改进遗传算法在DOA搜索中的应用

在对短时、低信噪比信号测向时。加权子空间测向算法（WSF）的统计性能优于Music，Esprit等测向算法，且可以对相干信号进行检测。但是因为算法实现涉及到非线性多维搜索。因此运算量很大。本文提出的改进遗传算法，是一种运算速度较快的加权子空间算法的实现算法。该算法将免疫算法引入遗传算法中，并与MVP算法相结合。解决了普通遗传算法（CGA）在DOA搜索中易早熟。后期搜索效率低的问题，拥有快速、全局的搜索能力。计算机仿真证明了该算法的有效性和鲁棒性。

Application of improved genetic algorithm in DOA search algorithm

In direction finding of burst and low SNR signal,the statistic performance of weighted subspace fitting algorithm(WSF) is better than that of Music and Esprit algorithms,which can also find direction of coherent signal.Since nonlinear multi-dimensional search is involved in this algorithm,a large quantity of computation is needed.An improved algorithm is proposed in this paper for realization of WSF.This algorithm can effectively exploit some experiences in evolution,so it greatly improved the convergence performance of GA algorithm.Furthermore,this algorithm combined immune MVP algorithm to solve the low efficiency problem in local search of GA algorithm,which had a rapid and global search capability.The simulation results proved the effectiveness and robustness of the improved algorithm.