Energy-Efficient Adaptive Cruise Control: An Economic MPC Framework Based on Constant Time Gap

In recent years, the rising concerns about climate change have increasingly driven automotive companies to invest in electric vehicle (EV) development. While EVs stand as a promising solution to enable a more sustainable form of transportation, they still present inherent limitations compared to traditional ones, particularly in terms of driving range. To address these challenges, this thesis develops a novel nonlinear Economic Model Predictive Control (EMPC) strategy based on a Constant Time Gap (CTG) approach for adaptive cruise control, aiming to simultaneously achieve optimal control performance and energy efficiency. The main challenge in designing the EMPC control strategy lies in managing multiple concurrent control objectives. This thesis approaches such a challenge by employing a carefully structured cost function for the optimal control problem, encompassing various aspects among which the temporal evolution of the battery state of charge, the ego vehicle energy consumption, and the ahead vehicle behavior. These objectives typically concur with each other: aggressive following behavior might lead to higher energy consumption and faster battery depletion, while strict energy saving could compromise the cruise control performance. Using experimental data collected from a real vehicle (a ``Fiat 500e''), different cost functions are designed and compared to evaluate their effectiveness in carrying out the cruse control task while preserving energy efficiency. The main goal is to find an optimal balance between these two crucial metrics. %The approach uses a simplified two-axle vehicle model, focusing on basic longitudinal dynamics and battery behavior. This choice allowed me to focus on the development of the control system, reflecting my academic background in mechatronics. The cost functions are tested in increasingly complex scenarios. First, validation is carried out considering constant and sinusoidal velocity profiles. Then, standard driving cycles -- based on real-world vehicle operation -- are employed, among which the WLTP class 3 cycle. Simulation results show that the CTG-based EMPC strategy achieves better energy efficiency that traditional MPC approaches while proficiently attaining the cruise control task. The research work carried out in this thesis demonstrates how CTG principles can be successfully integrated within economic objectives for EVs control. While the immediate focus of this study are EVs, the general formulation of the EMPC framework makes it adaptable to several other power management scenarios. The cost function structure, which balances energy consumption and control performance, could find potential application in other domains, like different hybrid powertrain configurations (e.g., fuel cell hybrid electric vehicles), smart grids, and HVAC systems. Each of these scenarios presents similar trade-offs between resource utilization and system performance, suggesting that the proposed EMPC approach could have a wider application beyond EVs.