The invention provides a freight train speed tracking and anti-skid control method based on an unscented Kalman observer. On the basis of a multi-mass-point model, the stress condition and the dynamic transfer process of the freight train in the running process are deeply analyzed, and a freight train dynamic model is established. Train running indexes such as energy consumption, punctuality and tracking precision during running of the freight train are fully considered, and optimal tracking of the target running speed is achieved by means of model prediction control. In order to enhance the adaptivity of the control system, a parameter adaptive adjustment mechanism is integrated in the MPC, and adaptive adjustment is performed on the optimization weight representing each driving index in combination with the driving condition of the freight train, so that adaptive model prediction control is formed. The parameter adaptive adjustment is realized through the Elman neural network, and the ENN can make correct weight adjustment on possible driving conditions through learning of excellent control weights.
本发明提出一种基于无迹卡尔曼观测器的货运列车速度追踪与防滑控制方法。基于多质点模型,深入分析货运列车行驶过程中的受力情况和动力学传递过程,建立货运列车动力学模型。充分考虑货运列车行驶时的能耗、准点和追踪精度等列车行驶指标,借助模型预测控制实现对目标行驶速度的最优追踪。为增强控制系统的自适应性,在MPC中集成了一个参数自适应调整机制,结合货运列车的行驶状况,对代表各行驶指标的优化权重进行自适应调整,构成了自适应模型预测控制。参数自适应调整是通过埃尔曼神经网络实现的,通过对优秀控制权重的学习,ENN能够对可能出现的行驶状况做出正确的权重调整。
Freight train speed tracking and anti-slip control based on unscented Kalman observer
基于无迹卡尔曼观测器的货运列车速度追踪与防滑控制
2023-08-22
Patent
Electronic Resource
Chinese
IPC: | B61L Leiten des Eisenbahnverkehrs , GUIDING RAILWAY TRAFFIC |
European Patent Office | 2022
|British Library Conference Proceedings | 2021
|SAE Technical Papers | 2021
|Unscented Kalman filter for visual curve tracking
British Library Online Contents | 2004
|