Model Based Road Slope and Mass Estimation of Road Vehicles

Abstract

Performance, fuel economy, driving experience, and safety are the most critical attributes automotive original equipment manufacturers (OEMs) focus on meeting customer requirements. OEMs are coming up with many features content to meet the customer requirements while meeting the stringent regulations regarding emissions and safety. Advanced sensors and embedded controllers are employed to meet the above needs. Performance of these algorithms can be significantly improved if we accurately predict a few important vehicle operational parameters. Vehicle mass and road slope are two crucial parameters essential for developing and optimizing control algorithms of powertrain, brake, and advanced driver assistance systems (ADAS). This study focuses on model-based road slope and vehicle mass estimation by integrating a state estimation algorithm and a longitudinal dynamics model. Drive-train torque, vehicle velocity, and vehicle design parameters are inputs for developing a longitudinal dynamics model. Initial approximate estimates of road slope and vehicle mass are given as initialization values for the state estimation algorithm, and future estimates are obtained using recursive adaptive filtering. A comparative study was performed between Extended Kalman filter (EKF) and Recursive Least Square (RLS) estimation algorithms for the convergence rate of state variables and error with respect to the actual state.