Master thesis presentation by Alexandre Vieira da Rocha and Daniel Alejandro Poveda Pi

Opponents: Logi Sigurdarson and Pierfrancesco Oselin
Examiners: Nikolce Murgovski to Alexandre and Mats Jonasson to Daniel
Supervisors: Dr. Derong Yang, Anand Ganesan, Juliette Torinsson, all at Volvo Cars

The thesis has been conducted at Volvo Cars.

Abstract:
Future electric vehicles (EV) are expected to exploit various sensors concerning environment perception and vehicle motion sensing, as well as different actuator configurations, where even more control degree of freedom is added due to the advent of autonomous driving. Nevertheless, EVs still suffer from a major limitation when compared with internal combustion vehicles: the driving range. Combining the autonomous capacity of vehicles with the need of energy-efficient vehicle motion control, energy-optimal trajectory planning urge as a promising technology to mitigate this problem. While energy-efficient longitudinal control technologies such as ECO-cruise control are already mature and widely implemented, combining lateral and longitudinal control is still being researched. In this thesis, nonlinear model predictive control (NMPC) is applied to optimize energy consumption of an EV given a reference path and longitudinal velocity profile, where focus is placed on efficient and fast computations for real-time implementation. Simulation results are presented with a high-fidelity vehicle model in a test track. The relation between lateral and longitudinal control and savings in energy consumption is described, in order to find out what is the energy consumption reduction potential from the control perspective. In specific, cornering scenarios are studied and the displacement from the center line of the path as well as the reference speed are related with energy savings, both individually and combined.

Welcome!

Logi, Pierfrancesco, Mats and Nikolce