A Maximum Entropy Based Approach to Fault Diagnosis Using Discrete and Continuous Features

This chapter presents a new maximum entropy (ME) based hybrid inference engine to improve the accuracy of diagnostic decisions using mixed continuous-discrete variables. By fusing the complementary fault information provided by discrete and continuous fault features, false alarms due to misclassification and modeling uncertainty can be significantly reduced. Simulation results using a three-tank benchmark system have clearly illustrated the advantages of diagnostics based on mixed continuous-discrete variables. Moreover, in the presence of significant measurement noise, simulation results show that the proposed ME method achieves better performance than the support vector machine classifier. A generalized structure of model-based FDI schemes consists of two stages of residual generation and residual evaluation. In the first stage, measurements of system variables and parameters are compared with the estimated values provided by a dynamic model to generate a residual vector. Then each residual is compared with a pre-designed threshold in the second stage of residual evaluation, and an FDI decision is made based on the binary residual responses. © 2007