Enhanced Force Estimation for Electromechanical Brake Actuators in Transportation Vehicles

Abstract

In transportation vehicles, the installation of force sensors for electromechanical brake systems causes cost issues and implementation difficulties. The elimination of the sensor is highly demanded. The main vulnerability of the existing force estimators is that the algorithms either rely on shifting characteristic curves or empirical data to determine the major quantities. Hence, they are less capable of dealing with initialization offset errors, fast-changing dynamics, and uncertainties. This article presents a novel super-twisting observer to accurately estimate the clamping force on the basis of a dynamic torque-balance model. In this observer, an extended state is defined on top of a super-twisting algorithm. The observer gains are computed through a smooth Lyapunov function to guarantee the fast convergence. It is experimentally validated that the observer enhances the performance required by the antilock braking systems in comparison with conventional methods.