Online estimation of the road grade andthe gross weight of an electric drive shuttlebased on longitudinal vehicle dynamics

The vehicle mass and road grade have significant influences on the longitudinaldynamics of commercial vehicles for passenger transportation, such as buses andshuttles. Therefore, estimating these factors plays an important role in both controllingthe vehicle’s motion and estimating its total energy consumption. Theroad grade varies as the vehicle moves, and the gross weight of a shuttle changesduring a trip due to the number of passengers on board, which requires an onlineestimation that can track these changes during the operation. The challenge ofmaking accurate online estimations has usually been overcome by placing additionalsensors on the vehicle and incorporating sensor fusion techniques.This study investigates the efficacy of conventional online parameter estimationapproaches, specifically Gradient Descent and Recursive Least Squares in estimatingthe road grade and vehicle mass without utilizing any sensors except theones typically included in standard commercial vehicles. A commercially availableelectric drive shuttle was converted into a test bed, and the performance of theproposed methods was evaluated under actual road conditions with varying grossweights. The results indicated that the estimation performance is highly sensitiveto model accuracy, warranting further study on identifying prominent factors thataffect the longitudinal dynamics of the vehicle.