密集信号环境下的雷达辐射源信号交叠概率增大,传统的聚类算法往往无法揭示信号特征集所表现出来的多样性结构,不能满足雷达辐射源信号分选识别的要求。因此,结合雷达辐射源信号分选的本质,以聚类的紧凑度和连通性指标构建多目标函数,在P系统优化理论的启发下,提出了膜框架下的粒子群多目标优化算法。该算法在各基本膜中融入粒子群算法,执行粒子多目标搜索策略;通过膜间溶解规则,在表层膜中利用非支配排序和拥挤距离机制提高算法的快速收敛性,保持解集的多样性和差异性 最终,提出的算法用于求解雷达辐射源信号符号熵特征数据集的Pareto最优解集,实现信号的多目标聚类分选。仿真结果表明:该算法获得了较好的雷达辐射源信号分选识别正确率,验证了此算法的有效性和可行性,其性能优于传统聚类方法。

Due to the increased overlapping probability of radar emitter signals in a dense signal environment, traditional clustering algorithms can′t reveal the diversity structure of the signal feature set or meet the requirements of radar emitter signal sorting and recognition. Therefore, combined with the essence of deinterleaving for radar emitter signals, a multi-objective function is developed with clustering compactness and connectivity indicators. Inspired by the P system optimization theory, a P system-based multi-objective particle swarm optimization algorithm is proposed under the membrane frame. This algorithm incorporates particle swarm algorithm into each elementary membrane and executes particle multi-objective search strategy; through the dissolution rules between membranes, the non-dominated sorting and crowding distance mechanism are used in the skin membrane to improve convergence of the algorithm and keep the diversity and difference of the solution set. In the end, the proposed algorithm is used to obtain the Pareto optimal solution set of the symbolic entropy feature set of radar emitter signal. Simulation results show that the algorithm has better accuracy rate of deinterleaving and recognition for radar emitter signal. Moreover, the proposed algorithm is effective and feasible, and its performance is superior to traditional clustering methods.

TN957.51

国家自然科学基金资助项目