机构地区: 深圳大学信息工程学院
出 处: 《电子与信息学报》 2011年第2期429-434,共6页
摘 要: 该文提出基于实数编码模式的混合蛙跳算法(Shuffled Frog Leaping Algorithm,SFLA)求解容量约束车辆路径问题(Capacitated Vehicle Routing Problem,CVRP);把具有极强局部搜索能力的幂律极值动力学优化(PowerLaw Extremal Optimization,τ-EO)融合于SFLA,针对CVRP对τ-EO过程进行设计和改进。改进的τ-EO采用新颖的组元适应度计算方法;采用幂律概率分布来挑选需要变异的组元;根据最邻近城市表,采用幂律概率分布挑选变异组元的最佳邻近城市,执行线路间或线路内的变异。求解测试库中的实例,证明该改进算法有效。 An improved Shuffled Frog Leaping Algorithm(SFLA) is proposed to solve the Capacitated Vehicle Routing Problem(CVRP)based on real-coded patterns.It is then combined with the power-law Extremal Optimization() to further improve the local search ability.The fitness for the components of an individual is carefully designed and the neighborhood for mutation is established according to power-law probability distribution.Experimental results show that the proposed algorithm outperforms other heuristic algorithms base on PSO and GA.
关 键 词: 智能优化 进化算法 混合蛙跳算法 极值动力学优化 车辆路径问题 收敛性
领 域: [自动化与计算机技术] [自动化与计算机技术]