| The transmission-line in power system is used to transfer electric energy, which high stability in the entire power system, is of major importance obviously. Transmission-line is also the most frequent faults in power system, if can’t take effective protective measures, may lead to paralysis of the entire power system. How to diagnose the failure of transmission-line fast and accurately, that has been an important issue among researchers.The types of failure and its mechanism of the transmission-line have been analyzed and extracted characteristics of mechanism analysis firstly, and used the analysis results as failure criterion. Secondly, for the issues that complex of information process and single method for diagnosis in real-time and accuracy problem, gave a serially method of plausible inference and neural network-- fuzzy neural network algorithm. Finally, the fuzzy neural network fault diagnosis model is build according to the combination method, and the fault diagnosis system by using diagnosis algorithm of transmission-line based on the Matlab, Simulated experiment. The system had the following characteristics:(1) By using sequence current and sequence voltage algorithm, which not only able to identify the fault phase, but also can identify whether the ground fault, greatly reduced the amount of calculation.(2) By establishing the model of series, both at the same time, the use of two diagnostic methods, foster strengths and circumvent weaknesses, but also avoided the interference of two methods. And in the fuzzy inference system, simplified the number of output vector, relieved the pressure of neural network training, increased the diagnosis efficiency of the system.For verifying the validity of the system model, by validating a large number of test samples in a laboratory environment. The results shows that the fuzzy-neural network system identified herein can make right recognition of all transmission-line short-circuit fault type, achieved the desired goals. |
PDF Full Text Request