Study On Multi-modal Control In High-speed Train

With economic development, population density increases and urbanization processspeeds up, rail transit has played an increasingly important role in relieving the trafficpressure both in urban and intercity. But bigger running density and higher speed along withthe continuous development of high-speed rail technology, that lead to traditional drivingmode has been difficult to meet the demands. So advanced computer control technology wasused to meet the train traction, braking and parking requirements by the ATO, to make surethat trains will be Safe and accurate, achieve the purpose of automatic train operation.Therefore, it’s the development trend of the future to replace artificial driving by ATO.Train speed control is the most important part of ATO, there are a variety of controlmethods applied to train speed control, but train operation is a complex nonlinear process andthere has no precise mathematical model to describe, in this case, conventional controlalgorithm is difficult to achieve a satisfied self-control effect. It’s found that, compoundintelligent control shows a huge advantage and potential in train speed control system.This thesis summarizes background and development direction of train speed control systemand control algorithm both at home and abroad, based on analysis of basic principles ofvarious control methods, and the pros and cons of using in train speed control. Calculated aideal running curve including starting, constant cruising and parking brake with data ofCRH2-300. Simulated high-speed train controlled by fuzzy predictive to track ideal curve,results show that fuzzy predictive control has a rapid responsiveness, but larger steady-stateerror. In order to solve this problem, using fuzzy PID control possesses high-precision insteadof fuzzy control, which has good control accuracy but weak responsiveness.In order to have both rapid responsiveness of fuzzy predictive and high-precision ofpredictive fuzzy PID, make these two kinds of control play a different role in different timeby switching. In view of threshold switch cause discontinues of controlling quantity, softswitching control was used, fuzzy predictive control plays a larger role when error is larger, atthe rest of the time predicted fuzzy PID control plays a major role. Simulation results showthat composite soft switching control both fast and highly accurate. To better utilize theadvantages of the two control modes, add proportional control to composite control, when theerror is very big, increase weight coefficient of P control. Compared with Multi-model controlwhich without predictive control. Simulation results show that this multi-modal controllerachieve satisfactory results in step response control and high-speed train track ideal curve.