Designing an Efficient Probabilistic Neural Network for Fault Diagnosis of Nonlinear Processes Operating at Multiple Operating Regions

Abstract

Neural networks have been used for process fault diagnosis. In this work, the cluster analysis is used to design a structurally optimized Probabilistic Neural Network. This network is called the Clustered-Based Design Probabilistic Neural Network (CBDPNN). The CBDPNN is capable of diagnosing the faults of nonlinear processes operating over several regions. The performance and training status of the proposed CBDPNN is compared to a conventional Multi-Layer Perceptron (MLP) that is trained on the whole operating region. Simulation results indicate that both schemes have the same performance, but, the training of CBDPNN is much easier than the conventional MLP, although it has about 50% more neurons in its hidden layer. Both schemes can reasonably handle an increase in fault deteriorations. However, the training time for CBDPNN is much less than that of MLP. This issue gets severely important when the number of measured variables, along with process faults, increases. Since, for plant-wide fault diagnosis, the reduction in training time is crucial, the advantages of CBDPNN make it more appropriate for fault diagnosis compared to other alternatives for such a case.