基于自适应蚁群算法的装配线平衡问题研究  

An Adaptive Ant Colony Optimization for Solving Assembly Line Balancing Problem

作　　者：邓福平[1] 张超勇[1] 连坤雷[1] 徐绍锬[1] 

机构地区：[1]华中科技大学数字制造装备与技术国家重点实验室,武汉430074

出　　处：《中国机械工程》2011年第16期1949-1953,1959,共6页China Mechanical Engineering

基　　金：国家自然科学基金资助重点项目(51035001);国家自然科学基金资助项目(50875101);国家高技术研究发展计划(863计划)资助项目(2007AA04Z107)

摘　　要：提出了一种自适应蚁群算法,用以求解装配线平衡问题。在该算法中,针对装配线平衡问题的具体特点,设计了一种蚂蚁分配方案可行解的构造策略,提出了一种比传统方法区分度更高的评价解质量的目标函数,同时为了克服蚁群算法易陷入局部最优和收敛速度慢等缺陷,通过自适应地调整算法的挥发度等系数,在保证收敛速度的条件下提高了解的全局性。最后,通过实例验证,证明了算法的可行性和有效性。An adaptive ant colony optimization was proposed to solve the assembly line balancing problem（ALBP）.According to the characteristics of the ALBP,a method of solution constructing strategy was developed,and a better differentiation of objective function was proposed to appraisal solution quality.However,general ant algorithm often falls into local optimal and consume excessive time,in order to overcome these shortcoming,an improved ACO was presented by adaptive adjustment of the parameters in the algorithm,which has a good ability of searching better solution at higher convergence speed.Finally,the proposed algorithm was tested and compared against best known algorithms reported in the literatures,and the experimental results indicate the feasibility and effectiveness of the proposed algorithm.

关 键 词：装配线平衡 蚁群算法 自适应 人工智能 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]