Study On Train System On-line Identification And Predictive Control

The Automatic Train Operation System(ATO)can control the train operation in place of experienced drivers.The ATO system is used to realize the safe,reliable and efficient operation of the train.An accurate description of the train model and effective control algorithm of the train speed is the basic of the ATO system with good performance.And the train operation quality such as stability and energy-saving can also be improved.From the perspective of the actual engineering application,a study on train system on-line identification based on the recursive least squares(RLS)algorithm is proposed.It can solve the problem of time-varying parameter identification and the problem of parameter identification with noise.On this basis,this paper studies the application of generalized predictive control(GPC)algorithm in the ATO system over the uncertainty of train operation process.So as to provide basis for improving the performance of automatic train control and realizing the optimal control of ATO system.The main contents of this paper include the following points:1.The ATO system was simplified through the study of the function and the basic structure of the ATO system.And the train model structure is established on the foundation of the train single point model and train operation force analysis.2.The AM-VFF-RLS algorithm is applied to identify the train system parameters.It can solve the problem of time-varying parameter identification and the problem of parameter identification with noise.The recursive least square method is used to identify the parameters of the model.The variable forgetting factor is used to enhance the tracking ability of the model.And the accuracy of parameter identification is improved by auxiliary model.The real-time running data collected from Hangzhou Line 4 are used as an example for the system identification simulation.The results prove the validity and superiority of the train model and parameter on-line identification method used in this paper.3.The principle and characteristics of GPC and its shortcomings in industrial application are studied.On this basis,the GPC algorithm is optimized.Based on the conventional GPC,stair-like control scheme with input constraints is used and the predicted future changes are considered as feedback.Besides,AM-VFF-RLS algorithm is used to identify the prediction model.And the actual train input constraints are considered to calculate the train traction/braking force.In this way,the accuracy and stability of the train automatic operation process is ensured.4.Simulation of ATO system based on GPC using MATLAB programming.Based on the running data from Hangzhou Line 4,the stair-like control based GPC optimal algorithm is applied to control the speed of the train.And the results are compared with the traditional GPC algorithm.The simulation results show that the train speed control algorithm proposed in this paper can meet the tracking requirements of the ATO system.And it provides practicability for its application in ATO system.