基于蚁群算法和遗传规划的跨单元调度方法

针对运输能力受限的跨单元调度问题，提出了一种基于蚁群算法与遗传规划的超启发式算法.通过蚁群算法搜索合适的启发式规则，并且利用遗传规划生成可以适用于问题模型的启发式规则，用以扩充规则集；同时引入时间窗的概念，用来决策每个小车运输时的等待时间.实验表明，提出的算法可以搜索出优质规则，并且通过遗传规划很大程度上改善了候选规则集，提升算法性能.同时时间窗策略的采用可以提高小车的利用率以及最小化总加权延迟时间. 

Intercell Scheduling Approach Based on Ant Colony Optimization Algorithm and Genetic Programming

To deal with the intercell scheduling problem with limited transportation capabilities, a hyper-heuristic approach was developed based on an ant colony optimization and genetic programming algorithm. The ant colony optimization algorithm was used to search the appropriate heuristic rules for the addressed problem. And the genetic programming was used to generate well-performing heuristic rules as an extension to the predefined candidate heuristic rules. Meanwhile, a time window was introduced into the proposed algorithm to determine the vehicle waiting time for a batch processing. Experimental results show that the ant colony optimization algorithm can search the outperforming combinations of the heuristic rules, the heuristic rules generated via genetic programming can obviously improve the quality of the candidate rule set, and that the time window can obviously improve the vehicle efficiency and minimize the total weighted tardiness, therefore providing better performance.