阵元位置自校准的微粒群算法参数优化设置研究

Parameters optimum setting of PSO algorithm for self-calibration of array element position

针对微粒群算法中参数设置问题还没有成熟的理论和实验方法,提出一种基于田口方法的参数最佳设置实验方法.仿真研究了PSO算法中参数设置对阵列自校准性能的影响,给出了0～20dB信噪比条件下阵列阵元位置自校准的PSO算法的最佳参数组合及不分信噪比条件的最佳参数组合,并比较了MUSIC算法在阵元位置误差校准前后的DOA估计性能.仿真结果证明该方法是有效、可行的.

Aimed at there are still no well-developed theory and experimental method for the parameter setting in the PSO algorithm, this paper presents an experimental scheme for parameter optimum setting based on the Taguchi statistic theory. By simulation, it deals with the influence of the parameter setting in PSO algorithm on the self-calibration performance of array element position, gives the optimum parameter combination of the self-calibration of array element position in PSO algorithm in the case of 0-20dB for SNR and the optimum parameter combination without considering the condition of SNR, and compares the DOA estimation performance of MUSIC algorithm before or after error calibration of array element. Simulation results show that this proposed scheme is of feasibility.