作　　者： ; ; ;

机构地区： 佛山科学技术学院电子与信息工程学院

出　　处： 《暨南大学学报（自然科学与医学版）》 2014年第1期82-87,共6页

摘　　要： 多峰函数的寻优能力一直是衡量算法优越性的依据之一,许多传统的算法对其求解容易误将局部极值当做全局极值,针对这一问题,以Schaffer多峰函数为例,运用改进的果蝇优化算法(Improved Fruit Fly Optimization Algorithm,IFFOA)对Schaffer函数进行寻优求解,同时对其他4个测试函数进行求解.通过matlab软件测试、分析了改进的果蝇算法寻优能力的影响因素,最后通过相同的参数,分别采用遗传算法(GA)和改进的果蝇算法进行对比,通过对比发现,改进的果蝇优化算法在收敛精度和收敛速度方面,均优于和声搜索算法(HS)算法,从而验证了IFFOA算法的优越性和有效性,可以将IFFOA算法应用于其他领域. The optimization capability of multi-peak function is one of the basis for judging algorithm superiority.The local extreme value often was regarded as the whole extreme value in many traditional al-gorithms.In order to solve this problem,the Improved Fruit Fly Optimization Algorithm (IFFOA )was used to solve optimization of schaffer function and the other four testing function by taking the schaffer function for example.The affect factors of optimization capability of IFFOA were analyzed by testing of Matlab software.The convergence accuracy and convergence speed of IFFOA were better than those of harmony search algorithm(HS)by comparing genetic algorithm(GA)and IFFOA with the same parame-ters.Therefore,the superiority and the effectiveness of IFFOA were verified.The IFFOA can be applied to other fields.

关 键 词： 果蝇优化算法 多峰函数 仿真 遗传算法 寻优能力

领　　域： [自动化与计算机技术] [自动化与计算机技术]