Model Predictive Path Integral Control for Multi-vehicle Formation System with Log-Normal Mixture Noise

A log-normal mixture noise-enhanced model predictive path integral control (Log-MPPI) scheme is proposed for the cooperative formation control of multi-vehicle formation system. First, a vehicle kinematic model based on the leader-follower framework is constructed, and a cost function is designed that integrates trajectory tracking error, obstacle avoidance constraints, and input limitations. Second, building upon the traditional MPPI algorithm, log-normal mixture noise is introduced to improve sampling efficiency and enhance exploration capability under asymmetric control constraints. Third, the Log-MPPI controller was designed to achieve smooth trajectory tracking and dynamic formation transformation in the multi-vehicle system. Finally, the effectiveness of the proposed method was verified through numerical simulation experiments.