Research On Nonlinear Control Of Lateral Vibration Of High-speed Elevator Car System

To meet the needs of high-rise buildings,the elevator,as a necessary component of highrise buildings,is developing towards higher speed and higher comfort levels.Elevator speed increase causes a series of problems,one of the most critical problems is the deterioration of comfort caused by the elevator car’s vibration.Therefore,it is imperative to find an excellent control method to restrain elevator vibration.In this thesis,the nonlinear control method is studied to control the vibration of high-speed elevator car system by using active guide shoes.In order to establish the car system dynamic model conform with the engineering practice,the simplification method of vibration system models（including guide shoe model,guide rail excitation model and car horizontal vibration model）is studied.Based on the rigid body dynamics theory,the car transverse vibration dynamic model is established.The car system model is simulated by MATLAB,and the time domain and frequency domain simulation data of car horizontal vibration are obtained.The car vibration acceleration data of 4 m/s high-speed elevator is collected on the 120 m high-speed elevator test tower,and compared with the simulation data.The data analysis result shows that the high vibration amplitudes of the built model and the real elevator are concentrated in the range of 0 10 Hz,the relative errors of the mean value and root mean square value of the horizontal vibration acceleration of the car are 4.63% and 2.05% respectively,and the determination coefficient of the goodness of fit of the model is 0.7253 > 0.4.The results show that the dynamic model of the car system can reflect the vibration response characteristics of the car well.Aiming at the problems that it is difficult to obtain the model parameters of active guide shoes accurately and the excitation of guide rail unevenness,based on the proposed dynamic model of active guide shoe,the control method of sprung mass vibration of active guide shoe based on fuzzy control theory is studied.Based on the fuzzy control theory,the spring mass vibration acceleration and spring mass vibration velocity of the active guide shoe are taken as inputs,and the fuzzy control rules are divided into seven variable grades.At the same time,the fuzzy and defuzzification processing are carried out to form a fuzzy controller.Under the excitation of step and random unevenness,the simulation and comparative analysis of passive control guide shoe and fuzzy control active guide shoe are carried out.The results show that under the two kinds of guide rail excitation,the root mean square value（RMS）of vibration acceleration of sprung mass is reduced by 26.2% and 30.1% respectively.Therefore,the fuzzy control method can significantly suppress the vibration of sprung mass of active guide shoe.Further,in order to realize the vibration control of the car system without knowing the accurate physical parameters of the control model,according to the established car horizontal vibration dynamic model,an active disturbance rejection control method is proposed to control the car vibration of high-speed elevator.In order to simplify the difficulty of control parameter design,the disturbance observation results of nonlinear extended state observer and linear extended state observer are compared and analyzed through simulation.The results show that the linear extended state observer can achieve the same observation effect of nonlinear extended state observer,and take into account the advantages of simple and easy parameter adjustment.Then,it is verified that the dynamic stroke of the active guide shoe meets the stroke constraints of the mechanical structure of the guide shoe,and the car vibration acceleration under ADRC and passive control is simulated and compared under full-load and no-load.The results show that the root mean square value（RMS）of the car horizontal vibration acceleration under noload and full-load are reduced by 38.2% and 53.2% respectively,which proves the effectiveness and robustness of the ADRC.The car system model not only has the problem of unknown parameters,but also the nonlinear output of the actuator will weaken or even deteriorate the control effect of the active guide shoe.To overcome these issues,an adaptive control algorithm based on extended state observer（ESO）is proposed to improve the control accuracy of ARG and then suppress the car vibration better.Based on the established uncertainty elevator system model and non-ideal actuator mathematical model,the proposed control algorithm divides system uncertainties and non-ideal factors into structured uncertainties（i.e.,payload mass and component parameters）and unstructured uncertainties（i.e.,actuator nonlinearities,external disturbances,unmodeled dynamics）.Adaptive control laws and ESO are designed to compensate for the structured uncertainty and the unstructured uncertainty,respectively,and integrated via a feedback law.The stability and convergence of the novel control strategy are proved by the Lyapunov stability theory.Finally,Several comparative experimental and simulation results demonstrate that the proposed control strategy can greatly improve ARG control performance.The comparison between the experimental data of the elevator and the simulation data under the random excitation of the guide rail shows that the RMS value of the horizontal vibration acceleration without control is reduced by 59.9%,and the control effect is remarkable.