蝙蝠算法在多目标多选择背包问题中的应用  

Application of Bat Algorithm in Multi-Objective and Multi-Choice Knapsack Problem

作　　者：李枝勇[1] 马良[1] 张惠珍[1] 

机构地区：[1]上海理工大学管理学院,上海200093

出　　处：《计算机仿真》2013年第10期350-353,共4页Computer Simulation

基　　金：国家自然科学基金资助项目(70871081);上海市研究生创新基金项目(JWCXSL1202)

摘　　要：研究蝙蝠算法在多目标多选择背包优化中的应用问题。针对传统的多目标多选择背包优化算法由于计算复杂度非常高,难以获得满意的解等问题,在蝙蝠算法的基础上,提出了一种改进的蝙蝠算法用于求解多目标多选择背包问题。算法设计中,首先引入了惯性因子作用于蝙蝠的速度,重新定义了蝙蝠的速度的更新方程,用来提高算法的收敛速度,然后给出了蝙蝠个体和群体更新的规则,引导蝙蝠向Pareto飞行。最后仿真结果表明,与粒子群算法相比,蝙蝠算法能够以更快的速度找到相同数目的 Pareto,体现出蝙蝠算法解决该问题的可行性和有效性以及蝙蝠算法性能的优越性,拓展了蝙蝠算法的应用领域。Study the application of bat algorithm in the multi - objective and multi - choice knapsack problem. The traditional algorithm for solving the multi - objective and multi - choice knapsack problem is of high computation- al complexity, and difficult to obtain satisfactory solutions. Based on the bat algorithm, an improved bat algorithm was proposed to solve the muhi - objective and multi - choice knapsack problem. In the algorithm design, an inertial factor was introduced to effect the speed of the bat, and the bat＇s speed update equation was redefined in order to im- prove the algorithm convergence speed. Then the bat＇s individual and group update rules were given to guide the bat to fly foreword the Pareto optimal solution. Finally, compared with the particle swarm algorithm, the simulation re- suits show that the algorithm can get the same number of Pareto optimal solution with faster speed, reflect the feasibil- ity and effectiveness of the bat algorithm to solve the problem and the performance advantages of the algorithm, and expand the application of bat algorithm.

关 键 词：蝙蝠算法 惯性因子 多目标 多选择 背包问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]