Research on Vehicle and Longitudinal Tire Model Parameters Identification Based on Electro-Mechanical Brake System

The Electro-Mechanical Brake (EMB) system is a novel type of brake by wire systems with independently controllable characteristics. This system aids in the decoupling analysis of the vehicle and actuator dynamics, thereby improving the accuracy of parameter identification. Therefore, this paper proposes an innovative parameter identification method for vehicle parameters and longitudinal tire model parameters, based on the characteristics of the EMB system and onboard sensors. First, based on the wind resistance and rolling resistance coefficients obtained from the vehicle coasting conditions, a decoupled constant clamping force sequence braking condition for the front and rear axles is designed by integrating the characteristics of the EMB actuator and vehicle dynamics. This approach enables the identification of vehicle and nonlinear longitudinal tire model parameters, significantly improving the accuracy of parameter identification. Next, considering the nonlinear characteristics of the longitudinal tire model, a factorial experiment is conducted to analyze the impact of the Particle Swarm Optimization (PSO) optimization algorithm parameters on the identification process from three perspectives: iteration count, computation time, and optimal function value. Furthermore, the effectiveness of three PSO variants: the Compressed Factor PSO (CF-PSO), the Adaptive Weight PSO (AW-PSO), and the Hybrid PSO (H-PSO), was investigated for identifying the nonlinear characteristics of the longitudinal tire model. Finally, through data simulation and real-vehicle experiments on both high-adhesion and low-adhesion roads, the effectiveness and accuracy of the proposed vehicle parameter and longitudinal tire model parameter identification method based on EMB system characteristics are verified through a comprehensive evaluation of multiple indicators, and the method’s validity is further confirmed using data backfill and model benchmarking.