Research Of Fault Location And Protection In Smart Distribution Grid

Since the 21st Century, the power industry has been developing rapidly with the demand for power quality and supply reliability increasing gradually in China. The distribution system is a connection between power transmission network and power consumer, which means a partial power cut or large area blackout would happen if a fault occurs. In our country the small current neutral system (SCNS) is mainly employed by the distribution network and the probability of single-phase ground fault accounted for 60% of the total faults or even higher. The rapid and accurate realization of single-phase earth fault line selection and location can effectively shorten the blackout time as well as reduce the human consumption, which is of the important practical significance. It is also one research direction for electric power department and scientific research workers.This paper studies the transient zero current section characteristics based on the analysis of steady and transient characteristics of the small current neutral grounding system. According to the difference of transient zero sequence current on both sides of the fault point and the identification of singular points through wavelet transform, a method to detect the fault section based on wavelet transform modulus value polarity was proposed. The simulation on the platform of PSCAD/EMTDC shows that the method can realize fault location very well.In the aspect of fault location, the method using single fault information is easily affected by external factors for the weak fault current. So a fault location model of wavelet neural network whose input includes zero sequence fundamental current, transient wavelet energy and transient reactive power is constructed. Meanwhile we can get the optimal initial value of neural network’s parameters and solve the convergence problem by using the immune particle swarm algorithm to optimize the wavelet neural network. The MATLAB simulation results show that the optimized WNN positioning method can effectively improve the learning efficiency and fault location accuracy compared with BP neural network and WNN. And at the end of the paper, a brief analysis of distributed and centralized control theory based on the protection of distribution network automation is given.