摘要
针对生产当中的热轧钢调度问题,使用热轧钢轧制批量计划问题的多旅行商模型(MTSP)进行建模,并且通过引入虚拟节点将转化为标准的ATSP问题。通过引入基于统计物理学中自组织概念的改进的分支策略,结合基于深度优先阈值加广度随机搜索的搜索策略,提出了一种改进的智能分枝定界算法,并应用于解决该ATSP问题,实验结果表现出了较高的效率和可行性。
The Hot Steel Rolling Plan is firstly modeled as Multi-Traveling-Salesman Problem(MTSP) model.Then,we transfer the MTSP model to the standard Asymmetric Traveling Salesman Problems(ATSP) model through introducing some virtual codes,meanwhile,an intelligent Branch-bound Algorithm(which combines a new branching algorithm based on SOC theory and a strategy that mix threshold constraint based depth-first-search and weighted random breadth-search) is proposed to solve the ATSP problem.Computational results using the data of the hot strips show the efficiency and promising future of this optimization method.
出处
《微型电脑应用》
2009年第4期39-41,38,共4页
Microcomputer Applications
基金
国家自然科学基金(60574063)
关键词
热轧调度
ATSP
分枝定界
阈值
修补算法
Hot Rolling Scheduling
ATSP
Branch-bound algorithm
Threshold
Patching algorithm
