JP2009067071A - Electronic control unit - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To guarantee identity of a physical quantity conversion value between the same data and simplify processing of a procedure for guaranteeing the identity, in an electronic control device handling the same data constituted of two types of a fixed-point type and a floating-point type. <P>SOLUTION: A microcomputer 30 in an engine control ECU (Electronic Control Unit) 10 is provided with a CPU (Central Processing Unit) 31, a FPU (Floating-Point Unit) 32, a RAM (Read Access Memory) 33, a ROM (Read Only Memory) 34 and a peripheral module 35. The CPU 31 executes an operation by fixed-point form data, while the FPU 32 executes an operation of floating-point form data. In the RAM 33, the same data converted from pulse width correction data of the fixed-point type stored in the ROM 34 into the floating-point type. When an operation is performed by using pulse width correction data in the CPU 31, the operation is executed by referring to data stored in the ROM 34. When an operation is performed by using the pulse width correction data, the operation is executed by referring to data stored in the RAM 33. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an electronic control device (hereinafter also referred to as “ECU”) having a floating point arithmetic function.

  In recent years, electronic control units (ECUs) applied to automobile engine control have been able to perform calculations based on floating-point type data by introducing a floating-point processor (FPU). As a result, it may be necessary to store two types of floating point type data and fixed point data type in the ECU for the same data.

  In an electronic control unit incorporating floating-point arithmetic, if both fixed-point type data and floating-point type data exist, if the data is stored in a RAM value, the fixed-point type data is converted to floating-point type data. Used for floating-point arithmetic. Or, conversely, a technique is used in which floating-point type data is converted to a fixed-point type.

  For example, as disclosed in Japanese Patent Application Laid-Open No. 2001-195233 (Patent Document 1), fixed-point and floating-point type conversion is performed every time the same data is calculated in advance, and the data is divided into storage areas. A method of storing data in each memory area in one data type has been considered.

JP 2001-195233 A

  However, in the above conventional technique, it is necessary to re-store data corresponding to the same physical quantity in two different areas of the fixed-point type data area and the floating-point type data area every time a data change occurs. Further, when the data is ROM area data, it is necessary to store the same physical quantity value in each of the two ROM areas.

  In view of the above problems, the present invention provides a fixed-point ROM by arranging fixed-point data in the ROM area and floating-point data in the RAM area when the same data is ROM area data. When the area data is changed, only one part of the ROM area is rewritten, so that the floating-point type data is calculated by the type conversion process in the microcomputer, and two data types that are different from the ROM area are calculated. The purpose is to reduce the mechanism for guaranteeing the data identity necessary for storing the same data.

  The present invention relates to a fixed-point arithmetic function for calculating fixed-point type data, a floating-point arithmetic function for calculating floating-point type data, a read-only memory (ROM) area for storing fixed-point type data, and floating-point type data In an electronic control device including a microcomputer that includes a random access memory (RAM) area that stores various types of control according to a predetermined control program, floating point type data stored in the random access memory (RAM) area is read The fixed-point data stored in the only memory (ROM) area is converted to a floating-point data and is the same data of different types.

  According to the present invention, it is no longer necessary to store two data corresponding to the same physical quantity in a read-only memory (ROM) area. For this reason, when changing data, the mechanism for storing the same physical quantity in two fixed-point and floating-point storage areas and the mechanism for confirming correct storage in the two data areas are reduced. This makes it possible to eliminate the need for a mechanism outside the microcomputer for guaranteeing the simultaneity of the two areas.

  A further feature of the present invention is that conversion from fixed-point data to floating-point data is performed using a conversion program.

  A further feature of the present invention is that conversion from fixed-point data to floating-point data and storage of the converted floating-point type data in a random access memory (RAM) area is performed during initialization processing of the microcomputer. It is in.

  A further feature of the present invention is that the conversion from fixed-point data to floating-point data and the storage of the converted floating-point data in a random access memory (RAM) area are performed at any time in the microcomputer. It is characterized by that.

  According to the above-described invention, in an electronic control apparatus that handles the same data consisting of two types of fixed-point and floating-point types, a mechanism for reducing the computation load and guaranteeing data identity between the two data types Can be simplified.

  Hereinafter, embodiments embodying the present invention will be described.

  FIG. 1 is a block diagram showing an outline of an embodiment of the present invention.

  The engine ECU 10 is connected to the engine 20 and processes various input information from the engine 20 and outputs fuel injection and the like. The microcomputer 30 in the engine control ECU 10 includes a CPU 31, an FPU 32, a random access memory area (RAM) 33, a read only memory area (ROM) 34, and a peripheral module 35.

  The CPU 31 has a calculation function for performing calculations with fixed-point model data. The FPU 32 has a calculation function for performing calculation of floating point type data. The ROM 34 stores fixed-point pulse width correction data necessary for operating the engine 20.

  The RAM 33 stores the same data obtained by converting the fixed-point pulse width correction data stored in the ROM into the floating-point type. Conversion of data from the fixed-point type to the floating-point type is performed in software by the CPU 31 or the like using a conversion program.

  The conversion program is supported by the microcomputer. The execution processing of the data conversion instruction such as the memory transfer instruction and the data conversion instruction is shown.

  As an example, the peripheral module 35 includes an input module such as an A / D converter, an I / O that outputs a fuel pulse, and the like.

  The floating-point data used for computation by the FPU 32 is configured according to, for example, the IEEE 745 standard, and has a 1-bit sign part, an 8-bit exponent part, and a 23-bit mantissa part as shown in FIG. Such a storage format of 4-byte data is defined as single-precision floating-point type data, and has a resolution of 7 digits (0.0000001) after the decimal point.

  FIG. 3 shows an example of the data arrangement of the RAM 33 and ROM 34, and FIG. 4 is a diagram showing data specifications shown in FIG.

  The pulse width correction data stored in the ROM 34 is expressed by 2 bytes, and is expressed by physical quantity conversion by the LSB shown in FIG. On the other hand, the pulse width correction data stored in the RAM 33 is a floating point type represented by 4 bytes. Note that the floating point type has no concept of LSB, and indicates a physical quantity as it is.

  FIG. 5 shows a flowchart for performing a type conversion process from fixed-point data to floating-point data.

  By supplying power to the engine ECU 10, the microcomputer 30 performs its own initialization process 100. The initialization process 100 performs an initialization process 101 for peripheral modules such as I / O and an initialization process 102 for the RAM 33.

  After the RAM initialization process 102, a type conversion process 103 from fixed-point data to floating-point data is performed. The type conversion process 103 multiplies the fixed-point type pulse width correction data value by the LSB and converts it to a floating-point type. As an example, fixed-point data 128 (decimal number) is converted as floating-point data 0.5.

  When the floating point data converted by the RAM storage processing 104 of the floating point type conversion data is stored in the RAM 33 and the calculation using the pulse width correction data is performed by the FPU 31 in the background processing, the RAM 33 is stored. Use the stored pulse width correction data.

  On the other hand, when the CPU 31 performs the calculation using the pulse width correction data, the data stored in the area of the ROM 33 is used.

  This configuration is useful when the CPU 33 and the FPU 32 perform calculations using the same ROM data.

As described above, according to the described embodiment, the following technical effects can be obtained.
(I). When it is necessary to use the same data for CPU and FPU calculations, fixed-point data used for calculation by the CPU is stored in the ROM area, and floating-point data used for calculation by the FPU is stored in the RAM area. The floating-point data stored in the RAM is generated by performing type conversion from the fixed-point data stored in the ROM area. Therefore, the identity of the physical quantity converted value in each type is the microcomputer. Guaranteed internally.
(B). As described above, the identity of the physical quantity converted values of the fixed-point data and the floating-point data is guaranteed by the processing inside the microcomputer, so that it is not necessary to store the two types of data in the two ROM areas. When it is necessary to change the same data, it is possible to change the floating-point data only by changing the fixed-point data, and it is possible to simplify the processing of the procedure for guaranteeing the identity of the physical value converted values.

That is, by changing only the fixed-point data in the ROM area, the change in the floating-point data is processed by software by the conversion program and stored in the RAM area, so that the physical value converted value of the changed data is identical. And the process of the procedure for guaranteeing the identity is simplified.
(C). In addition, by performing type conversion from fixed-point data to floating-point data within the initialization processing of the microcomputer, it is not necessary to perform type conversion every time the FPU is used for calculation, and the calculation load can be reduced. .

  In addition to the above, the present invention can be realized in the following forms.

  In the above embodiment, as shown in FIG. 6, in addition to initialization, type conversion from fixed-point data to floating-point type data may be performed by a background processing 105 in a fixed execution task every 100 ms. Also, the task can be realized at any time including execution tasks other than every 100 ms.

  Although the above embodiment has been described as an example regarding the processing of single precision floating point data, it can be similarly applied to an electronic control apparatus that handles double precision floating point arithmetic data.

It is a block diagram which concerns on the Example of this invention and shows the structure of an electronic controller. FIG. 4 is a data structure diagram of a single precision floating point according to an embodiment of the present invention. It is a figure which concerns on the Example of this invention and shows the data arrangement | positioning of RAM and ROM. It is a figure which concerns on the Example of this invention and shows the example of a specification of fixed point and floating point data. 6 is a flowchart for performing type conversion processing of fixed-point data to floating-point data according to an embodiment of the present invention. 6 is a flowchart according to another embodiment of the present invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Engine ECU, 20 ... Engine, 30 ... Microcomputer, 31 ... CPU, 32 ... FPU, 33 ... Random access memory area (RAM), 34 ... Read-only memory area (ROM), 35 ... Peripheral module 35.

Claims (4)

Fixed-point arithmetic function for calculating fixed-point type data, floating-point arithmetic function for calculating floating-point type data, read-only memory (ROM) area for storing the fixed-point type data, and storing the floating-point type data In an electronic control device including a random access memory (RAM) area and including a microcomputer that performs various controls according to a predetermined control program,
The floating point type data stored in the random access memory (RAM) area is the same data of different types obtained by converting the fixed point type data stored in the read only memory (ROM) area into a floating point type. Electronic control device characterized in that The electronic control device according to claim 1.
The electronic control device according to claim 1, wherein the conversion is performed using a conversion program. The electronic control device according to claim 1 or 2,
The electronic control device according to claim 1, wherein the conversion and the storage of the converted floating point type data in the random access memory (RAM) area are performed during an initialization process of the microcomputer. The electronic control device according to claim 1 or 2,
An electronic control unit characterized in that the conversion and storage of the converted floating-point data in the random access memory (RAM) area is performed at an arbitrary time in a microcomputer.

Images