求解混流装配线调度问题的蚁群算法  

An Ant Colony Algorithm for Mixed-model Assembly Lines Scheduling Problems

作　　者：杨田田[1] 蔡洌[1] 李世其[1] 

机构地区：[1]华中科技大学机械科学与工程学院工业工程系,武汉430074

出　　处：《计算机与数字工程》2005年第11期5-9,29,共6页Computer & Digital Engineering

基　　金：香港蒋氏工业慈善基金资助项目;国防863项目804-5-15.3

摘　　要：以最小化总的传送中断时间为目标函数的混流装配线调度问题是丰田生产方式中自动化概念的一个重要问题,而新颖的蚁群算法具有通用性、鲁棒性、并行搜索以及易于与其他启发式算法结合的优点,可以解决多种组合优化问题,对其进行了改进,以便更适于求解混流装配线的调度问题。实验表明:改进的蚁群算法解决了混流装配线的调度问题,得到了优于分支定界法、模拟退火法和遗传算法的可行解。Solving the mixed- model assembly lines scheduling problem, whose objective function is the minimization a total conveyor stoppage time, is the most important goal for the concept of Autonomation in Toyota production systems. The main desirable characteristics of ant colony algorithm are versatile, robust, parallel searching and easily combining with other certain heuristics. The ant colony algorithm also can solve the combinatorial optimization problems. It is improved in this paper, so that it is easily used to solve mixed - model assembly lines scheduling problem. The results show that the solution of this new algorithm is better than the one which branch- and - bound algorithm, simulated annealing algorithm and genetic algorithm.

关 键 词：混流装配线 自动化 调度 蚁群算法 

分 类 号：O22[理学—运筹学与控制论]