Identifying Lightning Channel-Base Current Function Parameters by Powell Particle Swarm Optimization Method

Abstract

This paper proposes the Powell (PPSO) method. The particle swarm optimization (PSO) algorithm is combined with the Powell algorithm to identify the parameters of lightning channel-base current function. The method can overcome the problem of premature convergence for PSO algorithm, and improve the analysis accuracy. The Heidler function is used to represent the lightning channel-base current. We have compared the PPSO method and genetic algorithm (GA) on reaching the desired peak value of the current I <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m</sub> and the maximum time rate of change of current (di/dt) <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> . The results show that the parameters of Heidler function evaluated by the PPSO method can achieve more accurate values of I <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m</sub> and (di/dt ) <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> than those evaluated by GA. Also, we have compared the PPSO method, the Nelder-Mead particle swarm optimization (NMPSO) method, and PSO method using the measured channel-base current. The results show that the PPSO method is better than the NMPSO and PSO method to determine the channel-base current function parameters. The PPSO method is an efficient method to identify the channel-base function parameters, which can improve the digitalization of lightning monitoring equipment. This approach is also useful in research related to lightning characteristics and lightning protection.