摘要
针对粒子群算法(PSO)在校正阵列天线阵元位置误差时容易陷入早熟与局部收敛的问题,结合入侵杂草算法(IWO)的搜索广度与PSO的搜索深度,提出一种新的基于IWO-PSO算法的阵元位置误差校正方法。根据毫米波辐射计热辐射信号弱的特点,该方法在低信噪比阵列误差模型下,通过一个方位精确已知的外部辅助源进行有源校正,将一组阵列视为一株杂草进行PSO中位置与速度更新,并繁殖新的种子正态分布生长在父代杂草附近从而完成循环迭代。实验结果表明,方法稳定提高了全局搜索能力,估计所得全局最优阵元位置与实际阵元位置相对误差仅为0.0006,并且收敛速度优于单独PSO校正方法。
Aiming at the problem that the particle swarm algorithm(PSO) is easy to fall into the premature and local convergence when calibrating the position errors of the array sensors, the search breadth of the invasive weed algorithm(IWO) is combined with the search depth of the PSO in this paper. It proposes a new sensor position errors calibration method based on the IWO-PSO algorithm. In view of the weakness of millimeter wave radiation signal's energy, an external auxiliary source which is accurately known azimuth information is used for active calibration under low SNR array error model. In this method, a set of arrays is considered as a weed. It updates position and velocity in PSO algorithm, and the new seeds are distributed normally around the parent weed, thus completing the cyclic iteration. The experimental result shows that the stability of this method improves the global search capability, and the error between the estimated global optimum sensor location with actual sensor location is only 0.0006. In addition, the convergence speed is better than the single PSO correction method.
出处
《微波学报》
CSCD
北大核心
2017年第S1期306-309,共4页
Journal of Microwaves
