Title :
Influence and Model of SNR and Prediction Time for the Credibility of Fault Prediction
Author :
Shen Jingjing ; Song Hua
Author_Institution :
Sch. of Autom. Sci. & Electr. Eng., BeiHang Univ., Beijing, China
Abstract :
One of the difficult and key problems which must be solved for fault prediction is to measure the credibility of fault prediction results. This paper uses Monte Carlo method to build SNR and prediction time model for the credibility of fault prediction. First, using satellite attitude control system as an example, fault prediction is achieved successfully. Then the curve between SNR, prediction time and credibility estimation indicators is got by Monte Carlo method. By the method of curve fitting, model of credibility estimation indicators is built. The results show that in a certain range of accuracy, the model is appropriate. The technique provides one way to measure the credibility of fault prediction.
Keywords :
Monte Carlo methods; artificial satellites; attitude control; estimation theory; fault diagnosis; Monte Carlo method; SNR; credibility estimation; fault prediction; satellite attitude control system; Accuracy; Attitude control; Error analysis; Fitting; Predictive models; Satellites; Signal to noise ratio; SNR; credibility; fault prediction; prediction time;
Conference_Titel :
Instrumentation, Measurement, Computer, Communication and Control (IMCCC), 2013 Third International Conference on
Conference_Location :
Shenyang
DOI :
10.1109/IMCCC.2013.140