期刊文献+

基于再生制动能高效利用的列车运行方案优化

Optimization of Train Operation Plans Based on Efficient Utilization of Regenerative Braking Energy
下载PDF
导出
摘要 随着城市轨道交通的快速发展,列车运行能耗问题已成为行业关注的焦点。传统的节能控制方法在大规模列车智能协同方面存在一定的局限性。为了优化总体能耗,尤其是有效利用再生制动能,研究提出了一种结合微观列车驾驶策略与宏观列车运行方案的优化求解框架。通过考虑列车运行过程中发车、站间操纵及停站等关键因素,深入刻画列车在单向运行过程中的全链路控制环境,构建多列车协同工况优化模型。引入一种多列车牵引制动重叠时间框架,提出一种融合多头注意力机制的多智能体协同深度确定性策略梯度算法(Multi-headed Attention Mechanism&Multi-Agent Deep Deterministic Policy Gradient,Mam-MADDPG)求解框架,快速实现多列车运行能耗优化求解。使用北京地铁亦庄线数据进行仿真验证,结果表明,提出的Mam-MADDPG方法在再生制动能利用方面提升近20%的节能效率,且具备较强的稳定性。这一研究为城市轨道交通的节能减排提供了新思路。 With the rapid development of urban rail transit,the issue of train operation energy consumption has been recognized as a focal point of concern in the industry.Traditional energysaving control methods have limitations in handling the intelligent coordination of large-scale trains.To optimize overall energy consumption,especially the effective utilization of regenerative braking energy,an optimization framework combining microscopic train driving strategies with macroscopic train operation plans was proposed in this study.By considering key factors such as train departure,inter-station maneuvering,and station stops during the entire train operation process,a comprehensive control environment for trains in a unidirectional operation was thoroughly characterized,and a multi-train coordination optimization model was constructed.A multi-train traction-braking overlapping time framework was introduced,and a solution framework based on a multi-headed attention mechanism&multi-agent deep deterministic policy gradient(Mam-MADDPG)algorithm was proposed,enabling the rapid optimization of multi-train operation energy consumption.Simulation validation using Yizhuang Line data of Beijing Subway shows that the proposed Mam-MADDPG method improves energy-saving efficiency by nearly 20%in terms of regenerative braking energy utilization while demonstrating strong stability.This research provides a new approach to energy conservation and emission reduction in urban rail transit.
作者 王若愚 周慧娟 秦勇 孙璇 张尊栋 张蛰 WANG Ruoyu;ZHOU Huijuan;QIN Yong;SUN Xuan;ZHANG Zundong;ZHANG Zhe(State Key Laboratory of Advanced Rail Autonomous Operation,Beijing Jiaotong University,Beijing 100044,China;School of Traffic and Transportation,Beijing Jiaotong University,Beijing 100044,China;Beijing Key Lab of Urban Intelligent Traffic Control Technology,North China University of Technology,Beijing 100144,China)
出处 《铁道运输与经济》 北大核心 2024年第12期76-87,共12页 Railway Transport and Economy
基金 国家自然科学基金重点项目(61833002)。
关键词 城市轨道交通 列车节能 再生制动能 多智能体强化学习 深度确定性策略梯度算法 Urban Rail Transit Train Energy Saving Regenerative Braking Energy Multi-Agent Reinforcement Learning Deep Deterministic Policy Gradient Algorithm
  • 引文网络
  • 相关文献

参考文献1

二级参考文献3

共引文献9

;
使用帮助 返回顶部