Modellazione integrata e strategie di gestione termica per celle a combustibile PEM con applicazioni ai Veicoli Pesanti

The objectives of this research are focused on the modeling, analysis, and control of the fuel cell at different levels of detail: from the thermal management of the whole vehicle to the modeling and control of fuel cell auxiliary system and, finally, to the analysis of fuel cell performance as a stand-alone component. In addition, several results are established to ensure the stability of a fuel-cell subsystem, along with tools for possible controller improvements. The proposed approach provides a mathematical framework that enables the inclusion of system non-idealities, such as quantization and sampling. Chapter 2 presents the fuel cell auxiliary systems and their integration within the vehicle, along with the corresponding mathematical models. Chapter 3 addresses the thermal control problem by introducing a comprehensive vehicle thermal model, including battery, electric motor, cabin, and fuel cell loops. An integrated configuration is proposed to exploit fuel cell waste heat for cabin heating, improving overall energy efficiency. HVAC system with refrigerant R134a is modeled through stationary equations and the transition phase inside condenser and evaporator are considered. Radiators, condenser, and evaporator are modeled as heat exchangers with ε -NTU method, while lumped parameters thermal models are used for electric motors, battery and fuel cell. Moving beyond heuristic or rule-based thermal management strategies, this work adopts a model-based approach based on a Nonlinear Model Predictive Control (NMPC). Chapter 4 develops a feedback control design methodology for the stabilization of nonlinear systems. The first section focuses on the oxygen supply control problem, addressing oxygen starvation avoidance while accounting for sampling and quantization effects typical of digital implementations. A static state-feedback, quantized, event-based sampled-data controller is proposed for a class of nonlinear systems, and semi-global practical stability is guaranteed under sufficiently fast sampling and accurate quantization through the stabilization in the sample-and-hold sense theory. Such methodology is then applied to the analyzed fuel cell subsystem. Since the above approach requires full state availability, the second section extends the results to output-feedback control of time-varying nonlinear systems not necessarily affine in the control input. A quantized, sampled-data, observer-based event-triggered control strategy is developed based on the Dynamic Output Steepest Descent Feedback (DOSDF) framework. The proposed methodology ensures semi-global practical stability under suitable sampling and quantization conditions through the stabilization in the sample-and-hold sense theory. The proposed results are validated using Sontag’s universal formula and demonstrated on several nonlinear systems, including an inverted pendulum, a glucose–insulin system, and a class of time-varying nonlinear systems that are non-affine in the control input. Chapter 5 investigates how electrochemical and fluid-dynamical processes under operating conditions affect its behavior. The study is carried out in the frequency domain through physics-based impedance spectroscopy. Electrochemical Impedance Spectroscopy is emulated via numerical simulations and applied to a pseudo-1D model. By applying a square-wave current excitation, the influence of varying operating conditions and the associated internal electrochemical and transport phenomena is examined, with particular emphasis on the role of water in fuel cell dynamics. Two models are compared: one accounting for water condensation and production within the membrane, and another in which water is present only in vapor phase and generated at the catalyst layer (CL). These modeling differences are sufficient to highlight critical operating conditions for the fuel cell and provide insight into potential avenues for further model refinement and performance improvement.