A Novel Localization Method for the Partial Discharge of Transformer based on UHF

Partial discharge(PD) diagnosis and location is helpful to find the fault location of high-voltage power device in the early aging stage, and has important reference value for making maintenance plan. Now, many ultra high frequency (UHF) are often used to form sensor array, and three-dimensional coordinate system is defined to locate the partial discharge of transformer. In this paper, the principle of capturing the initial pulse of signal based on energy accumulation method and positioning based on time difference of arrival (TDOA) algorithm is studied. A nonlinear programming problem with the coordinates of partial discharge points as unknowns is constructed, and the tabu search particle swarm optimization (TS-PSO) algorithm is used to solve the optimal solution. The algorithm can avoid the problems of nonconvergence and nonuniqueness in solving nonlinear equations and the high initial value requirements of least square method. It can not only ensure the speed of solution, but also ensure the uniqueness and accuracy of solution. Laboratory and on-site tests verify the effectiveness of the positioning results.