Research on Intelligent Vehicle Fleet Longitudinal Control Based on Fuzzy Model Predictive Control

Addressing the challenge of balancing tracking efficiency and stability in longitudinal cooperative control of intelligent vehicle platoons, this paper proposes a hierarchical control strategy integrating fuzzy logic with Model Predictive Control (MPC). The upper-layer controller, based on the platoon's longitudinal dynamics model, formulates an optimization objective function that considers both tracking performance and ride comfort. A fuzzy logic system is introduced to adaptively adjust the weighting coefficients within the MPC objective function in real-time based on the prevailing tracking state, outputting an optimal desired acceleration. The lower-layer controller, utilizing an inverse longitudinal dynamics model incorporating the engine's universal characteristic map and a brake system model, accurately translates the desired acceleration into throttle opening or brake pressure via a Proportional Integral Derivative (PID) control algorithm. To validate the control strategy, a co-simulation platform was built using CarSim and Simulink. Tests were conducted under high-speed cruising and emergency braking scenarios, comparing the proposed strategy against classical PID and conventional MPC methods. Results demonstrate that the proposed fuzzy MPC controller maintains all dynamic constraints within permissible limits, stabilizing the inter-vehicle spacing error within 8 meters. Compared to PID and conventional MPC control, the maximum velocity error is reduced by 6.6 m/s and 2.5 m/s, respectively. Particularly in emergency braking scenarios, velocity adjustments are smoother, significantly enhancing the platoon's driving stability and ride comfort.