RES脉内特征的差分进化粒子群投影寻踪评价模型  

Intrapulse Feature Evaluation Model of Radar Emitter Signal Based on Differential Evolution,Particle Swarm Optimization and Projection Pursuit Algorithm

作　　者：朱斌 金炜东[2] 余志斌[2] 

机构地区：[1]长江师范学院电子信息工程学院,重庆408100 [2]西南交通大学电气工程学院,四川成都610031

出　　处：《西南交通大学学报》2018年第1期189-196,共8页Journal of Southwest Jiaotong University

基　　金：国家自然科学基金资助项目(61134002;60971103)

摘　　要：针对雷达辐射源信号脉内特征综合评估存在标准单一、缺乏客观性等问题,提出了基于群体智能的雷达辐射源信号脉内特征综合评估模型.首先,通过投影寻踪算法将雷达辐射源信号脉内特征的综合评估问题转化为有条件限制的多元非线性目标函数的优化问题;其次,通过改进的粒子群优化算法与差分进化算法的结合得到新的智能算法;最后,利用该算法实现多元非线性目标函数的优化求解.仿真结果表明:该群体智能算法对Rosenbrock测试函数的最优适应度值最小,对Rastrigrin函数和Girewank测试函数的最优适应度值为0,说明该算法的计算精度优于其他算法.同时适应度值的方差比标准粒子群算法和差分进化算法小,说明该算法的收敛性和鲁棒性较好.通过与加速遗传算法对评估问题目标函数5次优化结果的比较,本算法的计算结果没有波动,说明基于群体智能的RES脉内特征综合评估模型能够更客观、更有效地实现对RES脉内特征的综合评估.To address the problems in comprehensive evaluation of radar emitter signal(RES)intrapulse features,such as incomplete evaluation criteria and the lack of objectivity,a new comprehensive evaluation model of RES intrapulse features was proposed based on swarm intelligence.First,the comprehensive evaluation problem of RES intrapulse features was converted into an optimization problem of the conditional multivariate nonlinear objective function through the projection pursuit algorithm.Secondly,the new swarm intelligence algorithm was obtained through the combination of the improved particle swarm optimization algorithm and the differential evolution algorithm.Thirdly,the optimization and solution of the multivariate nonlinear objective function was achieved using the proposed algorithm.The simulation results show that the optimal fitness of the Rosenbrock test function of this new intelligence algorithm is minimal,and the optimal fitness values of the Rastrigrin test function and the Girewank test function are zero,indicating that the calculationaccuracy of the proposed algorithm is better than the standard particle swarm optimization algorithm and the differential evolution algorithm.At the same time,the variance of the fitness value of the proposed algorithm is smaller than those of the standard particle swarm optimization algorithm and the differential evolution algorithm,which indicates that the convergence and robustness of the proposed algorithm are better.According to the objective function of the evaluation problem,the results of the proposed algorithm do not fluctuate when comparing the five optimization results with those of the accelerated genetic algorithm,which shows that the intrapulse feature evaluation model based on swarm intelligence can effectively achieve the objective of comprehensive evaluation of RES intrapulse features.

关 键 词：雷达辐射源信号 特征评估 群体智能 粒子群优化 投影寻踪 

分 类 号：TN957.51[电子电信—信号与信息处理]