Study On Modeling Of Pantograph-catenary System Based On Least Squares Support Vector Machine

With the rapid development of high-speed railway in today’s society,there is a growingrequirement for the performance of basic railway equipments. The pantograph-catenarysystem as a major component of electric locomotive power supply will affect the currentcollecting quality of electric locomotive, which is associated with train stability. In order tomake the pantograph-catenary system have higher quality of contact and current collection,the modeling problem for pantograph-catenary system is studied based on least square supportmachine(LSSVM), the main work of this thesis is as follows(1)At first,the discrete equation of pantograph-catenary system is obtained based on itscoupling dynamic equation. Second,the generalized model of pantograph-catenary system isbuilt by considering the inherent characteristics of pantograph-catenary system and theenvironment factors.(2)The generalized model of pantograph-catenary parameters are identified by the use ofLSSVM. And the generalized pantograph-catenary system is used to study the influence oftrain speed and pantograph equivalent mass for the contact force. In addition, empirical studyis implemented on the field measured data of pantograph-catenary used on existing railwayline. The simulation results show that, the pantograph-catenary model based on LSSVM caneffectively approximate the dynamic behavior of the contact force in pantograph-catenarysystem with higher prediction accuracy. In a word, the proposed method is effective inmodeling of train pantograph-catenary system.(3) In order to improve the prediction accuracy of generalized pantograph-catenarymodel, a method based on error correlation is proposed for the optimization of LSSVM model,which is used in the study on modeling of generalized pantograph-catenary system. Based onthe actual testing data of pantograph-catenary system, the contact force is modeled andpredicted by the use of error correlation. The results show that, it improves the predictionaccuracy with the method of LSSVM based on error correlation and achieves a goodprediction results, indicating that the proposed method for pantograph-catenary system isfeasible.