Dynamics Modeling for Modular Distributed-Drive Vehicles With Unknown Parameters

Modular distributed-drive (MDD) vehicles are an optimal solution for transporting oversized and overweight cargo. This study develops an extensible three-degree-of-freedom dynamics modeling approach for MDD vehicles. To calibrate unknown dynamics parameters, a novel Markov-based adaptive particle swarm optimization with hierarchical learning (MAPSO-HL) is designed. A six-axle MDD vehicle serves as a case study. The MDD dynamics model is calibrated across two driving scenarios, two adhesion conditions, and three load states using eight optimizers, resulting in 96 calibrations. The results indicate that the MAPSO-HL performs the best, with average calibration improvements of 54.636% and 40.78% over standard particle swarm optimization (PSO) and adaptive particle swarm optimization (APSO), respectively. Meanwhile, the calibrated dynamics model accurately characterizes the kinematic acceleration of the MDD vehicle, highlighting the effectiveness of the extensible dynamics model and its parameter calibrator. This research contributes a unified dynamics modeling approach for the MDD vehicles, which is beneficial for engineering applications.