Research On Control Of Laterally Interconnected Air Suspension Based On Multi-agent Theory

Air suspension has been widely used in modern automobile because of its excellent vibration damping performance and car body torsion elimination.However,the laterally interconnected structure can significantly reduce the ability of the vehicle to resist the roll,and seriously affect the handling stability of the vehicle.Therefore,the opening and closing control of the interconnected states need to be made according to the different working conditions.And the damping control has not previously considered the change of the interconnection state,resulting in a decrease in its control effect.In this paper,the interconnected state of the laterally interconnected air suspension and the damping control of the damper were studied based on Multi-agent and other control theories.The structure and working principle of interconnected air suspension were briefly introduced.According to the correlative theories,the model of interconnected air springs,a 7-DOF vehicle model and a road excitement model were built.Based on the MTS-320 four channel road simulation stand,the laterally interconnected air suspension vehicle test bench was built,and the vehicle simulation model of the laterally interconnected air suspension was verified.Based on structures of the existing vehicle control system and according to the control demand of laterally interconnected air suspension,a distributed laterally interconnected air suspension control system was designed,and the control process of the system was expounded.Meanwhile,for the orderly development of the following parts,the fuzzy control theory,neural network fuzzy control theory,multi-agent theory,genetic algorithm and cooperative game Shapley value method were introduced in this paper.The influence of interconnected state change on body roll was simulated and analyzed.On this basis,the interconnected state agent control system was designed by using the knowledge of agent.And the pavement adjustment factor is proposed,which made the control system can deduce the optimal interconnected state under different pavement conditions.At the same time,based on the fuzzy control theory,the D level road was used as the datum to design the interconnected state reasoning device in the system.The PWM technology was used to modulate the output signal of the interconnected state reasoning device,and the control of the solenoid valves in the interconnected pipelines was realized.The multi-agent damper control system was constructed based on the multi-agent theory.In this control system,the information publishing agent was constructed by using the reactive agent model for the purpose of realizing the pertinent information of the information.In order to reduce the vibration of the sprung mass,the BDI model was used to model the ride comfort agent.In order to improve the anti-roll ability of the whole vehicle,a hybrid agent model was used to build the handling stability agent.In order to coordinate the demand for the damping coefficient of the ride comfort agent and the handling stability agent,a cooperative game agent was established based on the cooperative game Shapley value principle.The simulation results showed that,compared with the traditional laterally interconnected air suspension,the multi-agent laterally interconnected air suspension can effectively promote ride comfort and handling stability of the vehicle.