Swarm Intelligence Optimization for Solving Combinational Optimization Problems with Complex Constraints and Its Practical Application to Railway Transportation

现实世界中的组合优化问题有许多复杂的约束条件。在铁路运输系统中，需要同时考虑车辆的运用和检修计划，在优化过程中必须构造出满足定期检修约束的组合解。在这项研究中，我们发展了一种基于蚁群优化的方法，作为一种处理周期性约束的有效方法，并将其应用于铁路客车车辆规划的实际应用中。此外，我们还实现了局部搜索，并将其扩展用于处理多目标问题。然后，我们的方法也被扩展到处理更复杂的动车组组合问题。

A real-world combinational optimization problem has many complicated constraints. In railway transportation systems, which need to consider operation of railway vehicles and their maintenance scheduling simultaneously, the combinational solutions which meet periodical maintenance constraints have to be constructed in optimization process. In this research, we developed the ant colony optimization-based approach as an efficient method which can treat periodical constraints, and applied it to a railway rolling stock planning for passenger trains as a real-world application. In addition, we implemented local searches in our approach and extended it for treating multi-objective problem. Then our approach was also extended for tackling the more complicated problem withtrain-sets composition.