JPH10143495A - Method and device for operating average value - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten processing time without lowering calculation accuracy. SOLUTION: Each time a sample value is inputted (S103), a microprocessor (MPU) sums it up (S105) and when a number (i) of times reaches a number (n) (S107), the next input is waited. When the next input is performed (S109), the oldest sample value is subtracted from the summedup value (S110), the latest sample value is added in place of it (S111), the added total value is divided by (n) (S114), and an average value is outputted. Thus, it is enough only to perform addition, subtraction and division once by once for operating the average value from the (n+1)th time.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for calculating an average value and an arithmetic device therefor. For calculating an average value of angular velocity data of an angular velocity sensor for calculating the average value.

[0002]

2. Description of the Related Art An example of an average value calculating method for calculating the latest n (n is a positive integer) average values of sequentially input numerical values is disclosed in Japanese Patent Application Laid-Open No. 59-221969.

According to the technique disclosed in this publication, an average value is obtained by subtracting an average value from a total value of n numerical values, adding a newly input numerical value, and dividing this by n. That is.

In the prior art disclosed in this publication, a total value of the latest n numerical values sequentially inputted is calculated, the oldest numerical value is subtracted from the total value, and the total value after the subtraction is added to the total value. There is disclosed a technique of performing a process of adding an input numerical value, dividing the sum after the addition by n to obtain an average value each time a numerical value is input.

[0005]

However, according to the technique disclosed in that publication, instead of subtracting the oldest number from the sum of the n numbers, the average value at the present time is subtracted. There is a drawback that a value obtained by dividing the difference between the oldest numerical value and the current average value by n is an error in the average value.

Further, according to the technique disclosed in the prior art of that publication, the total value of n numbers is calculated each time a new value is input, so that the time required for addition becomes longer. There is a drawback that an arithmetic processing unit having high arithmetic processing capability is required.

SUMMARY OF THE INVENTION An object of the present invention is to provide an average value calculation method and a calculation device capable of reducing the processing time without lowering the calculation accuracy.

[0008]

According to the present invention, there is provided an average value calculating method for calculating the latest n (n is a positive integer) average value of sequentially input numerical values. ,
A first step of summing the first n numbers, a second step of subtracting the oldest number from the sum obtained in the first step, and a sum obtained in the second step A third step of adding the next input numerical value to the value, a fourth step of dividing the total value obtained in the third step by n to calculate an average value, and the second to fourth steps. And a fifth step of repeating the step.

Another aspect of the present invention is an average value calculating device for calculating the latest n (n is a positive integer) average value of sequentially input numerical values, wherein the n input numerical values are input first. And summing means for summing the n numbers stored in the storage means, and subtraction means for subtracting the oldest number from the sum of the n numbers obtained by the summing means. An adding means for adding the next input numerical value to the total value after being subtracted by the subtracting means; an average value calculating means for dividing the total value obtained by the adding means by n to calculate an average value; And control means for performing processing each time the numerical value is input to the subtraction means, the addition means, and the average value calculation means.

According to the present invention, after the total value of the first n numerical values is calculated, the oldest numerical value is subtracted from the total value, the latest numerical value is added, and Are divided respectively.

According to another aspect of the invention, the total value of the first n numerical values is stored in the storage means, and after the total value is calculated by the totaling means, the total value is calculated every time a numerical value is input. , The addition of the latest numerical value is performed by the adding means, and the division of the total value after the addition is performed by the average value calculating means.

[0012]

Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a configuration diagram of a preferred embodiment of a satellite tracking device including an average value calculation device according to the present invention.

The satellite tracking device detects an angular velocity of a rocking angle of a moving body such as a vehicle and outputs an analog quantity proportional to the angular velocity, and converts the analog quantity obtained by the angular velocity sensor 1 into a digital quantity. An A / D converter 2;
A microprocessor (hereinafter, referred to as MPU) 3 for performing an average value calculation based on the digital amount input from the A / D converter 2, and a RAM (Random Access Memory) for storing data used for the average value calculation.
y) 4 and R in which the processing program for the average value calculation is stored
An OM (Read Only Memory) 5, a backup memory 6 for storing parameters when power is turned off (off), a motor drive circuit 7 for driving a motor 8 based on a command value from the MPU 3, and a satellite communication It comprises a tracking antenna 9, a motor 8 for controlling the directivity of the antenna 9, and a satellite communication transceiver 10 for performing satellite communication via the antenna 9 and communicating information such as voice. Further, the satellite communication transceiver 10 notifies the MPU 3 of the satellite reception level, reception status, and the like necessary for improving the satellite pointing accuracy.

Among them, A / D converter 2, MPU3, R
The AM 4, the ROM 5, and the backup memory 6 constitute an average value calculation device 11.

Next, a description will be given of an area required for calculating an average value, which is secured in the RAM 4.

The area secured in the RAM 4 is a latest sample storage (NEW) in which the latest sample data is stored.
21, a sample value storage array unit 22 in which sample values used for averaging are sequentially stored, and a sample value storage array unit 2
Storage position indicating unit (i) 23 indicating the latest storage position
And a total value storage unit (TOTAL) 24 in which the total added value of the samples stored in the sample value storage array unit 22 is stored.
And an average calculation result storage (AVE) 25 for storing the average value of the samples.

The array of the sample value storage array section 22 is D
It is represented by T [n]. N is a positive integer and represents the number of elements of the averaging process.

Next, the operation of the average value calculating device 11 will be described with reference to a flowchart. FIG. 2 is a flowchart showing the first operation of the present invention.

Note that, although common to all of the first to fourth operations described below, the average value calculation is performed by the MPU 3 using the RO.
The processing is performed based on the processing program stored in M5, and data during the processing and the processing result are stored in a predetermined area in the RAM 4.

Referring to FIG. 1, in a first operation, first, when the average value calculation is started by turning on the power (ON) (S101), each value of the RAM 4 is initialized (S102).
That is, the storage position indicating unit (i) 23 and the total value storage unit (TO)
TAL) 24 and sample value storage array section (DT [n])
22 is reset.

Next, the latest sample value is stored in the angular velocity sensor 1
Is input through the A / D converter 2 and the MPU 3 and stored in the latest sample storage (NEW) 21 (S10).
3), and a sample value storage array section (DT [n]) 22
(S104).

Next, the sample value storage array section (DT
[N]) The sum of the sample values stored in 22 is calculated (S105). Now, since there is only one sample, the total value is equal to the sample value. The total value is stored in a total value storage (TOTAL) 24.

Next, the indicated value i of the storage position indicating unit (i) 23 is updated from 0 to 1 (S106). Then, it is checked whether or not the indicated value i matches the number n of elements of the sample value storage array section (DT [n]) 22 (S107). If not, the process returns to S103 and input of the sample value is continued. .

On the other hand, if the indicated value i matches the number n of elements in S107, it means that the elements necessary for the averaging operation of the present invention have been completed, and the storage position indicating unit (i) 23 is first reset. (S108), the latest sample value is stored in the latest sample storage (NEW) 21 (S109).

The total value storage (TOTAL) 24
The oldest sample value is subtracted from the total value in the above, and the total value after subtraction is stored in the total value storage (TOTAL) 24 (S110). The oldest sample value is the oldest sample value stored in the sample value storage array unit (DT [n]) 22, and is the position indicated by the indicated value i of the storage position indicating unit (i) 23. Is the sample value at.

Next, the latest sample value is added to the total value in the total value storage unit (TOTAL) 24, and the total value after the addition is stored in the total value storage unit (TOTAL) 24 (S111). Then, the latest sample value is overwritten and saved at the position where the oldest sample value was stored in the sample value storage array unit (DT [n]) 22 (S112). Indicated value i is 0 to 1
Is updated to (S113).

Next, the total value stored in the total value storage (TOTAL) 24 is divided by the number n of elements to calculate an average value (S114). Then, the average value is transferred to a necessary processing system such as the MPU 3 (S115).

Next, it is checked whether or not the indicated value i matches the number n of elements of the sample value storage array section (DT [n]) 22 (S116). If not, the flow returns to S109.
The average value is calculated while sequentially updating the designated value i.

On the other hand, if the indicated value i matches the number n of elements in S116, the process returns to S108, resets the indicated value i, and then continues the average value calculation processing in S109 and subsequent steps. Thus, the first operation ends.

Now, consider the case where the number n of elements of the sample value storage array section (DT [n]) 22 in the first operation is 16.

Since n = 16, the array DT of the sample value storage array unit (DT [n]) 22 is DT [0] to DT [0].
In the conventional method in which all elements are simply added each time, there are 16 elements up to [15], and 16 pieces of array data are added and divided, so that the calculation is performed with 15 additions and 1 division. In the present invention, however, in the present invention, the operations required each time to obtain the total value include one subtraction of the oldest data, one addition of the latest data, and division for calculating the average value. Only once, the result is obtained in a processing time of 2/15 of the conventional example in terms of the addition and subtraction time.

Next, in the case where the initial value is continuously increased by 10 per sample from the initial value 100, a comparison is made with the conventional case where the current average value is subtracted from the total addition stored value and new data is added.

Assuming that 100 when n = 1, 110 when n = 2,..., And 250 when n = 16, are stored in the sample value storage array section (DT [n]) 22, respectively. The average value is 100 + 110 + 120 + ... + 25
Since 0 = 2800 and 2800/16 = 175, the average value is 175.

Assuming that the next sample value 260 is input, in the case of the present invention, 2800-100 + 260
= 2960, and 2960/16 = 185.

On the other hand, in the case of the conventional example, 2800-175
+ 260 = 2885, and 2885/16 = 18
0.3.

That is, in the case of the present invention, the average value data is also 1
85-175 = 10, which is surely increased by 10, but in the case of the conventional example, it increases only by 180.3-175 = 5.3, which indicates that the error is large.

Next, the second operation will be described. FIG.
Is a flowchart showing a second operation of the present invention.

The second operation differs from the first operation in that S201 and S202 are added between S106 and S107 of the first operation (see FIG. 2), and the other operations are the first operation. Is the same as Therefore, the operation is described in S201.
Steps S202 and S202 are performed, and the other steps are not shown in the drawings and the description is omitted.

In the second operation, every time one sample value is input, the average value (AVE) is obtained in S201, and
At 202, the average value is transferred to a necessary processing system such as the MPU3.

That is, in the first operation, in FIG.
Since the average value is not output until the process from 101 to S107 is completed, a predetermined time is required from the power-on to the output of the average value. However, in the second operation, the process from S101 to S107 is performed.
Since the average value is output in S202 even in the period up to this point, such a disadvantage is solved.

More specifically, in the first operation, the average value cannot be output until the number of elements n required for the averaging operation is accumulated, but in the second operation, the total value (TOTA
L) and the indicated value i, the provisional average value is calculated as (total value /
By calculating in i), the average value can be output each time the first sample value is input.

Next, the third operation will be described. FIG.
Is a flowchart showing a third operation of the present invention.

The third operation differs from the first operation in that S102 to S107 of the first operation (see FIG. 2) are replaced with S301 and S302. Others are the same as the first operation. Therefore, the operation is described in S301.
And S302, and the other steps are not shown in the figure and the description is omitted.

The third operation is as follows: after the power is turned on (S101),
When the first sample value is input (S301), the sample value is multiplied by n, the sample value multiplied by n is stored in the total value storage unit 24, and the n number of samples stored in the sample value storage array unit 22 are stored. The first sample value is stored in the section (S302). That is, the same sample values are stored in the n storage units of the sample value storage array unit 22.

Then, based on the total value stored in the total value storage section 24 and the sample values stored in the n storage sections of the sample value storage array section 22, the above-described processing of S108 to S116 is performed.

That is, in the third operation, it is possible to shift to the processing from S108 to S116 when the first sample value is input, so that the time required from power-on to output of the average value is reduced. It can be shortened.

In this case, the average calculation result based on the normal number of samples is obtained when n samples are obtained after the apparatus is started.

Finally, the fourth operation will be described. Referring to FIG. 1, the fourth operation is to store R in backup memory 6.
The MPU 3 controls so that the final result of the average value stored in the average value calculation result storage unit 25 in the AM 4 is stored. In this way, the final result of the average value is stored in the backup memory 6 even if the power of the apparatus is cut off. Can be read and used.

Therefore, by using the sample value read from the backup memory 6 as the first sample value (S301) of the third operation (see FIG. 4), the input of the first sample value can be performed without waiting for the input of the first sample value. It is possible to enter the standby for the processing of S108 and thereafter. That is, when the first sample value is input, the process is started from S109, so that the time required from the input of the first sample value to the output of the first average value is set as the first operation. It can be shortened.

Here, the backup memory 6 stores the EEP
ROM (Electrically Erasable)
Any type of readable and writable memory such as a programmable ROM (RAM) or a battery-backed-up RAM that can store stored data when the power is turned off can be used.

Although the MPU 3, RAM 4, and ROM 5 in the average value calculation device 11 (see FIG. 1) are described separately for convenience of explanation, a one-chip IC (Integrate) is used.
dCircuit) CPU (Central Proc)
An IC having the functions of the MPU 3, the RAM 4, and the ROM 5 may be used in one IC as in the case of an Essing Unit.

Further, since the angular velocity sensor 1 has an analog output, the digital conversion is performed by the A / D converter 2. However, a digital output sensor is used as the angular velocity sensor 1, and the angular velocity sensor 1 is connected by, for example, serial communication. MPU output
3 may be input.

Further, the average value is stored in the average calculation result storage unit 25. However, the average value is not stored in the average calculation result storage unit 25, and the processing system (MPU 3
Etc.). By doing so, the average calculation result storage unit 25 can be omitted.

On the other hand, although the total value is configured to be stored in the total value storage unit 24, if the calculation accuracy is not so required and the output result can be obtained only by the integer operation, the average value is used instead of the total value. If only the area is secured and the total value is obtained by multiplying the average value by n, the total value storage unit 24 can be omitted. Further, since the average value is 1 / n of the total value, the memory area to be used can be reduced. Further, if n is a power of 2, the total value can be obtained by shifting the corresponding data instead of multiplication, so that the processing time does not become a large burden.

[0055]

According to the present invention, there is provided an average value calculating method for calculating the latest n (n is a positive integer) average value of sequentially input numerical values, wherein the first n numerical values are summed. A first step of subtracting the oldest numerical value from the total value obtained in the first step, and a numerical value input next to the total value obtained in the second step. A third step of adding
An average value calculation method is composed of a fourth step of dividing the total value obtained in the third step by n to calculate an average value and a fifth step of repeating the second to fourth steps. Therefore, it is possible to reduce the processing time without lowering the calculation accuracy.

According to another aspect of the present invention, there is provided an average value calculating device for calculating the latest n (n is a positive integer) average value of sequentially inputted numerical values, wherein , A summing means for summing n numbers stored in the storage means, and n obtained by the summing means.
Subtracting means for subtracting the oldest number from the sum of the numbers, adding means for adding the next input value to the sum after the subtraction by the subtracting means, and adding means obtained by the adding means. An average value calculating device including an average value calculating unit that divides a total value by n to calculate an average value, and a control unit that performs processing each time the numerical value is input to the subtracting unit, the adding unit, and the average value calculating unit. With this configuration, it is possible to reduce the processing time without lowering the calculation accuracy.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a best mode of a satellite tracking device including an average value calculation device according to the present invention.

FIG. 2 is a flowchart showing a first operation of the average value calculation device according to the present invention.

FIG. 3 is a flowchart showing a second operation of the average value calculation device according to the present invention.

FIG. 4 is a flowchart showing a third operation of the average value calculating device according to the present invention.

[Explanation of symbols]

 Reference Signs List 3 microprocessor 4 RAM 6 backup memory 21 latest sample storage unit (NEW) 22 sample value storage array unit 23 storage position indicating unit 24 total value storage unit 25 average calculation result storage unit

Claims (9)

[Claims] 1. An average value calculation method for calculating the latest n (n is a positive integer) average value of sequentially input numerical values, comprising: a first step of summing the first n numerical values; A second step of subtracting the oldest value from the total value obtained in the first step, and a third step of adding the next input value to the total value obtained in the second step. And a fourth step of dividing the sum obtained in the third step by n to calculate an average value, and a fifth step of repeating the second to fourth steps. Average value calculation method. 2. The method according to claim 1, wherein the first step is a step of summing numerical values each time a numerical value is input, and the step of summing the total value when the number of input numerical values reaches n. 7. A method according to claim 1, further comprising the step of: 3. The first step includes a sixth step of summing numerical values each time a numerical value is input, and a number i (i is a positive number) of inputting the total value obtained in the sixth step. An eighth step of dividing by i ≦ n) and calculating an average value;
a ninth step in which the sixth and seventh steps are repeated until i becomes equal to n; and a tenth step in which the sum of the n numbers is passed to the second step. The method for calculating an average value according to claim 1, wherein 4. The first step comprises: an eleventh step of multiplying a first input numerical value by n; and considering the n-multiplied numerical value as a total value of n first input numerical values. The second
2. A method according to claim 1, further comprising a twelfth step of passing the step to the step. 5. An average value calculating device for calculating the latest n (n is a positive integer) average value of sequentially input numerical values, wherein the storing means stores the first input n numerical values. Summing means for summing the n numbers stored in the storage means; subtraction means for subtracting the oldest number from the sum of the n numbers obtained by the summing means; Adding means for adding a numerical value to the total value after being subtracted by the subtracting means, average value calculating means for dividing the total value obtained by the adding means by n and calculating an average value; Control means for performing processing each time the numerical value is input to the addition means and the average value calculation means. 6. The second summing means for summing numerical values each time a numerical value is input, and outputting the total value to the subtracting means when the number of input numerical values reaches n. The average value calculation device according to claim 5, further comprising an output unit. 7. The summing means includes: a second summing means for summing a numerical value each time a numerical value is input; and a number i (i is a positive number) of inputting the total value obtained by the second summing means. Where i ≦ n), the second average value calculating means for calculating the average value, and the summing means and the second average value calculating means perform processing until i becomes equal to n. 6. The average value calculation device according to claim 5, further comprising: output means for outputting the total value at that time when i becomes equal to n to said subtraction means. 8. The multiplying means for multiplying a numerical value inputted first by n, and a numerical value integrated by the multiplying means as the total value of the n numerical values input first. 6. The average value calculation device according to claim 5, further comprising an output unit that outputs the average value to the subtraction unit. 9. A backup memory for storing a final result of the average value calculated by the average value calculating means, wherein the totaling means calculates a final result of the average value stored in the backup memory for a next average value calculation. 9. The average value calculation device according to claim 8, wherein the average value calculation device is used as a numerical value input first when processing.