A NEURAL-NETWORK APPROACH TO FAULT-DETECTION AND DIAGNOSIS IN INDUSTRIAL-PROCESSES

Using a multilayered feedforward neural-network approach, the detectio n and diagnosis of faults in industrial processes that requires observ ing multiple data simultaneously are studied in this paper. The main f eature of our approach is that the detection of the faults occurs duri ng transient periods of operation of the process, A two-stage neural n etwork is proposed as the basic structure of the detection system, The first stage of the network detects the dynamic trend of each measurem ent, and the second stage of the network detects and diagnoses the fau lts, The potential of this approach is demonstrated in simulation usin g a model of a continuously well-stirred tank reactor. The neural-netw ork-based method successfully detects and diagnoses pretrained faults during transient periods and can also generalize properly, Finally, a comparison with a model-based method is presented.