Fault Detection And Isolation Using Statistic Signal Processing

Abstract

In this work, statistical signal processing are employed in fault detection and insolation scheme. The procedure was applied to detect the failed sensor and suppose the use of mathematical model of process and the correlation (covariance) function calculated in real time between the signals, which represent the commands, and the measurements. If the values of these statistical functions are great, that means the signal is correlated. When these values are small and this situation is preserved long time that means we have a failed sensor. The technique was tested by simulation and the results are good.