Study On Adhesion Control Method Of High Speed Motor Vehicle Based On Least Square Method To Estimate The Slope Of Adhesion-slip Curve

In order to ensure good adhesion performance between wheels and rails and give full play to the traction and braking force of EMU trains,it is necessary to implement reasonable and efficient adhesion control for EMU trains.In this dissertation,a recursive least squares algorithm with switchable forgetting factor is studied to identify the slope of stick-slip curve from the perspective of practical engineering application.In addition,aiming at the problem that the traditional gradient method deduces the slower speed near the extreme point,the step size is set as a function related to the curve slope,and the gradient method is selected to deduce the optimal creep speed,which provides a basis for the efficient adhesion control of EMU trains.The main contents of this paper include the following points:(1)The single axle longitudinal dynamic model of EMU train is established.The correctness of the model is verified by simulation based on the actual parameters of CRH2 A EMU in MATLAB.Aiming at the problem that the adhesion coefficient is difficult to obtain in real time,a full-dimensional state observer is constructed.The adhesion coefficient can be obtained indirectly by observing the load torque of the motor.The actual values calculated by the observed values and the adhesion force calculation models at different poles are compared in the simulation.The results show that the error between the observed values and the actual values is the smallest when the pole value is-300.(2)Aiming at the problem that the traditional recursive least squares(RLS)method does not track the identification parameters in time and lags behind when the condition of the rail surface changes suddenly,the traditional recursive least squares method is improved,and a recursive least squares algorithm with switchable forgetting factor is designed to track the slope value of the stick-slip curve.The simulation results show that the improved algorithm has the characteristics of fast tracking and small fluctuation.(3)In this dissertation,the gradient method is used to estimate the optimal creep velocity.In order to solve the problem that the traditional gradient method has a slower estimation speed near the extremum due to the smaller slope,the search step is set as a function related to the slope value,which can increase the search step to improve the search speed when the slope of the curve decreases.In addition,a fuzzy PI controller is designed.The adhesion control of trains is realized.The simulation can simulate the different road conditions of train operation by changing the stick-slip characteristic curve to simulate the idle or sliding state of wheelset.(4)In order to verify the effectiveness of the adhesion control method proposed in this dissertation,the adhesion control model is built in MATLAB/Simulink to simulate and verify,and compared with the combined correction method.The simulation results show that the control method can make the train run in 0.5 seconds after the sudden change of rail surface conditions,and its control performance is better than the combined correction method.