Research On Diagnosis Of The Insulator Running Status Based On Fuzzy Neural Petri Net

Insulator is an important component device in the transmission system, by the atmospheric pollution factors in normal working cases, the insulator pollution flashover accident occurred frequently, directly affected the safe and stable power supply system. At present the study indicate that. leading to the insulator fault many factors and mutual relationships are complex, difficult to effective and accurate monitoring, therefore, to defile the degradation of insulator discharging and flash reguar issues of diagnosis and prediction to come will be in the field of electric power transmission hot research topic.This article analysis of insulator reguar outside the occurring mechanism and its surface uncleanness discharge the development process, combination with intelligent diagnosis techniques and methods, and summarizes the malfunction caused insulator fault features and volume of the complex relationship between them, studies the pollution flashover discharge the characteristics of each stage, established a comprehensive accurately and insulator malfunction diagnosis reasoning system, the design process is shirt-sleeve fuzzy production rule type, Petri network and neural network theory, makes the design of insulator running state diagnosis system has good reasoning and self-learning ability, and the design of reasoning system to a certain research, finally the diagnosis system simulation program proved the diagnostic system accuracy and feasibility. This paper made three aspects of trial.(1) Sums up the causes of the factors and mutual insulator the complicated relationship, select a facilitate diagnosis system, use the diagnosis system after a comprehensive reasoning and analysis can be very accurate, diagnosed the insulator filthiness degradation degree, risk degree and flash collaterals.(2) Using fuzzy neural Petri nets diagnostic systems has the advantages of these theories, making the diagnosis reasoning and network learning to be more perfect. System modular structure, and the input parameters are defined and fuzzed, function in the form of computational reasoning ensure the accuracy of the diagnosis result.(3) Object-oriented technology is one of the characteristics of VC + + 6.0 language and use it to the characteristic of designed a set of design and in front of the diagnosis model corresponding diagnosis system simulation program, which proves this diagnosis model accuracy and feasibility. Final diagnosis analysis results, and find the system needs to improve and perfect the aspect, for future research direction.