SU1053113A1 - Gradual fault predictor - Google Patents

Abstract

PREDICTOR FAILURE OF FAILURES, containing the starting block, bluetooth / normalization and coefficient block, serially connected pulse counter, the first multiplication block, the first subtraction block and the display block, so that, in order to expand the area its application by providing the ability to predict quasideterministic processes, changing 101 lihs according to an arbitrary a priori unknown law, it contains a memory block of the mathematical expectation of the monitored parameter, the memory block is corrected function of the parameter being monitored, pulse generator, second subtraction unit, second multiplication unit, comparator, adder, trigger, AND element, the first input of which is connected to the pulse generator, and the output is connected to the input of the estimator and the second stroke with the predictor input, the memory block of the mathematical expectation of the monitored parameter is connected to the output of the And element and to the input of the memory block of the correlation function of the controlled parameter whose output is connected to the first input the second multiplier unit, the first input of the comparator is connected to the output of the normalization unit, the first input of the adder is connected to the outputs of the second multiplication unit, C the second input - from the output of the memory unit of the expected parameter of the monitored parameter, and the output - to the BTopi input of the comparator, the trigger input is connected to the comparator output, the zero input is connected to the output of the trigger unit, and the output is connected to the second input of the element I, the first input of the second block is ii; The subtraction is connected to the output of the memory expectation block of the controlled parameter pa, the second input -i ate with the output of the unit for registering and the output with the input of the coefficient block, oo the first output of which is connected to the second input of the first subtraction unit, the second output with the second input of the multiplication unit, and the third output g with W eye the input of the first multiplication unit.

Description

The invention relates to automation and can be used in the study of the reliability of automatic control systems and electronic equipment, in particular, to predict the time of the onset of gradual failures. A device for prediction is known that contains blocks of monitoring, control, and calculation of velocity parameters. , RAM, Forward and Inverse Functional Converters, Difference Calculations and Logic Elements f. A disadvantage of this device is the use of the exponential function to predict the change of the monitored parameter at future times, which limits the use of the device. . . The closest in technical essence to the invention is an exponent for a gradual failure predictor comprising voltage-to-code conversion blocks, controls, displays, a switch, a reversible counter, a subtraction register, a decoder, a prediction counter, and gates 2. A disadvantage of this device is the use of a prediction algorithm for extra polishing using a linear binomine in it, which significantly limits the scope of use of the device and makes it possible to use it to predict only linearly varying processes. The purpose of the invention is to expand the scope of the predictor by providing the possibility of predicting the implementation of quasi-deterministic processes that vary according to an arbitrary a priori unknown law. . The goal is achieved. the fact that the predictor containing the triggering unit, the normalization unit and the coefficient block, the pulse counter connected in series, the first multiplication unit, the first subtractor and the display unit, contain the memory block of the expected value of the monitored parameter, the correlation function memory of the monitored parameter pulse generator, second subtraction unit, second multiplication unit, comparator, adder, trigger, AND element, the first input of which is connected to the pulse generator, and the output is connected to the pulse counter input with the input of the normalization unit, the second input of which is connected to the input of the predictor, the input of the memory block of the mathematical parameter of the monitored parameter is connected to the output of the element And and to the input of the memory block of the correlation function of the monitored parameter whose output is connected to the first input of the second multiplication unit, the second input the comparator is connected to the output of the rationing unit / the first input of the adder is connected to the output of the second multiplication unit, the second input is connected to the output of the memory unit of the expected value of the controlled pair tra and an output - to a second input of the comparator unit ,, trigger input connected to the output of the comparator, the zero input - with the output. the start block, and the output — with the second input of the element I, the first input of the second subtraction unit is connected to the output of the memory unit of the mathematical expectation of the monitored parameter, the second input is connected to the output of the normalization unit, and the output is connected to the input of the coefficient block, the first output of which is connected with the second input of the first subtraction unit, the second output with the second input of the multiplication unit, and the third output with the second input of the first multiplication unit. The drawing shows the block diagram of the predictor gradual failures. The device contains a start block 1, trigger 2, pulse generator 3, element 4, block 5 of the mathematical expectation memory of the parameter being monitored, block 6 of the correlation function of the parameter being monitored, the second multiplication unit 7, adder 8, comparator 9, block 10, Factor step 6c, second subtraction unit 11, normalization unit 12, pulse counter 13, first multiplication unit 14, first subtraction unit 15, display unit 16. The predictor implements the algorithm for predicting the deviation of the monitored parameter out of acceptable limits using the formula Y "(i))) (YO -, g (G). Where YjJCS) is the predicted value of the V controlled parameter / ra; |) (t) - correlation function. predictable parameter; Y- is the value of the controlled parameter at 0; T. is the value of the mathematical expectation of the controlled parameter at O ;, r4 (t) is the current knowledge of the mathematical expectation of the controlled parameter. The solution of this equation allows one to determine the correlation interval i of the correlation function (Cf). The decay time of the correlation function determines the interval during which the i-linear correlation between the values of r eCiCa is practically attenuated. At the same time p (o) l O and Y (to)). If (Co) o6ecne4HTb-m, j Ci :) Yg (where Yg is the allowable value of the parameter being monitored), then the time until the equipment fails can be defined by the expression Ир - ог The irrognostic machine works as follows. When a signal is received from the start-up unit 1, a single input S of the trigger 2 causes the voltage from this trigger to be sent to the first input of an element 4, to the second input of the element i, clock pulses with a periodicity T {, from the generator 3, impulse from the output of the element i and when i) O arrives at the inputs of the memory block, mathematical expectation 5 and the correlation function 6 of the parameter being monitored, as well as to the inputs of pulse counter 1 and the normalization device 12. The values of the parameter Y and its mathematical expectation then go to the second block 11 .count . The value of L YO - Sd from the second block you read comes in block 10 coefficients. This also records the values of the repetition period of the clock pulses T of the generator 3 and the time of the correlation function of the parameter being run. The value of Y is the SD nocTyrtaet from the block of 10 coefficients to the second multiplication unit 7, where it is multiplied by the value of the correction function p, j (0) 1, coming from the memory block of the correlation function of the correlated parameter b. The resulting value of YO - K, Q is summed in adder 8 with the value of mg supplied from memory expectation unit 5 of the monitored parameter, and compared in comparator 9 with the value of YO coming from block 12 of the normalization. Thus, when O1, equation (1) degenerates into Yg YO. At the same time in the pulse counter. 13 (counting the number of intervals Tf,) will record the value of N 0. Therefore, the display unit will record the time / l m until the equipment LO has failed. From the comparator 9, the voltage is applied to the zero input R of the flip-flop 2 and brings it back to the initial state. This completes the preparatory preprocessing operation (first measurement point, 0). Upon further receipt of the signal from start unit 1 (second measurement point, 0) to the single input S of the trigger 2, the voltage from this trigger enters the element And 4 At the second input element And receive clock W pulses with a frequency of TP from the generator 3 pulses. Through the element And these pulses arrive at the inputs of the memory 5 block of the mathematical expectation of the controllable parameter. the memory block of the correlation function of the monitored parameter, the rating block 12 and the counter 13 pulses. Upon receipt of clock pulses, the current values of mu (t), the correlation function rh (1} and the normalized value Y () of the predicted parameter) are received from the outputs of these blocks. The value of C (C) is multiplied using the second block 7 multiplied by the value and YO - Schd / coming from the block of 10 coefficients. Products are summed with the help of adder 8 with the values of m (Z} i coming from block 5. The obtained values of p ,, (e) (Yo - Sho) + t, (0) are compared by a comparator 9 with the value of the normalized parameter Y (t), the incoming and From the normalization block 12. In this case, the clock pulses coming from element 4 are counted by pulse counter 13. The number of counted pulses is multiplied by the first block 14. multiplied by the repetition period of clock and multiples Tf, coming from, block 10 coefficients. NTf, is subtracted from the decay time of the correlation function with the help of the first subtraction unit 15. At the time when the voltages are equal at the inputs of the comparator 9, it is triggered, a pulse is generated at its output to the zero input R of trigger 2. Tr gger returns to the initial state. The signal at its output disappears, which leads to blocking of element 4. The receipt of pulses from the generator of 3 clock pulses to the counter of 13 pulses stops. : 1In this way, the modeling of equation (1) is carried out, the solution of which is proportional to the number of pulses received on the counter 13 pulses. Choosing accordingly the pulse repetition frequency of the generator 3, we obtain the code of the number of pulses recorded in the counter 13, proportional to the time of the controlled parameter deviation to the allowable one; limit. The calculated predicted value fg-jn to-i (where c is the root of equation (1)) is displayed on the display unit of the display 16. The advantage of the proposed device is the possibility of its use for predicting quasideterministic processes, changing according to an arbitrary a priori unknown law, according to two measured values of the monitored parameter.

   fff (f

w

/

13

eight

U-.

//

fff

//

Claims (1)

GRADIATE FAILURE PREDICTOR, comprising a start block, a normalization block and a coefficient block,; connected in series with a pulse counter, a first multiplication unit, a first subtraction unit and a display unit, with the aim that, in order to expand the scope of its application by providing the possibility of predicting quasideterministic processes that vary randomly a priori unknown law, it contains a mathematical expectation block of a controlled parameter, a correlation function memory block of a controlled parameter, a pulse generator, a second subtraction block, a second multiplication block, com Arator, adder, trigger, element,, the first input of which is connected to the pulse generator, and the output - to the input of the pulse meter and the input of the normalization unit, the second stroke of which is connected to the input of the predictor, the input of the mathematical expectation block of the controlled parameter is connected to the output - element And and with the input of the memory block of the correlation function of the controlled parameter, the output of which is connected to the first input of the second multiplication block, the first input of the comparator is connected to the output of the normalized block. Niya, the first input of the adder is connected to the output "! the second block of multiplication, the second input - from the output! ” the memory block is the mathematical expectation of the parameter being monitored, and the output is with the second input of the comparator, a single trigger input is connected to the output of the comparator, the zero input is with the output of the start block, and the output is with the second input g of the And element, the first input of the second ka se subtracting> one with the expectation output of the memory unit konFroliruemogo parameter η second input to the output of the valuation unit and taken the analysis - with the input koeffitsiontov ^block! ; the first output of which is connected to the second input of the first subtraction unit, the second output to the second input of the multiplication unit, and the third output to the second input! ” first block of multiplication.