JPH01149174A - Arithmetic processing circuit - Google Patents

Abstract

PURPOSE:To operate a function value always having constant accuracy by interpolate-operating the function value at an arbitrary point based on the curvature of a prescribed function stored in a function information storing means. CONSTITUTION:A ROM 1 previously stores the function value of each interval, into which a prescribed interval is subdivided arbitrarily, and the inclination information of a function corresponding to this function value based on the curvature of the prescribed function. When the interpolative operation of the function value at the arbitrary point is designed by an input part 4, an operation part 3 reads the function value of each interval and the inclination information of the function corresponding to the function value from the ROM 1 and interpolate-operates the function value at the arbitrary point. The arithmetic result is sent through a CPU 2 to an output part 5 composed of a printing part 5a and a displaying part 5b.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an arithmetic processing circuit that computes a preset function value, and particularly relates to an arithmetic processing circuit that approximates a function value in an arbitrary interval. be.

[Conventional technology]

Conventionally, devices having this type of arithmetic processing circuit, such as desktop calculators, have been configured to execute preset function values while referring to arithmetic data stored in a storage medium such as a ROM. There is.

That is, when calculating a function value corresponding to an argument given by a computer, etc., the value is stored in a storage area such as a table in advance, and the required calculation data is read from the above 1Δ area each time calculation processing starts. , is configured to operate by interpolation.

[Problems to be Solved by the Invention] Conventional arithmetic processing circuits calculate function values according to the process described above. If it is a function, a predetermined calculation must be performed on the slope value as well, which increases the number of calculation processing steps, requires wasted calculation processing time, and significantly delays the start time of the next program execution. There were problems such as the

Additionally, when calculating based on table data in which function values are carved at regular intervals, the error in the function value fluctuates depending on the curvature at a specific point of the function, making it impossible to calculate a function value with constant accuracy. was there.

This invention was made to solve the above-mentioned problems, and in interpolation calculation of function values, by subdividing the desired interval based on the curvature of the function and calculating the function value, the function value is always calculated. The purpose is to obtain an arithmetic processing circuit that can calculate function values with a certain precision.

[Means for solving problems]

The arithmetic processing circuit according to the present invention includes a function information storage means that stores in advance a function value of each section in which a predetermined section is arbitrarily subdivided based on the curvature of the predetermined function and slope information of the function corresponding to this function value. , comprising arithmetic processing means for interpolating a function value at an arbitrary point based on each interval function value of the predetermined function stored in the function information storage means and slope information of the function corresponding to each interval function value. It is.

(Operation) In the present invention, when an interpolation operation is instructed for a function value at an arbitrary point, the function information storage means stores the function value of each interval in which the predetermined interval is arbitrarily subdivided based on the curvature of the predetermined function. An arithmetic processing means reads the slope information of the function corresponding to each interval function value, and interpolates the function value at an arbitrary point.

(Embodiment) FIG. 1 is a block diagram illustrating the configuration of an arithmetic processing circuit showing an embodiment of the present invention. Reference numeral 1 denotes a ROM which also serves as a function information storage means of the present invention, and a ROM containing a control program and a curvature of a predetermined function. Function value information for each section in which the predetermined section is arbitrarily subdivided based on is stored. 2 is a CPU, and when calculation is started and numerical data (arguments) are input from the human power section 4, each interval function value and the slope (differential coefficient) of the function corresponding to each interval function value are calculated according to the control program stored in the ROMI. Information) Among the information, necessary specific interpolation information is read out and output to the calculation unit 3. The calculation unit 3 constitutes the calculation processing means of the present invention, and executes an interpolation calculation according to the following equation (1) according to the interpolation information output from the CPU 2, and transfers the calculation result to the CPU 2. Reference numeral 5 denotes an output section, which includes, for example, a print section 5a, a display section 5b, and a calculation section 3.
It is possible to output calculated values, numerical values related to calculations, and I control information.

In the interval [a, b], f'(X
) = (f (b) - f (a) )/ (b-a)
<a<=x<b>...(1) There exists an X that satisfies the following. Therefore, f'(X) is (a.

The distance between the straight line connecting the two points f (a)) and (b, f (b)) and the function f (X) is the maximum. Therefore, the function value f (Xl) corresponding to the fine increments obtained by dividing the interval a ONa n-1 according to the curvature of the function f (X) so that this distance is constant is stored in ROM 1 as interpolation information. .

Next, with reference to FIG. 2, the interpolation calculation process of the calculation section 3 shown in FIG. 1 will be explained.

FIG. 2 is a flowchart illustrating an example of a processing procedure for interpolation calculation according to the present invention. In addition, (1) to (3)
indicates each step.

First, for the argument X instructed by the human resources department 4, al<X
Determine ai such that <a+*t (1), then,
Read f (at) and f (aHl) from ROM1 and set their contents in registers A and B of CPU2 (2).

Next, according to the contents of register A and register B, ((
f (at) - f (at, +)) / (ai - at
-+)x(x-al)+f(as)) and set it in register Y3), and the process ends.

Therefore, a function approximation calculation can be performed according to the curvature, and the error in the function value in an arbitrary interval converges to a certain range.

〔Effect of the invention〕

As explained above, the present invention provides durability information storage means for storing in advance the function value of each section in which a predetermined section is arbitrarily subdivided based on the curvature of the predetermined function and the slope information of the function corresponding to this function value. and function information 1. (! Since the calculation processing means is provided for interpolating the function value at an arbitrary point based on each interval function value of the predetermined function stored in the means and the slope information of the function corresponding to each interval function value, the function It is possible to perform function approximation calculations that are not affected by the curvature of Demonstrate.

[Brief explanation of the drawing]

FIG. 1 is a block diagram illustrating the configuration of an arithmetic processing circuit according to an embodiment of the present invention, and FIG. 2 is a flowchart illustrating an example of a processing procedure for performing an interpolation operation according to the present invention. In the figure, 1 is a ROM, 2 is a CPU, 3 is an arithmetic section, 4 is an input section, and 5 is an output section. Figure 1

Claims (1)

[Claims] In an arithmetic processing circuit that executes predetermined four arithmetic operations according to a pre-stored program, a predetermined interval is arbitrarily subdivided based on the curvature of the predetermined function, and the function value of each interval and the slope of the function corresponding to this function value are calculated. a function information storage means for storing information in advance; and a function value at an arbitrary point based on each interval function value of a predetermined function stored in the function information storage means and slope information of the function corresponding to each interval function value. 1. An arithmetic processing circuit comprising arithmetic processing means for performing interpolation calculations.