基于模糊推理的粒子群优化算法

摘要:

针对无约束优化问题,提出了基于模糊推理的粒子群优化算法,该算法针对粒子群优化算法搜索能力的不足,先引入平均粒子,然后引入模糊推理来改进粒子群的速度更新公式,再利用模糊推理动态地改进算法惯性权重和速度更新公式的权重因子,再结合混沌扰动增加算法后期的局部搜索能力.数值试验采用12个测试函数并有5个算法进行对比,数值试验证明,改进算法的搜索能力有较大的提高.

It is well know that the standard PSO is unreliable to solve unrestraint problems with global optimization, due to the insufficient search ability nature of the PSO. In order to overcome the shortcoming, the particle swarm optimization algo-rithm based on fuzzy reason uses average particle, fuzzy reasoning to improve the speed of the particle swarm update formula. Fuzzy reasoning dynamically inertia weight and velocity updating formula of weight factor. Chaos disturbance increase algo-rithm local search ability in the late. A number of numerical examples which are used twelve trial functions have shown the present method is found to be ultra-accurate, capable of modeling global optimization.

作者:

史旭栋 高岳林 韩俊茹

Shi Xudong Gao Yuelin Han Junrun(School of Mathematics and Computer, Ningxia University, Yinchuan 750021,China Research Institute of Information and System Computation Science, Beifang University of Nationalities, Yinchuan 750021？China)

机构地区:

宁夏大学数学统计学院 北方民族大学信息与系统科学研究所

出处:

《betway官方app 学报：自然科学版》 CAS 北大核心 2017年第2期108-118,共11页

基金:

国家自然科学基金(61561001) 北方民族大学重点科研项目(2015KJ10)

关键词:

粒子群优化 模糊推理 信息共享

particle swarm optimization Fuzzy reasoning information sharing

分类号:

TP18 [自动化与计算机技术—控制理论与控制工程]

基于模糊推理的粒子群优化算法.pdf