Research On Interconnecting State Control Of Laterally Interconnected Air Suspension Based On Agent Theory

As one of the derived structures of the traditional air suspension,the laterally interconnected air suspension has better vibration isolation performance and torsion elimination performance,which can further improve vehicle riding comfort.However,the anti-roll ability of the suspension will be decreased when the air springs are laterally interconnected,which seriously weaken the handling stability especially when turning.If the shortcomings of the laterally interconnected structure can be compensated by reasonable interconnecting state control,the advantages can be fully utilized.At present,there are few studies on the interconnecting state control of laterally interconnected air suspension.Based on the in-depth analysis of the research status at home and abroad,this paper studies the intelligent control of the interconnecting state of the laterally interconnected air suspension.Firstly,the structure characteristics and working principle of the laterally interconnected air suspension are explained,and the mathematical model of the laterally interconnected air spring is established based on the theory of hydrodynamics theory and engineering thermodynamics theory.Considering the bounce,roll and pitch modes of the spring mass and the vertical modes of the unspring mass,a 7-DOF vehicle dynamics model is built,which conforms to the non-independent suspension structure of existing test car.The air spring characteristic test was carried out to obtain the air spring parameters.The simulation model of the laterally interconnected air suspension with four-wheel random road surface excitation model was established in the MATLAB/Simulink environment with the car parameters.Secondly,the conventional air suspension on the existing test vehicle is modified with laterally interconnected structure.Relying on the test car,the information acquisition and control system is designed and built based on Arduino platform.By comparing the simulation results with the test results,the accuracy of the vehicle simulation model established above is verified.The principle and implementation of the imitated sky-hook interconnecting state control strategy are theoretically analyzed.The validity and control effect of the imitated sky-hook interconnecting state control strategy is verified for the first time on a real vehicle.The test results showed that the selection of hysteretic range has great influence on vehicle riding comfort and handling stability.Finally,agent theory is introduced into the research of the interconnecting state control.The structure and behavior of the BDI agent are described in detail.Based on the imitated sky-hook interconnecting state control strategy,the agent is designed and its perception,reasoning and decision-making behavior are emphatically designed,so as to establish the intelligent system for laterally interconnecting state control.In this system,the agent perceives the environmental state information and form the evaluation of vehicle riding comfort and handling stability.In order to enhance the adaptability under dynamic environment,the Thompson sampling reinforcement learning algorithm is introduced.The agent learns the reasonable action combination of upper and lower bounds of hysteretic range under different environmental states through the evaluation index.The corresponding model is built in MATLAB/Simulink environment and simulated under mixed working conditions.The simulation results show that the contradiction between ride comfort and handling stability of the laterally interconnected air suspension vehicle controlled by the interconnecting state control agent system has been effectively alleviated.