改进模型预测控制的列车自组网运行控制  

Train Ad Hoc Network Operation Control Based on Improved Model Predictive Control

作　　者：宋宗莹[1] 杨迎泽[2] 王兴中 于晓泉 李烁 武悦 胡超[2] SONG Zongying;YANG Yingze;WANG Xingzhong

机构地区：[1]中国神华能源股份有限公司,北京100011 [2]中南大学电子信息学院,长沙410075 [3]中南大学交通运输工程学院,长沙410075 [4]长沙理工大学电气与信息工程学院,长沙410114

出　　处：《科技创新与应用》2025年第5期22-26,共5页Technology Innovation and Application

基　　金：国家自然科学基金项目(62177046)。

摘　　要：重载组合列车自组织网络是一种新型智能化的管理体系,能够实现列车的实时信息交互和协同运动。随着列车向长编组、重载化发展,对于重载组合列车自组织网络提出新的要求。该文提出一种改进模型预测控制的列车自组网运行控制策略。首先,建立重载组合列车自组织网络系统的单列车与多列车运动模型,其中多列车运动模型以3辆车为例具体化。其次,设计一种基于模型预测控制的多列车协同控制,以实现多列车的高效率安全运行,并且通过优化模型预测控制预测时域,改善控制效果。最后,仿真实验验证所提方案在加速、减速和不同预测时域的效果。结果表明,所提出控制效果的优越性,在保证运算性能的同时达到最优的控制效果。The self-organizing network of heavy-haul train is a new intelligent management system,which can realize real-time information interaction and cooperative movement of trains.With the development of long train formation and heavy load,new requirements are put forward for the self-organizing network of heavy-haul trains.This paper presents a train self-organizing network operation control strategy based on improved model predictive control.Firstly,the single-train and multi-train motion models of the self-organizing network system are established.The multi-train motion model takes three cars as an example.Secondly,a multi-train cooperative control based on model predictive control is designed to achieve high efficiency and safety of multi-train operation,and the control effect is improved by optimizing the prediction time domain of model predictive control.Finally,simulation experiments verify the effectiveness of the proposed scheme in acceleration,deceleration and different prediction time domains.The results show the superiority of the proposed control effect and the optimal control effect can be achieved while ensuring the operational performance.

关 键 词：重载组合列车 自组织网络 模型预测控制 优化预测时域 运行安全 

分 类 号：U296[交通运输工程—交通运输规划与管理]