Study Of Fault Detection And Diagnosis Of Radar-telemeter

Observing the working of equipment, Technology of fault detection and diagnosis (F.D.D.) can ensure whether equipment working normally or in error, detect the fault early, find out the cause of the fault, can predict the trend of the fault. F.D.D. consists of state inspecting, fault detecting, identifying and diagnosing, analyzing and forecasting and advice on dealing with fault. For 30 years, F.D.D. synthesizes technologies of various fields and takes out developments and advancements in theory and application, Involves theories of System, Cybernetics, Informatics, theory of Detecting and Estimate, Computer. F.D.D. is made up of fields of the cause of fault, technology of dealing with the info of fault, separating and ascertaining the position of fault, study of the application of Artificial intelligence. Plotting-cell of aim distance (P.C.D.) of radar is made up of Amplifier of video frequency (A.V.F.), Discriminator, Corrector, Amplifier of Error-voltage (A.V.E.), integrator of error-voltage, adjustable delayer of pulse (A.D.P.), Generators of pulses of distance. All parts of the cell are made up of a closed loop with a subtractive feedback. Signals of video frequency of aim (S.V.F.) is fed to cell circuit, when the plot of the S.V.F. is equality to the plot of the plot of Signals of double doors (S.D.D.), but is unequal in centers of the signals, Error voltage is produced by the Discriminator, its value and sign rest with the value and sign of difference of the S.V.F. and the S.D.D. After amplified by A.V.E., the integrator produces the Control-voltage by Error-voltage, according to the Control-voltage, the A.D.P. produces quadrate pulse which extent is controlled by the Control-voltage, the falling edge of the quadrate pulse activates generators of pulses of distance, so, all pulses are produced (plotting-pulse and pulse of double doors etc.). By simplifying becomingly, we can find out mathematical models of part-circuits of the cell and models of Automation. According to models, we can found the simulant models by software Simulink of Matalb. Simulating the normality and familiar fault, we can get the responses of Error-voltage and Control-voltage: response00, response0 1, ……, response16. To avoid the pollution by disturbance signals, we gather data at low frequency in order to avoid the influence of S.V.F. with high frequency. But we find that there are disturbance signals of low frequency at the outcome wave. Regard as outliers, data can be smoothed by multinomial slipping fitting (M.S.F.). Comparing the smoothed data and original data, we can find outliers and delete these exceptional points. Supply these data points by Linear Interpolation. After treatment, the worked signal with low degree distortion can reveal the real state of cell circuit. Aggregate of criterion faults signals is made up of familiar voltage waves of fault, comparing standard waves with waves gotten to be detected, we can get the data about the degrees between standard waves and waves to be detected. So, we can estimate whether the cell is in fault, which style fault is taking place. In experiment, four waves are well detected. In the final, the paper shows: detect the voltage waves of the cell circuit, we can know in which state the P.C.D. is working, can estimate whether the circuit is in fault, detect the cause and position of the fault. Extending these techniques, we will realize the F.D.D. of radar.