Research On The Train Lateral Semi-Active Suspension System Using Generalized Predictive Control

Currently, raising the running speed of the train constantly increases the amplitude of the vibration of the vehicle body during operation, so that smooth running of the vehicle is reduced. Practice shows that:China’s trains’vertical smooth basically meet the requirements of speed, however, the lateral vibration deterioration leads to steady poor performance by the traditional passive suspension system parameters limitations, which puts forward higher requirements to lateral suspension system of speed trains and high-speed trains [1]. Semi-active suspension control is an effective way to improve train lateral stability. Using semi-active control strategy to control the train’s suspension system can effectively improve the lateral vibration, thus improving the stability and ride comfort has great practical significance.In this thesis, using semi-active suspension to control the lateral vibration of the train was simulated. In order to build the train semi-active suspension control system’s simulation system, the thesis started from researching the basic principles of train vibration. The direction and the level of irregularity signal that cause the vehicle lateral vibration were simulated using mathematics method; Then simulink model of suspension system of the17-freedom-degree train was constructed in MATLAB environment; Semi-active control strategy for the train’s suspension system was studied, and an improved generalized predictive control algorithm (Improved General Predictive Control, IGPC) was designed; Finally, the control system of train’s semi-active suspension was built using MATLAB/Simulink tool and simulated in MATLAB.The specific content mainly consists of the following:(1) First, the need of semi-active suspension system for high-speed trains was analyzed. Through analysising the railway vehicle’s passive suspension system, active suspension control system as well as semi-active suspension system’s principles and characteristics, and comparing the advantages and disadvantages of the three kinds of suspension control systems, it is concluded that semi-active suspension control system in the running train is superior and necessary. Secondly, the common control strategies of semi-active suspension system were analyzed;(2) The direction and the level of irregularity signal that have greater affect on the vehicle lateral vibration were simulated using mathematics method, the analytical expressions of track irregularities’spectral density is the German track spectrum. And then the simulation model of lateral semi-active suspension control system of the17-freedom-degree high-speed train was built up in MATLAB/SIMULINK environment, which was based on motion differential equations that describe the characteristics of high-speed train.(3) The predictive control’s basic knowledge and generalized predictive control’s principle were studied, and the drawbacks associated with generalized predictive control were analyzed, such as, solving Diophantine equations and inversing matrix calculation will lead to large online calculation. The paper solved generalized predictive control’s performance index function, improved predicted output, and designed an improved Generalized Predictive Control Algorithm (IGPC). And the parameter’s selection of the algorithm was studied, which should be taken special attention in application. Finally, the control system for the train’s lateral semi-active suspension was designed by using the established the German’s track spectrum and high-speed trains’Simulink model, combined with improved generalized predictive control algorithm (IGPC).(4) Semi-active suspension system based on Improved generalized predictive control (IGPC), the semi-active suspension system based on generalized predictive control and PID (GPC_PID) and passive suspension system were simulated in different speeds cases. According to the related evaluation of high-speed train’s ride comfort, the simulation results were analyzed.The simulation results show that:the evaluation indicators of the high-speed train’s semi-active suspension system based on IGPC are better than the passive suspension system and semi-active suspension system based on GPC_PID, therefore, the method can effectively improve the lateral vibration of the train. The effectiveness of the method is verified in this paper.This thesis is supported by the key project<Study on active control for rolling stock’s vibration>(2008TPL_Z04) that is undertaken by Southwest Jiaotong University’s Traction Power State Key Laboratory.