基于协同指挥的高铁智能调度理论与方法研究

Train dispatching command system that consists of multiple subsystems including transportation, locomotive, infrastructure, electricity, and vehicle systems plays a critical role in operation of high speed railways. In order to promote the capability and ability of train dispatching system to meet the demand of “Intelligent High-speed Railways” in terms of intelligence and delicacy operations, in particular, under the events of emergence, the proposed project investigates dynamic dispatching in complex high-speed railway networks by addressing the following three fundamental questions: (1) evolution mechanism of a complex high-speed railway network’s operation situation, (2) delicacy and intelligent integrated planning and timetabling, (3)data-driven multi-disciplinary cooperation rescheduling. Based on the concept of Cyber-Physical Systems, this project is to develop a framework of multi-tiered cross-disciplinary collaborative commanding and intelligent dispatching system under the condition of tight spatio-temporal constraints and fast-changing transportation resource limitations. Main tasks are: (1) dry running and situation evolution study of high-speed railway operation; (2) integrated line planning and timetabling, (3) multi-granularity prediction and evaluation of operation status under disruptions. (4) reinforcement learning for intelligent rescheduling and (5) optimization and development of simulation platform and validation. In summary, the proposed project will be able to provide both systematic theories and enabling technologies to leverage the intelligence and collaborative abilities of high-speed railway dispatching command system to address disruptions, which is of both theoretical and practical importance.

高铁调度指挥系统是高速铁路运行的核心，是涉及车-机-工-电-辆多专业多工种的复杂信息系统。为进一步提调度指挥系统的智能化、精细化水平和突发事件处置能力，更好满足“智能高铁”的重大需求，本课题以成网条件下高铁调度指挥为研究对象，围绕高铁“智能调度”的复杂路网运营态势演化机理，动态精细化、智能化的高铁调度计划一体化编制和大数据驱动多专业协同的调度计划智能调整三个科学问题，从信息物理系统的角度出发，构建多级耦合、时空约束强、运输资源约束动态演化条件下，基于协同指挥的高铁智能调度理论体系。主要研究内容包括高铁运营态势推演、“开行方案-运行图”一体化编制、突发事件下运行状态的多粒度预测评估、基于强化学习的高铁智能调整策略学习与优化和仿真平台搭建与实验验证。本项目为提高突发事件下调度指挥系统的协同能力和智能化水平提供理论支撑和技术保障，具有重要理论意义和工程应用价值。

高铁调度指挥系统是高速铁路运行的核心，是涉及车-机-工-电-辆多专业多工种的复杂信息系统。为进一步提高调度指挥系统的智能化、精细化水平和突发事件处置能力，更好满足“智能高铁”的重大需求，本项目以成网条件下高铁调度指挥为研究对象，围绕高铁“智能调度”的复杂路网运营态势演化机理，动态精细化、智能化的高铁调度计划一体化编制和大数据驱动多专业协同的调度计划智能调整三个科学问题开展研究，并取得了一系列理论和应用成果，包括：提出了动态运输资源约束下高铁开行方案和运行图一体化编制理论和技术，构建了多时空粒度和路网拓扑特征的高速铁路运营态势推演方法，构建了数据驱动的集成化、规范化的运行状态在线综合评估和预测系统，以及建立了时空强约束条件下的高铁智能优化调度理论方法与快速实现技术，最终形成基于协同指挥的高铁智能调度理论体系，并在此基础上，完善了智能调度集中系统，研发了列车开行方案与运行图一体化编制云平台系统、调度集中系统车务终端仿真培训系统、以及高铁建设铺轨调度原型系统，并实际应用于部分高铁线路。本项目取得的理论成果和提出的技术方案，有效提升了突发事件下高速铁路调度指挥系统的协同能力和智能化水平，为高速铁路调度的智能化奠定了坚实基础。