Fuzzy fault diagnosis system for a 200 MW turbo-generator set

Aiming at the low accuracy of current fuzzy diagnosis system, the paper analyzes the main cause of the low accuracy, and then proposes a new diagnosis strategy based on the varying-weight model of comprehensive evaluation and expanded membership function for fault diagnosis. The new diagnosis strategy has some essential difference from conventional systems in evaluation model and definition of fuzzy membership functions. Based on the new diagnosis strategy, a fault diagnosis system for 200 MW turbo-generator set is designed and programmed. The application results prove that the original idea is effective in solving practical problems. This project is supported by Natural Science Foundation of Guangdong Province (No. 994227, 980874).