中文摘要：

针对花朵授粉算法易陷入局部极值、收敛速度慢的不足,提出一种具有量子行为的花朵授粉算法.该算法通过引入量子系统的态叠加特性,用波函数描述种群个体的位置,利用势肼场使种群个体以一定的概率密度在可行空间任何区域进行搜索,并且利用种群的平均最优位置使种群间存在等待效应,提高种群的协同工作能力,从而使算法能有效地避免陷入局部最优,增强全局寻优能力,提高收敛速度.通过8个CEC2005benchmark测试函数进行测试比较和3个数值积分的求解,并对结果进行分析,仿真结果表明,改进算法的全局寻优能力明显优于基本的花朵授粉算法、差分进化算法和蝙蝠算法等,其收敛精度、收敛速度和鲁棒性均比对比算法有较大提高.

英文摘要：

In order to resolve the problems of easily relapsing into local extremum and low speed of convergence of flower pollination algorithm,aquantum-behaved algorithm was proposed.In this algorithm,the superposition feature of quantum system was introduced,the position of the population individual was described with wave function,the individual in population was made to be searched in feasible space,at any area,and with a certain probability density by means of potential hydrazine,and the optimal mean position of the population was used to make the waiting effect exist among the propulation,so that the cooperative work ability of population would be improved,local optimization was avioded global searching ability was enhanced,and convergence speed was increased.By means of measurement and comparison of 8CEC2005 benchmark functions,solution of 3numerical integration,and analysis of simulation result,the simulation of flower pollination was conducted and its result shows that the proposed algorithm would have better global searching ability,faster convergence and more precise convergence than basic flower pollination algorithm,difference evolution algorithm and bat algorithm,exhibiting agreater improvement in convergence accuracy and speed,and robustness as well.