| The traditional passive buffering device generates damping force through the throttle holes in the flow passage.The buffering force cannot be controlled in real time and can only be effectively buffered for specific impact.Once the impact conditions change,the passive buffer device can not achieve a good buffer effect.Magnetorheological energy absorber(MREA)uses the magnetic fluid as the control medium to generate continuous and controllable buffering force by taking advantage of its adjustable damping under the action of magnetic field.In this thesis,MREA under the impact load of drop hammer is taken as the research object.By combining the method of test and numerical simulation,the inverse dynamic model,excitation control unit,buffer control strategy and other key issues of the magnetorheological buffer under high speed impact are mainly studied.The specific research contents are as follows:(1)The controllable characteristics of MREA were studied according to the impact test results.A drop hammer impact test platform was built to test the MREA under different current excitation,and dynamic parameters such as buffer force and displacement were measured synchronously.By analyzing the dynamic characteristics of MREA,it is found that MREA can be controlled under fixed current excitation.However,the buffer force curve of MREA presents the shape of " mountain peak ",and further research on the control algorithm is needed to obtain the optimal buffer characteristics.(2)The inverse kinetic model of MREA is studied by using the impact test data.By analyzing the mechanical characteristics of MREA,it is found that it is difficult to accurately describe the dynamic characteristics of MREA by using the physical model and parameterized dynamic model,so the inverse dynamic model of MREA is established by using the neural network method.The inverse dynamics model of MREA was established by using BP neural network and GRNN neural network.The model took the buffer force,piston velocity and piston displacement as the input and the excitation current as the output.Aiming at the easy local optimization of BP network,the genetic algorithm was used to find the optimal weight of BP,and GA-BP inverse model was established.Statistical index was used to analyze the errors of the three inverse models,and the GRNN inverse model had the highest accuracy.(3)Based on the neural network inverse model,the preset buffer control method of MREA is studied.By analyzing the buffering demand of falling hammer impact,the best buffering characteristic is the "platform effect",that is,the buffering force remains unchanged in the whole displacement.According to the motion equation of MREA and the initial impact velocity,the buffer force to be maintained and the corresponding motion curve are calculated.Using the neural network inverse model to calculate the required excitation current,and stored in the controller’s memory.The high inductance of the MREA coil can seriously affect the current response.The excitation current control unit of MREA is designed by combining the super capacitor with the Buck converter with fast charge and discharge speed to improve the current response speed.(4)The impact control test of falling hammer was carried out.The drop hammer impact control test system is built,and the preset control software is designed.The proximity switch is used to trigger the excitation control unit and output a preset current curve to buffer the MREA.The buffer effect of the preset current control is verified by experiment and simulation. |
