基于协同控制理论的列车间隔实时调整优化方法研究

Train headway adjustment is an important approach to decrease train delay,improve rail transportation system’s efficiency and service quality. Based on given line and train condition,there are many feasible choices, which are larger than minimum headway, for train headway adjustment. Currently, rail transporation system has a hierarchical architecture, which is from Control Center to Train.Under the limitation of the hierarchical architecture, traditional methods are insufficient on response time and efficiency. Moreover, minimum headway between successive trains is not considered.In this proposal,we focus on two problems: to propose an effective method to optimize train headway based on cooperative control, and to analyse the impact of commmunication on the proposed method.By giving sensible assumption that trains could communicated with each other,we will propose a train headway adjustment model based on cooperative control with considering minimum headway. Then, we will design an hierarchic algorithm, which satisfies the real-time requirement, to find proper solutions. Secondly, the effect of random communication delay on the proposed method will be analysed, especially when one train can’t communicated with others. Finally, the proposed method will be verified in a real wireless communication system environment for rail transportation system.The proposal will study train headway adjustment with train-to-train communication， that will enrich the theory on train headway adjustment.

列车间隔调整是减少列车延误时间，提高轨道交通运输效率和服务质量的重要手段。给定线路条件和列车条件，列车运行间隔在最小追踪间隔的约束下存在多个追踪间隔值可供列车间隔调整选择。现有间隔调整方法受限于轨道交通“中心-列车”层次化控制结构，响应时间较长，效率不高，调整时很少考虑最小追踪间隔约束。本项目在提出车车通信假设的基础上，针对“面向系统实时状态变化，设计快速有效的算法”和“通信环境对此列车间隔调整方法的影响”问题展开研究，内容包括：基于协同控制理论，考虑最小追踪间隔约束，建立列车间隔调整模型，设计满足实时性要求的分层智能求解算法；分析随机通信延迟对列车间隔实时调整效果的影响以及单列车失去通信情况下列车行车效率问题；基于实际通信环境验证列车间隔实时调整算法效果。本项目的研究将探索车车通信条件下列车间隔调整方法，进一步丰富列车间隔调整理论。

列车间隔调整是减少列车延误时间，提高轨道交通运输效率和服务质量的重要手段。给定线路条件和列车条件，列车运行间隔在最小追踪间隔的约束下存在多个追踪间隔值可供列车间隔调整选择。现有间隔调整方法受限于轨道交通“中心-列车”层次化控制结构，响应时间较长，效率不高，调整时很少考虑最小追踪间隔约束。为此，本项目基于现场采集的数据，首先，构建了考虑最小追踪间隔约束的基于协同控制的多列车控制模型，对现场采集数据进行拟合和聚类分析，提出了基于Gamma分布的车车通信延误模型假设。其次，设计了分层智能求解算法，上层构建双阶段随机规划模型来确定运行曲线；下层提出了一种小波包滤波和迭代学习相结合的列车自动驾驶控制方法。随后，对考虑通信延迟的列车协同控制效果进行了分析，最后，基于实际通信系统验证了方法的有效性。本项目的研究丰富了车车通信条件下列车间隔调整方法和理论。