An Economic Model Predicive Control Approach for Adaptive Cruise Control Design in Full Electric Vehicles

In the ever-evolving scenario of present days, the road transport industry is in the middle of a period of great changes. The urgent need for more energy-efficient transport has led to significant advancements in the introduction of vehicles powered by cleaner energy forms than fossil fuels: the Electric Vehicles (EVs). Simultaneously, road safety remains a critical concern, where the human factor represents the primary cause of traffic accidents. So, the increasing demand for safer solutions has driven the automotive industry toward the development of autonomous driving vehicles and driving assistance systems. Autonomous driving, combined with electric propulsion, represents a promising future solution to these challenges by reducing road accidents and optimizing energy consumption. However, the introduction of electric autonomous vehicles leads to new technical challenges, particularly in terms of efficient energy management and precise vehicle control. This thesis specifically addresses these challenges by focusing on the development of a Nonlinear Model Predictive Control (NMPC) strategy with the aim of improving both driving safety and energy efficiency. Traditional controllers like PID and LQR have inherent limitations: the PID is not an optimal control strategy, while the LQR, despite being optimal, is limited by its linear nature, making it unsuitable for handling nonlinear dynamics and constraints on states and inputs. In contrast, NMPC predicts and optimizes future control actions while explicitly considering physical and safety constraints, making it particularly well-suited for autonomous driving applications. In order to obtain an NMPC controller able to predict the behavior of the vehicle and suggest the best control sequence, such that a safe and energy-efficient scenario is achieved, a detailed vehicle model is developed by incorporating the vehicle longitudinal dynamics and the electric powertrain characteristics. Other than system model, the NMPC controllers also need to considers accurate real-world constraints, such as battery SOC limits, motor power capabilities, and maximum allowable speed, that allow the NMPC to ensure feasibility and efficiency in real-time applications. In this thesis, NMPC has been chosen as suitable control tool for designing an optimal Adaptive Cruise Control. The NMPC- based ACC performances have been compared with respect to traditionally used controller, the Constant-Time-Gap (CTG) controller, producing good results in terms of both driving safety and energy efficiency. The aim of this work is to contribute to the development of control strategies for future electric autonomous vehicles, in particular by focusing on the energy saving, that may be seen as one of the main concerns that limits the shift to electric vehicles, while not losing sight of safety, that remains a fundamental objective.