Integrated Estimation Strategy of Brake Force Cooperated with Artificial Neural Network Based Road Condition Classifier and Vehicle Mass Identification Using Static Suspension Deflections

Abstract

Brake forces and maximum static road friction coefficients for each wheel of the vehicle are essential information for vehicle safety systems including adaptive cruise control, electronic stability control (ESC), and collision avoidance system, etc. Many studies have been performed to estimate brake force and road friction using well-known model-based approaches, but none have unambiguously guaranteed an accurate performance in all ranges of driving conditions and road ones. In addition, the investigation of the integrated estimation approach of road friction and brake force including mass estimation has not been clearly addressed so far. Therefore, in this study, a novel integrated estimation strategy based on a data-driven technique and artificial neural network (ANN) classifier along with a compact mass identification has been proposed to acquire the accurate road friction and brake force of individual wheel. Specifically, it includes an instant mass estimation by monitoring static suspension deflections, an artificial neural network (ANN) classifier for road friction coefficient based on the average data set from available standard sensors, and a brake force estimation using the data-driven technique. The performance of the proposed technique is validated by a co-simulation environment between Carsim and MATLAB/Simulink. It is found that the integrated estimation strategy guaranteed an accurate estimation of brake forces and road friction for a wide range of variations of road frictions, vehicle velocities, and masses. This work will be a valuable asset for those who wish to develop an integrated estimation system for such crucial parameters of the vehicle system.