Operation Curve Optimization And Sliding Mode Tracking Control For Urban Rail Trains

The urban rail transit has been developing rapidly and increasingly for its large capacity,high efficiency,safety,comfort and punctuality.However,the costs for construction and maintenance of an urban rail transit are rather high.Especially,traction power consumption accounts for a large part of the maintenance cost.Therefore,it is of great significance to optimize the train speed curve.At the same time,due to the high train running speed and operating density,the working requirements for drivers are becoming stricter as well.The Automatic Train Operation(ATO)system can greatly reduce the pressure of drivers and thus improve the operational efficiency and performance accordingly.For an ATO system,the most important function is the target speed curve tracking.Designing a reasonable and effective ATO tracking control algorithm is the key to realize the automatic operation of the train.In this thesis,the optimization and control of urban rail trains are studied from two aspects:the speed curve optimization and the speed curve tracking.The main contributions of this thesis are as follows:(1)The single particle motion model and train operation calculation model are established,and the energy-saving operation optimization model based on Warning Speed Curve(WSPC)is designed.Compared with the traditional multi-objective energy-saving optimization model,the proposed model effectively reduces the search interval,and thus improves the solving efficiency.This lays the solid foundation for solving the optimization model.(2)To address the shortcomings of traditional genetic algorithms widely used in train running curve optimization,a Random Reinforcement Genetic Algorithm(RR-GA)is proposed to solve the energy-saving optimization model based on the warning speed curve.Without losing elite individuals,the novel RR-GA increases the group diversity and reduces the probability of falling into the local optimum.(3)For the ATO target speed curve tracking,most existing PID and fuzzy controllers rely highly on complex parameter adjustment,while the robust adaptive method can hardly ensure the finite time convergence and is not suitable for the fixed time interval trains.To tackle this issue,a Double Sliding Surface Guidance(DSSG)is developed for the finite time tracking of velocity curves with bounded disturbances.The Lyapunov theory is used to prove that the DSSG can be progressively stable in a finite time.Finally,with the operating background and line data of Beijing Yizhuang subway,multiple scenario simulations are carried out.The simulation results verify the effectiveness and superiority of the proposed RR-GA and DSSG.