A Study on Parameter Optimization of a Vehicle Dynamics Model Based on Real-Vehicle Lateral Response Shin Ho Yeong Advisor : Prof. Sung Woo Suk, Ph.D. Department of Smart Vehicle System Engineering Graduate School of Chosun University This study propo...
A Study on Parameter Optimization of a Vehicle Dynamics Model Based on Real-Vehicle Lateral Response Shin Ho Yeong Advisor : Prof. Sung Woo Suk, Ph.D. Department of Smart Vehicle System Engineering Graduate School of Chosun University This study proposes a parameter optimization procedure for a vehicle dynamics model using real-vehicle lateral response data. To improve the reliability of simulation-based autonomous driving controller verification, real-vehicle speed and steering angle profiles were applied to the simulation model, and yaw rate and roll angle were used as the main response indicators. candidate parameters related to the steering system, tires, vehicle body, and suspension were selected based on vehicle dynamics knowledge. A one-at-a-time sensitivity analysis was performed to identify influential parameters, and a genetic algorithm was applied to minimize the response error between the real vehicle and the simulation model. The optimization was performed using 39 training cases. The objective function decreased from 0.3626 to 0.1256, corresponding to a reduction of approximately 65.35%. Validation using nine cases not included in the optimization process showed that the optimized model reproduced the overall trends of yaw rate and roll angle responses under unseen input conditions. These results indicate that the proposed procedure improves the response agreement between real-vehicle data and simulation results and can support more reliable simulation-based verification of autonomous driving controllers.