Mass and Grade Estimation of Commercial Vehicle using Kalman Filter

Abstract

In today’s world, everyone is in race to make efficiently performing vehicles. For this performance improvement, it is important to know real-time vehicle parameters. This thesis presents a method to achieve some of those parameters. In this thesis, the considered parameters for estimation are mass of vehicle and slope of the road on which vehicle is present. Extended Kalman Filter (EKF) is a nonlinear estimator which has been used to estimate mass and road slope simultaneously. Use of any additional sensors to collect this information will be costly. So, EKF is used with already existing sensors in the vehicle to achieve the desired result. For virtual testing, co-simulation was carried out between Simulink and AVL Cruise. The input data to Simulink model was supplied from AVL cruise. The vehicle model was also verified through this co-simulation, by suppling all inputs for known mass and slope of road to satisfy the mathematical governing equation. The mass estimated is more than 80% accurate, which is enough for performance enhancing applications on the vehicle. If the algorithm runs for longer time. It corrects itself in the background so that all estimation reach above 90% accuracy with time, but due to constraint of convergence time, the mass estimation needs to be locked within certain time limit.