Abstract: Front/rear axle cornering stiffness coefficient is an important parameter in automobile dynamic control system, which is linked closely with automobile maneuverability. In the real world front/rear axle linear cornering stiffness often changes with the tire properties. To overcome the deficiencies of the traditional method in the estimation of linear cornering stiffness for wide-spread application, a method based on adaptive forgetting-factor recursive least square(FFRLS) is proposed to estimate this parameter in which only common sensors in modern automobiles are used. A lateral velocity estimation method with the robustness of tire properties is given, which is further used to calculate axle sideslip angle. Maximum lateral force is proposed and applied in the computation of axle lateral force. Some adequate toggle conditions are set to identify front/rear axle linear cornering stiffness. Simulation results based on veDYNA show that adaptive FFRLS has better parameter convergence performance than normal FFRLS when tire properties suddenly change. Finally the estimation method is validated when the toggle condition is limited via real-time simulation in the human-in-the-loop test bench.

Key words: Cornering stiffness, Human-in-the-loop hybrid simulation, Parameter estimation, Recursive least square