Game Control of Multi-agent Damper System for Laterally Interconnected Air Suspension

• Published in: 淡江理工學刊 Vol. 21; no. 2; pp. 241 - 252
• Main Authors: Zhong-Xing Li, Jian-Yu Huang, Hong Jiang, Hong-Tao Xue
• Format: Journal Article
• Language: English
• Published: 淡江大學 01.01.2018
• Subjects: Multi-agent; Fuzzy Neural Network; Laterally Interconnected Air Suspension (LIAS); Cooperative Game; Genetic Algorithm
• ISSN: 2708-9967
• DOI: 10.6180/jase.201806_21(2).0012

To improve the performance of vehicles equipped with laterally interconnected air suspension (LIAS), a damper system is constructed based on multi-agent theory and Shapley value principle of cooperative game. The multi-agent control system consists of an information fusion agent, a ride comfort agent, a handling stability agent and a game coordination agent. The information fusion agent obtains vehicle status information from environment, and carries out the information transmission according to the requirements of other agents; on the basis of suspension dynamic deflection and its changing rate, the ride comfort agent indicates the intention of damping coefficients. The handling stability agent triggers the reasoning module according to the current interconnection state information, and indicates the intention of the damping coefficients by the vehicle roll angle, in which, the reasoning module is formed by the self-learning of fuzzy neural networks according to the damping coefficients optimized by genetic algorithm. The game coordination agent receives the intention of the damping coefficients from the above agents, then corrects the damping intention through the rules of cooperative game, and outputs the global optimum damping coefficients. Based on a bench test, a simulation model is established to perform the effectiveness of the multi-agent damper system proposed in this paper. The results have verified that the proposed multi-agent damper system not only improves ride comfort of vehicle with LIAS, but also restrains the roll motion of the vehicle body.