Property Experiment And Mechanical Model Optimization Of Magnetorheological Damper

In recent years,magnetorheological dampers assemi-active control devices having excellent propertyhave been used in cable-stayed bridges.But damping control technology of the magnetorheological dampers is far from mature,there are many problems to be solved.In the actual vibration process,the external excitation conditions of oblique cable in the actual vibration process are changing constantly,but many of the existing dynamic models of magnetorheological damper are obtained by parameter identification under a single excitation.Without considering the influence of changing excitation conditions,the prediction accuracy is not high enough.The actual application of the results are not ideal.Therefore,it is necessary to study the effect of excitation on the magnetorheological damper,and to establish a more accurate dynamic model of MR damper.This paper has done a lot of research on this problem:(1)In this paper,MR-60 magnetorheological damper made by ourselfe is used to do the experriments under low frequency and high amplitude excitation.The main research of the two aspects:(1)current's lnfluent on the dynamic property of magnetorheological damper.(2)Effect of excitation properties on the dynamic property of Magnetorheological Damper.From the experimental results,it is found that as the current increases,the output damping force gradually increases and reaches saturation at 1.25 A.The excitation amplitude and the excitation frequency have obvious influence on the mechanical properties of the magnetorheological damper.These effects are mainly manifested in the hysteresis part.The influence of frequency and amplitude on the mechanical properties of MR damper can be simplified to study the effect of excitation velocity amplitude on the mechanical properties of magnetorheological damper.(2)Based on the dynamic experiment of magnetorheological damper,the parametric models of magnetorheological damper considering the effect of excitation were establishedby MATLAB software.In this paper,the hyperbolic tangent model and the extended hyperbolic tangent model were identified and simulated by SIMULINK toolbox.From the simulation results,it can be seen that the hyperbolic tangent model does not consider the effect of the stimulus.When the excitation condition changes,forecast results were not ideal.The extended hyperbolic tangent model considers the effect of the excitation property.The results show that the genetic algorithm can not accurately and quickly identify the above parameters models.The paper proposes a genetic algorithm to combine the least squares algorithm to identify the parameters of the model.From the results,the method can improve the recognition accuracy,shorten the recognition time.(3)Aiming at the problem that the parameter model has low prediction accuracy and difficulty in parameter identification,BP neural network prediction model considering the effect of excitation is established by using BP neural network toolbox in MATLAB.The results show that the BP neural network prediction model can track the output damping force of the magnetorheological damper well,and the model does not have complicated calculation process,and it does not need carry out parameter identification.The modeling has some Engineering application value.(4)The hyperbolic tangent model,the extended hyperbolic tangent model,and the BP neural network prediction model are compared and analyzed from the aspects of model precision and model complexity.From the comparison results,it can be found that the error of the BP neural network prediction model is 3.777% on average,and the error of the hyperbolic tangent model is 16.4919%,and the average error of the hyperbolic tangent model is 29.6178%.In this paper,the BP neural network prediction model is established with the highest efficiency and the experimental requirements,which can provide some guidance for the application of the magnetron damper.