Memory/Learning Based Control Method And Applications To Pantograph-Catenary Systems

The operating mileage and the train speed are the two improvements of China Railway,which lies great importance on the reliability,stability and safety of train operation.The relationship of pantograph-catenary is one of the three key factors that restrict the train speed and affect the dynamic performance of the train.So the evaluation criteria,namely,the pantograph-catenary current collection quality,has been widely concerned.As the method of parameter optimization can not anymore meet the requirements of pantograph-catenary current collection quality under the high speed situation,so the introduction of active control and semi-active control is inevitable.The application of semi-active control is rare in pantograph-catenary system while it is common in many fields such as vehicles,bridges and buildings.Therefore,the study on active control and semi-active control of pantograph have great practical significance and widely application prospect.In addition,most of the existing research is based on linear model,which do not conform the reality of the pantograph-catenary system.In this thesis,considering the external disturbance and the uncertainty of pantograph-catenary system,a control method is designed,which is independent of the exact system model.The main research contents of this thesis are as follows:Firstly,since the structure of catenary is complicated,it is not easy to establish the model.So the catenary model is simplified,and then the catenary is regarded as a variable stiffness spring in this thesis.In order to make the pantograph-catenary system model closer to reality,the pantograph is dealed with the flexibility model,instead of simply equivalent to the spring-mass-damper system,so that a pantograph-catenary coupling non-linear model is established in this thesis.Secondly,in view of the non-linearity of the system,memory control is introduced,which does not rely on the exact model.Memory control also has the characteristics of small storage space and the storage space will not increase with time going on.A memory/learning based controller is designed after the human memory learning process of human and the principle of memory control.The analysis of the stability proves that the tracking error is bounded.Matlab/Simulink simulation results validate the control effect of the control method applied to pantograph-catenary system and the comparison with the PID control shows that the proposed method has great advantage over PID.The influence of the controller parameters on the control effect is analyzed,then a group of better parameters are obtained.Subsequently,the simulation results are further analyzed in this thesis.It is found that the contact force has a large initial error,which is successfully improved by introducing PI control.Finally,due to semi-active control requires much less energy than active control,but its control effect is similar to active control,so the semi-active control of pantograph-catenary system is studied.In this thesis,the magneto rheological damper is used to realize the semi-active control of the pantograph-catenary system,and the semi-active control model is established.After that the controller designed above is applied to the semi-active control of the pantograph-catenary system.Then,the simulation results of Matlab/Simulink unfold the better performance of memory/learning based semi-active control than PID semi-active control.