基于列车运行仿真和SVR组合的地铁新线牵引能耗预测方法  

Traction energy consumption prediction of new metro lines based on simulation combining support vector regression

作　　者：周姗姗 柏赟[1] 袁博[1] 汪茜 李佳杰 ZHOU Shanshan;BAI Yun;YUAN Bo;WANG Qian;LI Jiajie(Key Laboratory of Transport Industry of Big Data Application Technologies for Comprehensive Transport,Ministry of Transport,School of Beijing Jiaotong University,Beijing 100044,China)

机构地区：[1]北京交通大学综合交通运输大数据应用技术交通运输行业重点实验室,北京100044

出　　处：《铁道科学与工程学报》2021年第10期2733-2740,共8页Journal of Railway Science and Engineering

基　　金：国家自然科学基金资助项目(71571016)。

摘　　要：新线牵引能耗预测可用以辅助线路的规划设计及列车选型,进而实现牵引节能。地铁新线牵引能耗受线路条件、列车属性等因素的综合影响,传统的仿真方法简化多种参数,而大数据预测方法忽略列车物理运动过程,均难以保证新线牵引能耗的预测效果。因此,提出一种将列车运行仿真与支持向量回归(SVR)相结合的组合预测方法预测地铁新线牵引能耗,并采用交叉验证方法及遗传算法对SVR参数进行寻优。研究结果表明:仿真与SVR相结合的组合预测方法优于单一预测方法,在95%的置信水平下,地铁新线牵引能耗预测精度可达90%。Traction energy consumption(TEC)prediction for new metro lines facilitates the decision-making of the line design and selection of train types,thus reducing traction energy.TEC of new lines is influenced by many factors,such as track profile,train characteristics,etc.Traditional simulation method simplifies many parameters,while the data-driven method ignores train movement,leading to relatively worse performance in energy consumption prediction.Therefore,this paper proposed a prediction method which combined simulation and support vector regression(SVR).Cross-validation method and genetic algorithm were applied to optimize SVR parameters.The case study shows the prediction accuracy reaches 90%with the confidence level of 95%by using the proposed method,which is better than the normal method.

关 键 词：地铁新线 牵引能耗 列车运行仿真 支持向量回归 

分 类 号：U29-3[交通运输工程—交通运输规划与管理]