Precise Calibration of Steering Feel Based on Model Identification

In the development of Electric Power Steering (EPS), calibration is the final step and is crucial for a good steering feel. Previous studies focus on design optimization of the mechanical system and control strategy of the EPS in the early phase of the development process, rarely addressing on-site calibration with physical prototypes. In this paper an EPS calibration method is proposed to meet strict objective targets of steering feel. The method is based on a theoretical model established between EPS assist parameters and steering torque characteristics. Parameters of the theoretical model depend on the electro-mechanical property of each vehicle, but they are identified with field test data during the calibration process instead of being calculated. The identified model is then used to obtain a new set of assist parameters that meet the steering feel targets. It is shown with test data that strict objective steering feel targets can be fulfilled with no more than two field tests, thus improving calibration efficiency without compromising accuracy.