Research On Adhesion Control Method Based On Slip Velocity Estimation

With the development of the world economy,there are higher requirements on the carrying capacity,speed limit,safety and stability of heavy haul trains,which also causes a lot of research on the operation performance and safety of heavy haul trains by many scholars at home and abroad.In the process of train operation,its maximum traction force or braking force is actually limited by the adhesion between the wheelset and the rail surface.Once the limit is exceeded,even if the traction or braking torque of the motor is increased,it will not help,but will lead to idling or coasting phenomenon of the train,which will greatly endanger driving safety.Therefore,an optimal adhesion control method based on slip velocity estimation is designed in this thesis to realize the search for the maximum adhesion point,and the torque controller is used to keep the vehicle at its peak efficiency.First of all,considering that it is difficult to obtain longitudinal speed of heavy haul train in actual operation,and the wheel speed obtained by wheel speed sensor is not exactly equal to the speed due to the existence of creep,a slip velocity estimation method of single-axle heavy haul train based on Square Root Cubature Kalman Filter is proposed.Compared with the filtering method without square root form,this method has higher estimation accuracy and stability.At the same time,due to the square root form,this method can guarantee the filtering does not diverge under the multi-axis train dynamics system.And train dynamics model is established based on MATLAB/Simulink to verify.Furthermore,in order to ensure the slip velocity estimation accuracy of multi-axle load train and minimize the communication consumption between filters,a slip velocity estimation method of multi-axle load train based on Distributed Square Root Cubature Information Filter was proposed.This method uses the information form of four one-axle filters to achieve the estimation accuracy only second to the centralized filter through the principle of average consistency,and can ensure that the filtering algorithm does not divergence,finally the algorithm is verified based on MATLAB.Next,the adhesion characteristic curve and train dynamics model may have certain uncertainties in actual operation and the influence of unknown noise of sensors,the slip velocity estimation method based on kalman filter framework needs relatively accurate adhesion model.So first run data design based on heavy haul train the gelling parameter identification method based on differential evolution algorithm,the characteristic of this method is able to stick together with a certain uncertainty given parameter identification result of the error as small as possible,and to train dynamics model and calculation model for the adhesion of tiny uncertainty,The slip velocity estimation method based on noise adaptive Square Root Cubature Kalman Filter is designed,and the simulation is verified by MATLAB/Simulink.Finally,in order to achieve the maximum adhesive coefficient of search,used to estimate the creep coefficient of speed observer to estimate and gelling establish gelling characteristic curve,was designed based on maximum power point tracking variable step search method,and through as a torque controller to realize PID control for search algorithm for a given slip velocity and creep rate of tracking.Finally,the algorithm is verified in a hardware-in-the-loop simulation environment based on MT PXI simulator.