Study On Starting Torque Control Strategy Of Elevator Traction Machine Based On Model Predictive Control

There are many advantages of gearless permanent magnet synchronous traction machine,such as small size,high mechanical efficiency,low speed,high torque etc.so that more and more gearless permanent magnet synchronous traction machines are applied in roomless elevator systems.However,during the start of the elevator traction machine,the traction wheel will be affected Non-linear time-varying perturbation torque,there will be a problem of starting shock.During the zero-servo period of the traction machine,if the output torque of the traction machine can not balance these disturbance torques in real time,the traction wheel will rotate in the reverse direction and the car will reverse slip,which seriously affects the smoothness of elevator starting and the comfort of riding.Traditional traction machine control systems use load cells to obtain external load information in real time,but their signals are susceptible to electromagnetic interference and increase costs.Therefore,this paper will mainly study the starting torque control strategy of gearless permanent magnet synchronous traction machine for vertical elevator without load cell.Aiming at the problem that the load torque of the elevator traction machine is strongly disturbed during the starting process,the load torque characteristics of the traction machine during the starting process are analyzed.The disturbance torque simulation model is established,and under the action of ideal disturbance torque and actual disturbance torque,the PI traction control system of elevator traction machine with different loads is simulated and analyzed.The problems of the traditional PI vector control strategy in the starting torque control system of elevator traction machine subjected to nonlinear time-varying disturbance torque are pointed out.In order to reduce the degree of car and traction wheel slip,and improve the stability of the elevator starting,dual-loop predictive control strategy is proposed,in which the speed loop adopts single vector model predictive control and the current loop adopts deadbeat predictive control.The dual-loop predictive control system under different load torques is simulation analyzed to verify its superiority to the traditional PI control strategy in the elevator traction machine starting torque control system.In order to further improve the anti-interference ability of the traction machine control system,the predictive control strategy based on an expanded state observer is proposed.The load torque feedforward compensation based on the expanded state observer is introduced into the predictive controller to form a two-degree-of-freedom control system to reduce the influence of load torque disturbance and unmodeled dynamic model mismatch on the control system,and the stability and effectiveness of the load torque observer are verified through theoretical and simulation analysis,which further reduces the starting shock and improves the smoothness of elevator starting.Aiming at the problem that the inertia perturbation of the traction machine causes the mismatch of the predictive controller model,which affects the stability of elevator starting,a variable parameter predictive control strategy based on the identification of inertia is proposed.The influence of rotational inertia on the traction machine predictive control system is analyzed,and a variable parameter predictive control strategy based on the identification of rotational inertia is proposed.The MRAS is used to identify the rotational inertia of the traction machine online,and the identification value of the rotational inertia is introduced into the speed prediction controller and the expanded state observer to eliminate the parameters related to the rotational inertia of the control system,and to modify the controller model online.Through the simulation analysis of the variable parameter predictive control system with different load torque and different rotational inertia,it is verified that there is better control performance than the predictive control system with accurate rotational inertia parameters,and the overshoot of the speed output with the large rotational inertia is suppressed.