列车运行调整微粒群算法研究

Study on the Particle Swarm Optimization Algorithm for Train Operation Adjustment

列车运行调整问题是铁路行车调度指挥工作的重要内容,决定着区段内行车秩序的优劣。这一问题的计算机自动求解算法是我国铁路信息化建设的一个核心技术和难点问题。本文依据我国铁路行车组织体制的特点,建立了相应的模型。在模型的求解过程中,先运用大系统理论将列车进行分层分级,从而将待解的原始问题分解成若干个子问题,在对分解后的问题进行求解时,设计了微粒群算法,运用该算法可快速得到各子问题的近似最优解。然后,应用系统原理对问题进行还原,即可快速得到一个满意度高、可用性强的列车运行调整方案。最后,采用现场数据,应用该算法对列车运行调整问题进行求解,并与遗传算法进行比较,结果表明微粒群算法解决列车运行调整问题高效、实用。

As part kind of the important work for railway dispatching, train operation adjustment is the key problem to keeping the orderliness in sections. The automatic solution algorithm for the problem is a core technology and also a difficult issue in informationization construction of China railways. A model is set up according to the characteristics of the train operation management system in our country in this paper. During the model solving process, trains are divided into different classes and grades by the overall system theory, and the original problem is split into sub problems which are then solved by the PSO （Particle Swarm Optimization） algorithm rapidly, yielding approximate optimal results. The problem is restored by applying the system theory. As a result, the plan for train operation adjustment is obtained with high satisfaction and applicability. Site data are calculated by the proposed algorithm and comparison is made with the genetic algorithm for train operation adjustment. The result demonstrates that PSO is more efficient and practicable.