Research On WSN Intelligent Fault Diagnosis With Neural Network And Fuzzy Logic

With the development of wireless communication,sensor,MEMS and distributed technology,wireless sensor network(WSN)technology integrated with the above technology,has a wide application prospect in the natural environment protection,health monitoring,military security of flexible deployment,strong reliability,good economy and so on.Due to the special application background,WSN nodes are vulnerable,which will result in WSN performance degradation or failure,and cause network paralysis seriously.Therefore,the fault nodes of WSN and causes are diagnosed accurately and timely,the proposed restoration scheme can effectively ensure the reliable operation of WSN.This paper has carried on the thorough research to the WSN fault diagnosis,and the research content has the following several aspects:? With the research of rough set theory,the fault types and corresponding fault characteristics of WSN fault nodes make the decision tables,and by the advantage of Sets Rough(RS)theory to eliminate the redundant attributes,the improved attribute reduction algorithm,and fully integrated the advantages of the improved RBF neural network to overcome the noise interference and parallel computation,the integrated rough set and improved RBF algorithm(IRSRBF)is proposed to diagnose the WSN nodes with limited energy and uncertain fault types.Experimental results show that the IRSRBF algorithm can effectively improve the reliability and practicability of the system.? Fault diagnosis is further studied in the wireless data transceiver unit which is the most difficult WSN node fault diagnosis,this paper proposed a fault diagnosis technology of wireless data transceiver unit on fuzzy neural network.According to the relationship between the transmitting current consumption and temperature,supply voltage of the wireless data transceiver unit,a current model is build up;Then the structure of the fuzzy neural network model is determined by the clustering algorithm,and the front and back parameters of the fuzzy rules are optimized by the hybrid learning algorithm;Finally,the fuzzy neural network parameters are extracted to establish the WSN node fault diagnosis model.The experimental results show that the fault diagnosis method of the wireless data transceiver unit has the characteristics of small amount of calculation and high diagnostic accuracy.