多点收缩混沌优化方法及全局收敛性证明

Multipoint shrinking chaos optimization algorithm and its global convergence

针对目前混沌优化算法在选取局部搜索空间时的盲目性,提出一种具有自适应调节局部搜索空间能力的多点收缩混沌优化方法.该方法在当前搜索空间搜索时保留多个较好搜索点,之后利用这些点来确定之后的局部搜索空间,以达到对不同的函数和当前搜索空间内已进行搜索次数的自适应效果.给出了该算法以概率1收敛的证明.仿真结果表明该算法有效的提高了混沌优化算法的性能,改善了混沌算法的实用性.

A multipoint shrinking chaos optimization algorithm which local searching space can be decided under an self-adaptive contral strategy is proposed. The method keeps multiple better searching points at present searching space to decide its local searching space later. By this way the method have a self-adaptive on different functions and different times the search has carried out before. The global convergence of the algorithm are proved. Simulation results show that the algorithm can improve the chaos optimization algorithm＇s performance effectivly as well as make the chaos optimization more practical.