Adaptive fuzzy control of electric multiple unit for accurate stopping by considering power loss partly
The partial power loss caused by unknown faults in train actuators,model uncertainties and unknown external disturbances may reduce stopping accuracy,consequently impacting operational efficiency. To address the issues of accurate stopping and velocity tracking control problem under the aforementioned conditions,a saturated adaptive control method was proposed for Electric Multiple Unit (EMU). First,considering the nonlinearity of the train,the multi-mass train dynamics models with unknown actuator failures was established. A smooth function with a bounded approximation error was used to approximate the saturation function and converted the non-affine system into an affine system based on the mean-value theorem. Second,combining adaptive fuzzy logic system and back-stepping method,an adaptive fuzzy fault-tolerant controller that does not need actuator fault information was designed based on the fault model. The adaptive fuzzy logic systems are used to approximate unknown external disturbances,model uncertainties,and other system functions. By using the norm estimation method,each fuzzy logic system only needs one adaptive parameter,reducing the complexity of the system adaptive update law. Then,it was proven via the Lyapunov's stability theory that all the signals in the closed-loop system are bounded. By selecting appropriate controller parameters,the tracking error was converged to any small neighborhood near the origin. Final,the CRH5 EMU was selected as the control object for simulation. Simulation data shows that the controller can ensure the smooth operation of the EMU,while keeping the stopping error within±0.5 cm,and the control input is always within the restricted range. The simulation results demonstrate that the designed approach can achieve the control objective and suppress the influence of external interference and model uncertainties in the presence of actuator faults and input saturation.
electric multiple unitmulti-mass modelaccurate stoppinginput saturationfault-tolerant controladaptive fuzzy control
万方数据
维普
