A Game Theoretical Chassis Domain Approach to Trajectory Tracking for Automated Vehicles

Trajectory tracking which is performed by multiple chassis actuators remains a significant-yet-challenging issue in automated vehicles. The high nonlinearity of chassis dynamics and multi-actuator integrated control underlying this task is intractable, and this paper aims to address this crucial issue. We propose a chassis domain method of trajectory tracking based on the game-theoretical paradigm for automated vehicles. Unlike most prevailing studies which mainly focused on lateral control for path tracking, the proposed approach develops a unified control architecture to provide optimal tracking commands to longitudinal and lateral dynamic motions. Global chassis execution with fully adjustable chassis actuators called chassis domain control in this paper has dramatic potential to handle complex mobility situations in an efficient and flexible manner. This work develops a game-theoretical chassis domain approach to track both longitudinal and lateral trajectory profiles. The approach defines all chassis actuators as a set of players with associated objectives. Then it formulates a game-theoretical optimal problem to optimize multiple objectives including longitudinal tracking, lateral tracking, and vehicle handling dynamics. Various experimental tests are carried out, and the results suggest the feasibility and effectiveness of our proposed chassis domain approach.