Data driven modelling of fluid mechanical systems for fuel injection control

Stringent emission regulations, such as Euro 7, are driving the automotive industry to enhance the precision of their fuel management systems. Achieving accurate fuel injection is critical to meeting these standards, which requires advanced and computationally efficient modelling of injection dynami...

Descripción completa

Detalles Bibliográficos
Autor: Putri, Andrea Eka
Tipo de recurso: tesis de maestría
Fecha de publicación:2025
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/442173
Acceso en línea:https://hdl.handle.net/2117/442173
Access Level:acceso abierto
Palabra clave:Fluid mechanics
Fluid dynamics
Nonlinear theories
Mecànica de fluids
Dinàmica de fluids
Teories no lineals
Àrees temàtiques de la UPC::Enginyeria mecànica
Descripción
Sumario:Stringent emission regulations, such as Euro 7, are driving the automotive industry to enhance the precision of their fuel management systems. Achieving accurate fuel injection is critical to meeting these standards, which requires advanced and computationally efficient modelling of injection dynamics. This thesis focuses on simplifying the complex fuel injection process by developing a one-dimensional model of the fuel system that balances accuracy with computational efficiency. Building on this, the project applies Sparse Identification of Nonlinear Dynamics (SINDy) to extract reducedordermodels(ROMs)fromsimulationdata. TheseROMs,whichareessentially 0D, capture the essential fluid mechanical behavior of the injection system at a fraction of the computational cost, enabling faster simulations while retaining key dynamics. By combining physics-based modeling with data-driven methods, the work highlights the potential of ROMs for real-time control and system optimization. Thethesis was carried out at Scania in Södertälje, Sweden, and contributes to ongoing efforts to integrate accurate yet lightweight models into embedded automotive applications.