A fusion estimation method of clamping force for electro-mechanical brake system based on adaptive Kalman filtering

Abstract

Clamping force estimation can reduce cost and space design complexity of electro-mechanical brake (EMB) system without a sensor. However, existing estimation methods lack the ability to adapt to environmental change such as temperature, humidity, and noise, which result in low estimation accuracy. To address this challenge, a state estimator based on adaptive Kalman filter (AKF) is first proposed. By setting the residuals sliding window, the noise covariance matrix is adapted to reduce the interference of noise on state estimation. Second, an estimation compensation fusion mechanism is proposed. The clamping force compensation value is calculated by building a vehicle dynamic prediction model to reduce the disturbance caused by variations of EMB parameters. Finally, a comparison is made with the existing estimation method to demonstrate the superiority of the proposed method.