Personalized Dynamic Model for a Car-Like Vehicle "Robucar" Used in Localization

Vehicle dynamic modeling is involved in many research domains and oriented according to the need and application type. This paper describes the adopted approach to develop a personalized dynamic model for an autonomous mobile robot called "Robucar", which is a non-holonomic electric vehicle with four drive and steer able wheels. The aim is to determinate the dynamic equations of the forces and torques acting on the Robucar, using a tire-road contact model. The developed dynamic model considers three movements of a Robucar: longitudinal, lateral and yaw, taking into account the technical characteristics related to the electric traction engines and also to the vehicle geometry and its bi-steering system. The proposed model can be used in several fields as motion planning, localisation, control and study of the Robucar stability. In this paper, we exploited the proposed dynamic model for the localisation. Simulation and experimental results are conducted and a comparison with paths obtained by a kinematic model is made. The accuracy provided by the proposed dynamic model is clearly highlighted.