A Study on Vehicle Cornering Stiffness Estimation Based on Recursive Least Squares Method

This paper investigates a vehicle cornering stiffness estimation method based on recursive least squares (RLS), aiming to enhance automotive active control capabilities. Conventional parameter identification schemes reliant on onboard sensors are constrained by sensing range and cost limitations. This study explores an estimation algorithm utilizing generic motion signals, derives the theoretical relationship between yaw rate response and yaw stiffness, and designs an RLS approach for online identification of relevant model parameters. In a simulation environment, applying an ideal steering angle excitation demonstrated that the RLS algorithm converges within approximately one second and successfully estimates the cornering stiffness with an error of less than 5% relative to the set true value. Results indicate that under known vehicle basic parameters and ideal input conditions, the proposed algorithm can effectively complete cornering stiffness estimation. This work establishes an algorithmic foundation for motion-signal-based vehicle parameter identification and analyzes challenges such as input dependency and parameter sensitivity encountered in practical applications.