PhD thesis of BUCKENMEYER André

Summary : 

This project aims to develop a digital twin of vehicle dynamics to enhance safety and comfort functions,as well as fatigue design. The objective is to have real-time access to precise information about the dynamic state of the vehicle and its ability to anticipate road profiles. To achieve this, we are building a digital model from a complete multibody model, enriched with data from various sensors such as accelerometers, a gyroscope, GPS, and a camera. This reduced model allows for the reconstruction of information not directly measurable and refines data for integration into vehicle control and piloting systems.

However, the development of these reduced models must overcome significant technological constraints, particularly for real-time assimilation of sensor data on a complete multibody model. The approach focuses on two main areas : applying model reduction techniques from the complete multibody model and using synthesis techniques to create substitute models, employing Machine Learning tools while adhering to the constraints imposed by dynamic equations.

Supervision :

Under supervision of Pr. Jean-Luc DION, MCF HDR Franck RENAUD and M. Laurent ROTA (Stellantis)