Vibration Control of Light Rail Vehicle Active Suspension System Using Bargaining Game-Based Distributed Model Predictive Control

This paper proposes a distributed model predictive control (DMPC) scheme with preview disturbance based on bargaining game theory for the active suspension control of a light rail vehicle (LRV). The objective is to improve ride comfort and safety by minimizing the deflection of suspension system and the vertical acceleration of the car bodies. The LRV is modeled as a multi-body dynamical system consisting of two power carbodies and one trailer carbody. The proposed DMPC framework decomposes the vehicle into three dynamically interacting subsystems, each governed by a local MPC controller. The optimal control inputs are determined by formulating the control problem as a bargaining game, which is then solved by a negotiation algorithm. The proposed approach is evaluated through numerical simulations under sinusoidal track disturbances and compared with a centralized MPC scheme. The simulation results demonstrate that the proposed DMPC effectively attenuates vertical acceleration and suspension deflections, thereby enhancing ride comfort and operational safety while achieving superior overall control performance. These findings indicate that the proposed DMPC provides an effective and computationally attractive alternative to centralized MPC.