Actuator-to-Vehicle Joint Estimation of Clamping Force for Sensorless Control of the Electro-Mechanical Brake System

The electro-mechanical brake (EMB) is a promising brake actuating system for electrified vehicle. To enhance the system function safety while saving space from redundancy sensors, this paper studied sensorless climbing force control for the EMB where a new climbing force estimator is proposed by fusing the information from vehicle dynamics and EMB states. The work was done with three contributions: 1) The priori clamping force characteristics were implemented to build the estimator with two parallel models, one of which was derived from the actuator rigid-body dynamics while the other was derived from vehicle longitudinal dynamics model; 2) a proportional-integral (PI) observer utilizing wheel speed residual signals was developed to correct the initial estimates iteratively; 3) a fuzzy control controller was proposed to optimize the key parameters of the PI observer. Comparative study was conducted on a co-simulation platform and the results showed that the actuator-to-vehicle joint estimation method can reduce more than 28% root mean square error (RMSE) compared with the conventional actuator model-based estimation method by utilizing PI observer. After the optimization of key parameters, the optimal ratio can reach 32%. Robustness analysis demonstrated that the climbing force estimations accuracy across the studied distinct braking scenarios were consistency.