SU1674157A1 - Statistics analyzer - Google Patents

Abstract

The invention relates to computing and can be used to determine the sample size required to compute an estimate of the average value of a random signal with a given accuracy. The purpose of the invention is to increase speed. The statistical analyzer contains block 1 for calculating extreme values, block 2 for dividing, first functional converter 3 of the form Y = (X - 1 / X + 1) ² , pulse generator 4, second functional converter 5 of the form Y = F ² / T ² X (F = 10 @ ^{/ 2} ° - 1, Α-confidence interval, T - Student coefficient), comparison unit 6 and counter 7. When estimating the sample size, the primary process variation indicator is used as the ratio of the extreme values of the input sequence counts. 1 il.

Description

one

(21) 4663041/24 (22) 12.20.88 (46) 08.30.91. Bul №32 (72) V.N. Streams (53) 681.3 (088.8)

(56) USSR Author's Certificate No. 1003096, cl. G 06 F 15/36, 1981.

(54) STATISTICAL ANALYZER

(57) The invention relates to computing and can be used to determine the sample size required by i to compute an estimate of the average value of a random signal with a given accuracy. The purpose of the invention is to increase speed. The statistical analyzer contains a block 1 for calculating extreme values, a block 2 for dividing, a first functional converter 3 of the form Y (X-1 / X + 1) 2, a pulse generator 4, a second functional converter 5 of the form U F2 / t2 x (F 10/2 ° - 1, «-confidence interval, t - Student coefficient), unit 6 comparison and counter 7. When estimating the sample size, the primary indicator of the process variation is used as the ratio of the extreme sample values of the input sequence. 1 il.

w

e

WITH

h |

Ј joint venture

Vj

The invention relates to computing and can be used to determine the sample size required to compute an estimate of the average value of a random process with a given accuracy.

The purpose of the invention is to increase speed.

The drawing shows a structural diagram of a statistical analyzer.

The statistical analyzer contains block 1 for calculating extreme values, block 2 for dividing, first functional / X-1 v2 converter 3 of type V

pulse generator 4,

pTFTj is the second functional converter of the form Y – y X 5,

Comparison unit b, counter 7.

Statistical analyzer works as follows.

Start the next. operation occurs after the arrival of a triggering pulse, according to which a reset of counter 7 into O occurs and block 1 returns to its initial state. After that, a zero code is generated at the output of the maximum values of block 1, and a single code at the output of the minimum values. As a result, in accordance with the values of the dividend (Hmax) and the divisor (Xmin), a zero code is generated at the output of block 2, which determines the zero code at the first input of the compare block 6. The zero state of the output of the counter 7 determines the zero code at the output of block 5, which, in combination with the zero code arriving at the first input of the comparison block 6, determines the zero state of the output of the compare block 6. The zero stuck at the output of the comparison block 6 starts the pulse generator four.

After the initial start signal, sequential counts of the random process {Xn} begin to arrive at the information input of the device. Each count, followed by a clock pulse, is analyzed in block 1 in order to isolate the maximum H max and minimum X Min values. The count number is recorded in the counter 7, depending on the number of clock pulses arriving at its input.

If the condition Xn Hmin is fulfilled in the next reading, then the value Xmin is replaced by the value Xn again arriving at the input, a similar correction Hmaks also occurs when Xn Hmaks.

At the output of dividing unit 2, a value K appears, corresponding to the expression

v - Hmaks

Y

AMINE

The value of K, which enters the input of the functional converter 3, determines at its output the code / 9 in accordance with the expression

, (к-п2 р- (тггт /

The magnitude / characteristic of the variation of the countable process.

At the output of block 5, in accordance with the number of samples received at its input n, a code is formed, which is determined from the expression

y-

4ie where F 10 -1;

a is the specified confidence interval of the deviation of results, dB;

t is the student coefficient for a given confidence probability P, which is a function of n.

Thus, the quantity y specifies 5 requirements for the accuracy of estimating the average value of the process.

Codes / 3 and y, respectively, arrive at the inputs of block 6 of the comparison. If the condition ft y is fulfilled, an active signal is generated at the output of the 6 0 comparison unit, signaling that the accuracy criterion is met and, therefore, the number of incoming samples n is sufficient to achieve the specified reliability of the estimated 5 mean process value {Xn}. The active signal at the output of the comparison unit 6 blocks the operation of the pulse generator 4, as a result of which the output of the device permanently contains a code corresponding to 0 the number of samples n forming the sample.

If the accuracy criterion is not met, the evaluation procedure continues, and in the counter 7 the number of 5 samples continues to accumulate.

After the criterion of accuracy is fulfilled, the device states are blocked until the next Start signal arrives, bringing the device to the initial state.

Claims (1)

Claims of the invention Statistical analyzer containing a dividing unit, a counter and a comparison unit, characterized in that in order to improve speed, a unit for calculating extreme values, a pulse generator, the first functional / X-1 2 type transducer A + / 5 16741576 the second functional converter with the first input of the comparison unit, p2 afa output of which is connected to the input of the variable Y-n (F 10 -1, and a truster-start of the pulse generator, the output of which is connected to the clock input of the unit t interval, Student's t-factor, g calculation of extrinsic values and n is the number of samples), with the information-counting input of the counter. the output of which The input of the extremal functional conversion unit is informational input / 1S POD | TERMINAL to the second input of the unit house of the analyzer, outputs of maximum and comparison, input of the start of the analyzer of the minimum value of the calculator block 10 with the installation input of the block comp. of extreme values are connected to the E1 extrema ratio and with the input responsible for the inputs of the divisible and divisor sb counter1 whose output is dividing unit, the output of which is through the ln-output of the analyzer Choice number 0, the new functional converter is co