Marotta, Raffaele (2024) Towards Autonomous Driving: Integrating Data-Driven Approaches into Model-Based Framework for AI-Enhanced Vehicle Dynamics. [Tesi di dottorato]
![[thumbnail of marotta_raffaele_37.pdf]](https://www.fedoa.unina.it/style/images/fileicons/other.png "marotta_raffaele_37.pdf") |
Documento PDF
marotta_raffaele_37.pdf Download (10MB) |
| Tipologia del documento: | Tesi di dottorato |
|---|---|
| Lingua: | English |
| Titolo: | Towards Autonomous Driving: Integrating Data-Driven Approaches into Model-Based Framework for AI-Enhanced Vehicle Dynamics |
| Autori: | Autore Email Marotta, Raffaele raffaele.marotta3@unina.it |
| Data: | 5 Dicembre 2024 |
| Numero di pagine: | 265 |
| Istituzione: | Università degli Studi di Napoli Federico II |
| Dipartimento: | Ingegneria Industriale |
| Dottorato: | Ingegneria industriale |
| Ciclo di dottorato: | 37 |
| Coordinatore del Corso di dottorato: | nome email Grassi, Michele michele.grassi@unina.it |
| Tutor: | nome email Strano, Salvatore [non definito] Terzo, Mario [non definito] |
| Data: | 5 Dicembre 2024 |
| Numero di pagine: | 265 |
| Parole chiave: | Autonomous Vehicles; Vehicle Dynamics; Model-Based Estimation; Artificial Intelligence; Machine Learning; Control Systems; Kalman Filter; Neural Networks; Tire-Road Interaction Force; Sideslip Angle; Unsprung Mass; Roll Dynamics; Pitch Dynamics; SAE Automation Levels; Adaptive Control; Reinforcement Learning; Transportation Safety; Efficiency; Passenger Comfort. |
| Settori scientifico-disciplinari del MIUR: | Area 09 - Ingegneria industriale e dell'informazione > ING-IND/13 - Meccanica applicata alle macchine |
| Informazioni aggiuntive: | Il sistema non mi dà la possibilità di selezionare il ciclo 37. Ho quindi selezionato il ciclo 36. Vi prego di correggere. |
| Depositato il: | 18 Nov 2025 14:49 |
| Ultima modifica: | 02 Set 2026 08:06 |
| URI: | https://www.fedoa.unina.it/id/eprint/16333 |
Abstract
The advancement of autonomous vehicle technology presents an opportunity to revolutionize transportation by enhancing safety, efficiency, and convenience. At the heart of this innovation lies the integration of model-based estimation and control techniques with artificial intelligence (AI) to address the complex challenges of vehicle dynamics. This thesis proposes novel hybrid methodologies to bridge the gap between these approaches, aiming to overcome the limitations of current systems and support the progression towards higher levels of driving automation as defined by the SAE classification. Vehicle dynamics—governing a vehicle's motion through the interplay of forces and moments—are critical to safety, efficiency, comfort, and reliability in autonomous systems. Model-based techniques, such as Kalman filters, Linear Quadratic Regulators (LQR), and Model Predictive Control (MPC), offer predictive capabilities, systematic design, and interpretability in dynamic estimation and control. Meanwhile, AI, particularly machine learning (ML), enhances these methods through data-driven insights, adaptive control strategies, and complex decision-making capabilities. This research focuses on integrating these complementary approaches to address key challenges in autonomous driving. It develops advanced solutions for critical tasks, including tire-road interaction force prediction, sideslip angle estimation, unsprung mass displacement estimation, and the stabilization of roll dynamics. By leveraging physical models alongside AI-driven enhancements, the proposed methodologies aim to improve performance across diverse operating conditions. The findings of this thesis are particularly relevant for achieving higher levels of driving automation, where precise and reliable dynamic control systems are indispensable. The integration of model-based and AI-driven techniques lays a robust foundation for ensuring stability, efficiency, and passenger comfort in future autonomous vehicles, thereby contributing to their safe and effective deployment.
Downloads
Downloads per month over past year
Actions (login required)
 |
Modifica documento |
