摘要
为进一步提高虚拟应答器(VB)技术在基于卫星定位的列车运行控制中的运用效果,分析常规VB捕获原理,从完善其不足的角度提出1种基于列车运行状态组合预测的新型VB捕获方法。首先,明确新型方法的捕获原理;其次,考虑列车运行状态的复杂性和规律性,采用交互多模型(IMM)结合容积卡尔曼滤波(CKF)的自适应IMM-CKF方法,对列车位置进行滤波估计和短期前向预测;然后,将长短期记忆方法引入列车运行状态预测问题,基于实时数据更新预测模型;最后,得到基于列车运行状态组合预测的VB捕获方法,有效改善常规方法中的VB漏捕获和捕获精度低等问题。基于京沈高铁现场试验数据,对比并验证新型方法和其他列车运行状态预测方法、VB捕获方法的性能。结果表明:尽管周期捕获计算平均用时较长,但新型方法下的VB捕获识别位置与下一个VB位置较近,不仅能保证100%的VB捕获率,而且在VB捕获误差均值上最大可实现70.23%的优化效果。
To further improve the application effect of Virtual Balise(VB)technology in train operation control based on satellite positioning,the conventional principle of VB capture is analyzed,and a new VB capture method based on the combination prediction of train operation state is proposed from the perspective of improving its shortcomings.Firstly,the capture principle of the new method is clarified.Secondly,by considering the complexity and regularity of the train operation state,the adaptive IMM-CKF method based on the combination of Interactive Multiple Model(IMM)and Cubature Kalman Filter(CKF)is adopted to conduct the filtering estimation and short-term forward prediction of the train position.Then,the Long Short Term Memory method is introduced into the prediction for the train operation state,and the prediction model is updated based on the real-time data.Finally,the VB capture method based on the combination prediction for train operation state is obtained,which effectively improves problems including VB missed capture and lowaccuracy capture in conventional methods.Based on the field test data from the Beijing-Shenyang high-speed railway,the performance of the new method is compared and verified with other methods for train operation state prediction and VB capture.The results show that although the average time of periodic capture calculation is comparatively long,the recognition position of VB capture with the new method is close to the next VB position,which can not only ensure the 100%VB capture rate,but also achieve a maximum optimization effect of 70.23%on the average VB capture error.
作者
王剑
王思琦
蔡伯根
刘江
程君
WANG Jian;WANG Siqi;CAI Baigen;LIU Jiang;CHENG Jun(School of Electronic and Information Engineering,Beijing Jiaotong University,Beijing 100044,China;Beijing Engineering Research Center of EMC and GNSS Technology for Rail Transportation,Beijing Jiaotong University,Beijing 100044,China;School of Computer and Information Technology,Beijing Jiaotong University,Beijing 100044,China;Signal and Communication Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China)
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2023年第1期202-213,共12页
China Railway Science
基金
国家自然科学基金资助项目(61873023,U1934222,62027809)
中国国家铁路集团有限公司科技研究开发计划项目(P2020G005)
北京市自然科学基金-丰台轨道交通前沿研究联合基金资助项目(L191014)。
关键词
列车运行控制系统
虚拟应答器
卫星定位
列车运行状态
组合预测
Train operation control system
Virtual Balise
Satellite positioning
Train operation state
Combination prediction
