铝型材自动化立体仓库优化调度算法的研究

【作者】 曹辉；

【导师】 熊红云；

【作者基本信息】 中南大学 ， 控制科学与工程， 2012， 硕士

【摘要】 自动化立体仓库出入库调度问题是近年来的研究热点,目前关于铝型材自动化立体仓库出入库调度问题的研究比较少。本文以南平铝业自动化立体仓库为背景,对立体仓库的出入库调度优化问题进行了较深入的研究,建立了该立体仓库的出入库数学模型,提出了最短路径优化算法以及最少倒垛优化算法,仿真结果验证了算法的有效性。论文对南平铝业自动化立体仓库的布局和结构进行了简单阐述,根据具体情况建立了出入库数学模型。针对立体仓库的货位分配问题提出了基于货品-货位耦合分配的分类均匀存储策略,实现了同种型材可以多台堆垛机作业,加快了型材的出入库速度。在货位分配已优化的基础上,出入库货位给定的情况下,提出了一种启发式遗传算法,解决了出入库调度的最短路径优化问题,采用最佳个体保存策略加快了算法的收敛速度,通过MATLAB仿真验证了算法的有效性,其优化效果强于传统遗传算法。将出入库可优化的时间与倒垛时间进行了比较,得出倒垛优化更有利于提高仓库作业效率,为此建立了倒垛数学模型。结合基于动态垛位的最优垛位选择,提出了一种求解最少倒垛次数的遗传算法。算法采用的最优垛位选择法是选择可交换型材组中倒垛次数最少的垛位,加快了算法的搜索收敛速度,提出了自适应交叉与变异策略,改善了群体结构的多样性,仿真实例证明了该遗传算法能很好地解决铝型材立体仓库的倒垛问题,对立体仓库实际调度系统开发具有指导性。更多还原

【Abstract】 The scheduling problem of AS/RS has been a hotspot research in recent years, the relatively studies on scheduling problems of Aluminum AS/RS is not sufficient. This thesis has Nan ping Aluminum AS/RS as the research background, Conducts a more in-depth study on access schedu-ling optimization problem of AS/RS.Establishes a mathematical model of the Aluminum AS/RS, pproposes the shortest path optimization algorit-hm and the least turned-out slab stack optimization algorithm. The simul-ation results have demonstrated the effectiveness of the algorithm.The thesis briefly introduces the layout and structure of Nan ping Aluminum AS/RS, establishes the mathematical model of the stacker task scheduling according to its own characteristics.focus on cargo space problem of the Aluminum AS/RS, it proposes the uniform classification storage strategy based on good-cargo coupled allocation, makes the same kind of profiles which can be operated in different stackers come true,and accelerates the speed of enter and out of storage.In the basis of the optimized of cargo space allocation and the certain position of warehouses out-in storage, we put forward a new theory about using a heuristic genetic algorithm to solve the shortest path of cargo’s in-out. We uses the best individual preservative strategy to speed up the algorithm convergence and verified the algorithm’s efficiency through MATLAB simulation. In conclusion, the optimization effect is stronger than the traditional genetic algorithm.According to compare the time can be optimized of stacker shortest path problem and the time of turn-out slap stack problem, the turn-out slap stack optimization is more conducive to improve the efficiency of warehouse operations, so establish a mathematical model for the turn-out slap stack optimization. Combinating the optimal stack selection method based on dynamic stack, we use genetic algorithms to reduce the times of the turn-out slap stack optimization to the least. The optimal stack selection method is to select the least times of the turn-out slap stack in the exchangeable Aluminum group.Using adaptive crossover and mutati-on strategy in order to improve the diversity of the group structure and accelerate the speed of convergence of the genetic algorithm. The examp-le shows that the turn-out slap stack problem of the Aluminum AS/RS can be optimized by the genetic algorithm. This paper can be a guidance for optimize the scheduling system in AS/RS in practice.更多还原