一种优化稀疏分解的雷达目标识别方法

摘要

雷达目标识别中雷达回波数据巨大,因此利用稀疏分解的方法对回波数据进行稀疏化处理.但稀疏分解中的匹配追踪算法存在计算复杂、计算量大的问题,所以汲取了粒子群优化算法(PSO)全局搜索能力强、收敛速度快的优点对最优原子的搜索过程进行优化,并且针对粒子群优化易陷入局部最优的问题,提出一种惯性权重自适应改变的改进解决方法.通过对雷达高分辨率距离像(HRRP)信号的稀疏表示实验仿真发现,基于粒子群优化的匹配追踪算法能大大缩短匹配追踪的时间,同时惯性权重自适应改变的方法也有效解决了PSO优化的"早熟"问题.%For the huge radar echo data in radar target recognition,the sparse decomposition method is utilized to perform the sparse processing for the echo data. The matching pursuit algorithm in sparse decomposition has the problem of complex com-putation and large calculated quantity,so the strong global searching ability and fast convergence speed of the particle swarm op-timization(PSO)algorithm are adopted to optimize the search process of the optimal atom. Since the PSO algorithm is easy to fall into the local optimization,an improved solution for the adaptive change of inertia weight is proposed. The sparse representation experiment of radar′s high resolution range profile(HRRP)signal was performed with simulation. It is found that the matching pur-suit algorithm based on PSO can significantly shorten the time of matching pursuit,and the adaptive change method of inertia weight can solve the "prematurity " problem of PSO algorithm effectively.