Real-time diagnosis of hydrogen gas system and seal-oil system of a large thermal turbo-generator using a fuzzy-set expert system

A diagnostic expert system with real-time knowledge-based functions for monitoring the performance of both hydrogen gas system and seal-oil system of a large thermal power turbo-generator is presented in this paper. The proposed diagnostic system evaluates the operating conditions of the studied systems in real time using several inputs from a number of sensors such as oil pressure, oil level, hydrogen purity, electrical signals, etc. When any abnormal situation occurs, the proposed diagnostic system may quickly and extensively analyze the operating conditions and discover the abnormalities. The proposed system can also provide intelligent suggestions to infer the specific equipment status and recommendations for remedial action. The operating data of a practical thermal power plant, the Shen-Ao 3rd generator set of Taiwan Power Company (TPC), Taiwan, are employed to justify the performance of the proposed diagnostic expert system. It can be concluded from the simulation results that the proposed expert system has the ability to make correct diagnosis in real time on the studied hydrogen gas system and seal-oil system under abnormal operating conditions.