Designing a Hierarchical Neural Network Based on Fuzzy Clustering for Fault Diagnosis of the Tennessee-Eastman Process

The paper proposes a hierarchical artificial neural network (HANN) for isolating the faults of the Tennessee-Eastman process (TEP). The TEP process is the simulation of a chemical plant created by the Eastman Chemical Company to provide a realistic industrial process for evaluating process control and monitoring methods The first step in designing the HANN is to divide the fault patterns space into a few sub-spaces by using Fuzzy C-means clustering algorithm. For each sub-space of fault patterns a special neural network was trained in order to do fault diagnosis. A supervisor network decides which one of each special neural network should be triggered. Each network in the proposed HANN has been given a specific duty so the proposed procedure is called Duty-Oriented HANN (DOHANN). The type of artificial neural networks used in the DOHANN is multilayer perceptron (MLP). Tennessi-Eastmann (TE) process was used to generate the training and test data. The performance of DOHANN was evaluated and compared to that of a conventional single neural network (SNN) and the method of dynamic Principal Component Analysis (DPCA). The simulation results indicate that the DOHANN diagnoses the TEP faults considerably better than SNN and DPCA techniques. Training of each neural network in the DOHANN is carried out more convenient than that of SNN because it includes structurally simpler neural networks.

کلیدواژه ها:

Hierarchical Neural Network ، Fuzzy C-means Clustering ، Tennessee-Eastmann Process ، Process Fault Diagnosis