Fbase of the integer type data into the fuel injection amount f without using the FPU 15, this processing is performed. finish.
[0033]
Next, the calculation process of the ignition timing e will be described with reference to FIG. Note that the load correction amount eHL used in the process is also floating point type data obtained by floating point calculation in the same manner as the load correction amount fHL in the calculation process of the fuel injection amount f.
[0034]
First, in step 300, the CPU 12 calculates a basic ignition timing Ebase of integer type data based on the engine speed and the intake pressure, and in subsequent step 310, determines whether or not the load correction amount eHL is non-numeric. judge. If a negative determination is made in step 310, the CPU 12 proceeds to step 320 and calculates the ignition timing e using the FPU 15 by the following floating point calculation.
[0035]
e = Ebase x eHL
That is, the ignition timing e is calculated by multiplying the basic ignition timing Ebase by the load correction amount eHL, and the value is calculated as floating point type data. Thereafter, in step 330, the CPU 12 determines whether or not the calculated value is non-numeric, and if it is determined that the calculated value is not non-numeric, the value of the ignition timing e obtained by the floating point calculation is assumed to be correct. This process ends.
[0036]
On the other hand, if an affirmative determination is made in step 310 or step 330, the CPU 12 proceeds to step 340 and completes the process after substituting the basic ignition timing Ebase of the integer data into the ignition timing e without using the FPU 15. To do.
[0037]
In the present embodiment, the processing in steps 210 and 230 in FIG. 3 and the processing in steps 310 and 330 in FIG. 4 correspond to non-numerical determination means, and the processing in step 240 in FIG. 3 and step 340 in FIG. It corresponds to.
[0038]
According to the embodiment described in detail above, the following effects can be obtained.
(1) When floating-point type data (load correction amounts fHL, eHL) used for floating-point arithmetic is non-numeric, backup processing in place of the floating-point arithmetic is performed. In this case, the result of four arithmetic operations including non-numeric values in floating-point arithmetic is invalid and invalid, but when floating-point data used for floating-point arithmetic is non-numeric, the invalid floating-point arithmetic is prohibited. can do. In this way, it is possible to avoid the fuel injection amount f and the ignition timing e as control data from becoming innumerable, and it is possible to prevent a control failure caused by the occurrence of the innumerable number.
[0039]
(2) In the present embodiment, as a backup process in the case of occurrence of an innumerable number, the basic injection amount Fbase of integer type data is substituted for the fuel injection amount f, and the basic ignition timing Ebase of integer type data is substituted for the ignition timing e. I made it. That is, a value that does not hinder the control is substituted as a backup value of the control data. In this way, an increase in ROM capacity associated with the backup process can be suppressed.
[0040]
The second to fourth embodiments of the present invention will be described below. However, the configuration of the ECU 10 in the second to fourth embodiments is the same as that of the first embodiment shown in FIG. 1, and the processing executed by the CPU 12 is different from that of the first embodiment.
[0041]
(Second Embodiment)
A second embodiment of the present invention will be described with reference to FIGS.
In the first embodiment, it is determined whether or not the floating-point type data (load correction amount fHL) used for the floating-point arithmetic is non-numeric. It is determined whether or not the floating-point data that affects the operation is not a number, rather than the floating-point data used in the above. Further, in the present embodiment, the load correction amount fHL is calculated by a different arithmetic expression at the time of acceleration and deceleration, and the determination of acceleration or deceleration is made by a parameter fDL corresponding to the load change amount. Has been. The parameter fDL is floating point type data calculated based on the engine speed and the intake pressure. If the parameter fDL is a positive value, it means acceleration, and if it is a negative value, it means deceleration.
[0042]
Here, the calculation process of the load correction amount fHL in the present embodiment will be described with reference to FIG. The flowchart in FIG. 5 is a modification of part of FIG. 2, and steps 100, 110, and 130 in FIG. 5 show common processing.
[0043]
In short, when any of the conditions that the engine speed is less than 3000 rpm or the water temperature is less than 60 ° C. is satisfied (when a negative determination is made in step 100 and step 110), the load correction amount fHL is set to a fixed value in step 130. “1.0” which is floating point type data is substituted and this processing is terminated.
[0044]
On the other hand, when the engine speed is 3000 rpm or higher and the water temperature is 60 ° C. or higher, in step 400, the CPU 12 determines whether or not the parameter fDL is a non-number. If it is determined that the parameter fDL is not a non-numeric value, the CPU 12 proceeds to step 410 and determines whether the parameter fDL is to be accelerated or decelerated using the parameter fDL. Here, if the parameter fDL is a value larger than “0”, the process proceeds to step 420 and the FPU 15 is used to perform the floating point calculation at the time of acceleration (fNe × MAF × fGain1), and the load correction amount of the floating point type data After calculating fHL, this process is terminated. On the other hand, if the parameter fDL is a value equal to or less than “0”, the process proceeds to step 430 and the FPU 15 is used to perform the floating point calculation at the time of deceleration (fNe × MAF × fGain2) to obtain the load correction amount fHL of the floating point type data. After the calculation, this process is terminated. Note that fGain1 is an acceleration constant set with floating-point type data, and fGain2 is a deceleration constant set with floating-point type data.
[0045]
On the other hand, if it is determined in step 400 that the parameter fDL is non-numeric, the process proceeds to step 440, where the fixed value “1” of the integer type data is substituted for the load correction amount FHL, and then the present process is terminated. To do. Here, the load correction amount FHL of integer type data is stored in a storage area secured in the RAM 13 separately from the load correction amount fHL of floating point type data.
[0046]
Next, the injection amount calculation process in the present embodiment will be described with reference to FIG. The flowchart in FIG. 6 is obtained by changing a part of FIG. 3, and steps 200, 220, 230, and 240 in FIG. 6 show common processes.
[0047]
That is, after calculating the basic injection amount Fbase in step 200, it is determined in step 500 whether or not the parameter fDL corresponding to the load change amount is non-numeric. If it is determined that the parameter fDL is non-numeric, the CPU 12 determines in step 510 that the load correction amount FHL and the basic injection amount Fbase into which the fixed value “1” of the integer data is substituted without using the FPU 15. Is multiplied by the fuel injection amount f (= Fbase × FHL).
[0048]
On the other hand, if it is determined in step 500 that the parameter fDL is not a non-numeric value, the CPU 12 calculates a fuel injection amount f (= Fbase × fHL) by a floating point calculation using the FPU 15 in step 220. Then, the CPU 12 determines in step 230 whether or not the calculated value is non-numeric. If it is not non-numeric, the process ends. If it is non-numeric, the fuel injection amount f is set in step 240. After substituting the basic injection amount Fbase, this process ends.
[0049]
In the present embodiment, the processing in step 400 in FIG. 5, step 500 in FIG. 6 and step 230 in FIG. 6 corresponds to the non-numerical determination means, and the processing in step 440 in FIG. 5, step 510 in FIG. Corresponds to backup means.
[0050]
In short, if floating-point data that affects floating-point arithmetic, that is, floating-point type data used as a precondition for floating-point arithmetic, is not a number, the precondition cannot be determined correctly and floating-point arithmetic is erroneously performed. There is a risk of being. On the other hand, in this embodiment, when it is determined that the floating-point type data used as the precondition is non-numeric, a process in place of the floating-point operation is performed. As a result, it is possible to avoid the inconvenience that the controllability of the engine 1 is deteriorated by erroneously performing floating point arithmetic and calculating inaccurate control data.
[0051]
(Third embodiment)
Next, a third embodiment of the present invention will be described with reference to FIGS.
In this embodiment, when any one of all the floating point type data becomes a non-numeric value, all the floating point operations are prohibited and a backup process is performed instead.
[0052]
As a specific example, first, the calculation process of the fuel injection amount f will be described with reference to FIG. The flowchart in FIG. 7 is obtained by changing a part of FIG. 3, and steps 200, 220, 230, and 240 in FIG. 7 show common processes.
[0053]
In FIG. 7, the non-number determination flag XALL is used, and the flag XALL is set to “1” when any one of all the floating-point type data becomes non-number. For example, in the floating point arithmetic processing (not shown) for calculating the load correction amount fHL, eHL, etc., when the calculated value of the floating point arithmetic is non-numeric, “1” is set to the non-numeric determination flag XALL. The
[0054]
Referring to the flowchart of FIG. 7 in detail, after calculating the basic injection amount Fbase in step 200, the CPU 12 proceeds to step 600 and determines whether or not the non-number determination flag XALL is “1”. When it is determined that the non-number determination flag XALL is “0”, the CPU 12 proceeds to step 220 and calculates the fuel injection amount f (= Fbase × fHL) by the floating point calculation using the FPU 15. At the same time, it is determined in step 230 whether or not the calculated value is non-numeric. Then, when it is determined in step 230 that the calculated value is not a non-numeric value, the present process is terminated. set. Then, the process proceeds to step 240, and after substituting the basic injection amount Fbase as the fuel injection amount f, this processing is terminated.
[0055]
Further, when “1” is set to the non-number determination flag XALL, an affirmative determination is made in step 600, and after the basic injection amount Fbase is substituted as the fuel injection amount f in step 240, this processing is terminated. In other words, when “1” is set in the non-number determination flag XALL, the process of step 240 is performed as a backup process instead of the floating point calculation, bypassing the floating point calculation of step 220.
[0056]
Next, the calculation process of the ignition timing e will be described with reference to FIG. The flowchart of FIG. 8 is obtained by changing a part of FIG. 4, and steps 300, 320, 330, and 340 of FIG. 8 show common processes.
[0057]
Also in the processing of FIG. 8, before performing the floating-point operation in step 320, it is determined in step 700 whether or not the non-number determination flag XALL is “1”. Then, when it is determined that the non-numeric determination flag XALL is “1”, the floating point calculation in step 320 is bypassed, and the basic timing is set to the ignition timing e in step 340 as a backup process in place of the floating point calculation in step 320. Set the ignition timing Ebase. Also, when it is determined that the value calculated by the floating-point operation in step 320 is non-numeric, the process proceeds to step 710 and the non-numeric determination flag XALL is set to “1”, and then the processing of step 340 is performed.
[0058]
That is, in this embodiment, when it is determined that the calculation result is non-numeric in all floating-point arithmetic used for engine control, the common non-numeric determination flag XALL is set. Then, for the floating point operation, before the operation is performed, it is determined whether or not the non-numeric determination flag XALL is set. If the flag XALL is set, all floating point operations are prohibited. Therefore, backup processing is implemented instead.
[0059]
In the present embodiment, the processing in step 230, step 600, FIG. 8, step 330, and step 700 in FIG. 7 corresponds to the non-number determination means, and the processing in step 240 in FIG. 7 and step 340 in FIG. Corresponds to backup means.
[0060]
Actually, when any one of the floating point type data becomes non-numeric, it can be determined that the microcomputer 11 is operating under an environment where the generation factor (specifically, noise) exists. Therefore, at that time, since it is not wrong when a non-number occurs, all floating point operations are prohibited, and a backup process in place of the floating point operations is performed.
[0061]
In the present embodiment, since the non-decision determination flag XALL is used to determine whether or not it is non-numeric, it is determined whether or not all floating-point data are non-numeric depending on the value. Compared with the case of determination, the process can be simplified and is practically preferable.
[0062]
As described above, when any floating point data has a non-number, and the flag XALL is set accordingly, the non-number generation factor such as noise is eliminated, and the floating-point data returns to a normal value. The flag XALL may be reset to resume the original floating point operation. As a result, when a non-number is temporarily generated due to noise or the like, the backup process is not continued for a long time, and the high-precision floating point calculation can be restored.
[0063]
(Fourth embodiment)
Next, a fourth embodiment of the present invention will be described with reference to FIG. The flowchart in FIG. 9 is obtained by changing a part of FIG. 2, and steps 100 to 130 in FIG. 10 show common processes.
[0064]
More specifically, when any of the conditions that the engine speed is less than 3000 rpm and the water temperature is less than 60 ° C. is satisfied (when a negative determination is made in step 100 and step 110), the load correction amount fHL is fixed at step 130. As a result, “1.0” which is floating point type data is substituted, and this processing ends. On the other hand, when the engine speed is 3000 rpm or higher and the water temperature is 60 ° C. or higher (when both step 100 and step 110 are affirmatively determined), the CPU 12 proceeds to step 800 to calculate the engine calculated as floating point type data. It is determined whether or not the rotation speed fNe is a non-number. When it is determined that the engine speed fNe is not a non-numeric value, in step 120, a floating point calculation is performed using the FPU 15, and a load correction amount fHL (fNe × MAF × fGain) of the floating point type data is calculated. After this, this processing is finished. On the other hand, when determining that the engine speed fNe is a non-number, the CPU 12 determines in step 810 that the engine speed Ne calculated as integer type data and the intake pressure MAF of the integer type data without using the FPU 15. The map data is interpolated using as a parameter. Then, the CPU 12 converts the interpolated value obtained as integer type data into the load correction amount fHL as floating point type data, and ends this processing.
[0065]
In the present embodiment, the process at step 800 corresponds to the non-number determination means, and the process at step 810 corresponds to the backup means.
As described above, in this embodiment, when it is determined that the floating-point data is non-numeric, an operation using integer data is performed instead of the floating-point operation. In this case, although the accuracy of the operation result is inferior to that of the floating point operation, the control data corresponding to the integer type data can be calculated as a backup value, which is preferable in practice.
[0066]
In addition to the above, the present invention can be embodied in the following forms.
In the first embodiment, in step 210 of FIG. 3, it is determined whether or not it is a non-numeric value based on the load correction amount fHL of the floating point type data. Good. That is, if the calculated value of the floating point calculation in step 120 in FIG. 2 is non-numeric, a non-numeric determination flag indicating that the load correction amount fHL is non-numeric is set, and in step 210 in FIG. It is determined whether or not the load correction amount fHL is non-numeric by referring to the flag. Similarly, in the processing of step 310 in FIG. 4, step 400 in FIG. 5 and the like, it may be determined whether or not the floating-point type data is non-numeric using a non-numeric determination flag. In this case, it is possible to determine whether or not the floating-point type data is non-numeric by referring to the non-numeric determination flag. Therefore, if the same floating point type data is used for a plurality of floating point operations, the processing can be simplified, which is practically preferable.
[0067]
In the second embodiment, the parameter fDL is embodied as the floating-point type data that affects the floating-point arithmetic. However, the present invention is not limited to this. For example, when using floating-point type data such as an air-fuel ratio correction amount as a precondition for performing floating-point arithmetic, it is determined whether the air-fuel ratio correction amount is non-numeric. Prohibit floating point operations and perform backup processing instead.
[0068]
Also, for example, to determine whether floating-point data that affects each floating-point operation is non-numeric, when non-numeric occurs, it is divided into various types such as ignition control and injection control. Make a decision. Specifically, a common non-numeric determination flag is prepared for each type of control such as ignition control and injection control. For example, when a non-numeric value is generated in a floating point calculation of ignition control, a non-numeric value for ignition control is used. A determination flag is set, and all floating point calculations performed in the ignition control are prohibited using the non-number determination flag. Even in this case, it is possible to prevent the floating point arithmetic from being adversely affected by the occurrence of the non-number.
[0069]
In each of the above-described embodiments, the electronic control unit (ECU) 10 that handles floating-point type data in the single-precision storage format is exemplified, but the present invention can also be applied to an electronic control unit that handles floating-point type data in the double-precision storage format.
[0070]
In the above description, the present invention has been described by taking automobile engine control as an example. However, by applying the present invention to an electronic control apparatus that is involved in the traveling of the automobile as described above, the reliability of the automobile control system can be improved. it can.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an overview of an engine control system in an embodiment of the invention.
FIG. 2 is a flowchart showing a load correction amount calculation process in the first embodiment.
FIG. 3 is a flowchart showing an injection amount calculation process in the first embodiment.
FIG. 4 is a flowchart showing an ignition timing calculation process in the first embodiment.
FIG. 5 is a flowchart showing a load correction amount calculation process in the second embodiment.
FIG. 6 is a flowchart showing an injection amount calculation process in the second embodiment.
FIG. 7 is a flowchart showing an injection amount calculation process in the third embodiment.
FIG. 8 is a flowchart showing ignition timing calculation processing in the third embodiment.
FIG. 9 is a flowchart showing a load correction amount calculation process in the fourth embodiment.
FIG. 10 is a diagram showing a configuration of floating point type data.
FIG. 11 is a flowchart showing a rotational speed calculation process.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 ... Engine control ECU (electronic control apparatus), 11 ... Microcomputer, 12 ... CPU, 15 ... FPU (floating point arithmetic processor).

Claims (5)

In an automotive electronic control device having a floating point arithmetic function for calculating floating point type data and performing various controls according to a predetermined control program,
Non-numeric determination means for determining whether floating-point data is non-numeric;
When that is not a number is determined by the a-number determining means and the cash strong point to floating-point operations, carried out interpolation calculation map data the calculated value as a parameter as an integer data, obtained by this Backup means for performing backup processing using the interpolation value as a backup value ;
An automobile electronic control device comprising: 2. The automotive electronic control device according to claim 1, wherein the non-number determining means determines whether or not the floating-point type data used at that time is a non-number for each floating-point operation. 2. The automotive electronic control device according to claim 1, wherein the non-numeric determination means determines whether or not the floating-point type data that affects each floating-point operation is non-numeric. 2. The electronic control apparatus for an automobile according to claim 1, wherein the non-number determining means determines whether any one of all floating-point data is a non-number. If the floating-point type data is non-numeric, a non-numeric judgment flag indicating that is set each time,
The non-number determination means determines whether the number is non-number with reference to a non-number determination flag, and the backup means includes:
The automotive electronic control device according to any one of claims 1 to 4, wherein backup processing is performed according to a determination result of the non-number determination flag.

Images