成像侦察星座的优化对于侦察时效性具有重要意义.当前侦察星座优化采用基于Pareto支配的进化算法,针对此类算法在解决优化目标函数维度大于 3 的侦察星座优化问题上出现的选择压力不足、多样性差等问题,提出一种改进的基于模糊关联熵的粒子群算法(Improved Particle Swarm Optimization Algorithm Based on Relative Entropy of Fuzzy Sets,IFREM-PSO).算法对自适应惯性权重策略进行改进,增强了收敛速度与收敛精度;引入变异策略,有利于跳出局部最优解;对外部档案维护策略进行改进,增强多样性.以面向区域目标的可见光侦察星座的设计与优化为背景,分别使用多目标粒子群算法 MOPSO(Multiple Object Particle Swarm Optimization)、基于模糊关联熵算法FREM-PSO(Particle Swarm Optimization Algorithm Based on Relative Entropy of Fuzzy Sets)和本文提出的IFREM-PSO对侦察星座进行优化.实验结果表明,FREM-PSO 算法在该问题上具有更好的表现,而相比 FREM-PSO 算法,IFREM-PSO算法在收敛速度上有显著提升,在收敛效果和多样性上表现更好.
Optimization of imaging reconnaissance constellation based on relative entropy of fuzzy sets
The optimization of imaging reconnaissance constellation is of great significance for reconnaissance timeliness.At present,evolutionary algorithms based on Pareto are often used in the optimization of reconnaissance constellation.In order to solve the problems of insufficient selection pressure and poor diversity of such algorithms in the optimization of reconnaissance constellation whose objective function dimension is greater than three,the improved particle swarm optimization algorithm based on relative entropy of fuzzy sets(IFREM-PSO)is proposed.The algorithm improves the adaptive inertial weight strategy and enhances the convergence speed and accuracy.The introduction of variation strategy is conducive to jumping out of the local optimal solution.Improve external archive maintenance policy to enhance diversity.Based on the design and opti-mization of regional target-oriented visible light reconnaissance constellation,MOPSO,FREM-PSO and the IFREM-PSO are used to optimize the reconnaissance constellation.The experimental results show that the algorithm based on fuzzy relative en-tropy has a better performance in this problem,and compared with FREM-PSO algorithm,IFREM-PSO algorithm has a sig-nificant improvement in convergence speed,and a better performance in convergence effect and diversity.
国家科技期刊平台
NSTL
万方数据
