Abstract: To alleviate the contradiction between ride comfort and handling stability of laterally horizontally interconnected air suspension vehicles, an intelligent system for interconnecting state control agent system of laterally horizontally interconnected air suspension is constructed based on the agent theory. Firstly, a horizontally interconnected air suspension vehicle model is established and its accuracy is verified through experiments. Then, based on the structure of traditional BDI (belief-desire-intention) agent, Thompson sampling is adopted to build the imitated skyhook interconnecting state control agent with online self-learning ability. The environmental state information is sensed from the sensor information acquisition module firstly, and then self-learning is performed through the internal BDI reasoning process and learning behavior, and finally the key parameters of imitated skyhook interconnecting state control adapted to different environmental states are output (hysteresis range to the interconnected state control module). The simulation is carried out under the mixed condition and the results show that the system balances the ride comfort and handling stability of vehicles

Key words: horizontal interconnection, air suspension, agent, BDI model, Thompson sampling