Rough Set Application for Reliable Fault Diagnosis of Power Plant Heater
Information shortage and redundancy coexist in the power plant fault diagnosis. Rough set theory is applied to analyze and to optimize the redundant information for improving reliability of expert system. Each long rule in the knowledge base is decomposed to many short rules according to the rough set reduction. Then the selected short rules are recombined to a new rule to replace the original one. With less preconditions, the new rule can give more reliable diagnosis conclusion. Moreover, optimum redundant information is remained by recombining the short rules. Reliability analysis proves that the recombined rule is more reliable than the original one. The method is applied to power plant feed water heater fault diagnosis system. The application shows that the more unreliable preconditions are, the more reliable recombined rule is than the original one.
rough set fault diagnosis reliability power plant
Yu Gang Chen Qing
Key laboratory of Condition Monitoring & Control for Power Plant Equipments, North China Electric Po School of Computer Science and Technology, North China Electric Power University Baoding, PR China
国际会议
太原
英文
165-168
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）