A New Filtering-Based Actuator Fault Diagnosis Approach Based on NN Models of PDFs

In many practical processes, the measured information is the stochastic distribution of the system output rather than its value. In this paper, following the new development for the fault diagnosis (FD) problem of stochastic processes (L. Guo and H. Wang, 2005), an improved FD method with H_infin performance optimization is considered by using the output stochastic distributions. A multi-layer perceptron (MLP) neural network is adopted to approximate the probability density function (PDF) of the system outputs. The measure of estimation errors represented by the distances between two output PDFs, would be optimized to find the diagnosis filter gain. Simulation example is given for the weighting dynamics to demonstrate the effectiveness of the proposed method