Study of Fault Tolerance Performance in Fault Diagnosis System Based on NN Model and Data Mining Model
In practical application of intelligent fault diagnosis, mis-diagnosis may be caused by realtime information that is distorted in the process of their generation or transfer. The probability of mis-diagnosis is dependent on the fault tolerance performance of intelligent principle used in diagnosis system. This paper researches the fault tolerance performance of fault diagnosis system based on neural network (NN) model and data mining (DM) model. It is discussed by diagnosis system of High voltage transmission line system (HVTLS). In DM model, the qualitative analysis ability of rough set (RS) theory is used to analyze knowledge region data set and the redacts of RS are solved by Genetic Algorithm (GA). In order to get the assurance of fault tolerance performance of tested diagnosis system and have practical value of studied system, this paper proposes the theory criterion of building test samples. The high fault tolerance performance of proposed approach is proved through comparison with that of NN-model based fault diagnosis system.
HVTLS fault diagnosis fault tolerance performance neural network data mining rough set
Sun Yaming Liao Zhiwei
School of Electrical Automation and Energy Engineering,Tianjin University, Tianjin, 300072,China
国际会议
8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)
上海
英文
620-625
2001-11-14（万方平台首次上网日期，不代表论文的发表时间）