Research of Sensor Fault Diagnosis Method Based On Immune Network and Fuzzy ART Neural Network
A new immune network model and its diagnosis algorithm were studied. With the deficiency of model algorithm in the identification of sensor correlation, a new algorithm used in the quantitative extraction and recognition of sensor correlation were proposed based on Fuzzy ART neural network, of which the diagnosis system was consisted and the immune network..By the simulation of temperature sensor fault in certain thermal control system, the method was validated. The simulation result shows that the system could recognize and diagnose the faults accurately , regardless of single or multiple sensor faults. The accuracy of recognition and diagnosis is above 90 percent when the sensor output is less than±5 percent deviation.
Immune network models Diagnosis algorithm Fuzzy ART neural networks Sensor Correlation identification Fault diagnosis
Jihai Gu Ye Tian Xiangyang Jin
Department of Mechanical Engineering Harbin University of Commerce Harbin Heilongjiang, China, 15002 School of Mechatronics Engineering Harbin Institute of Technology Harbin Heilongjiang, China, 150001
国际会议
长沙
英文
1524-1527
2009-10-10（万方平台首次上网日期，不代表论文的发表时间）