Development of PEM FCS model for Vehicle Digital Twin

Kurzfassung

Proton exchange membrane (PEM) fuel cells are promising clean energy technologies because of their high power density and efficiency. However, developing efficient PEM fuel cell systems requires overcoming the challenges of multidomain complexities. Computer modelling and simulation helps by playing a crucial role in accelerating PEM fuel cell design and performance optimisation before moving to physical prototyping. One such tool is OpenModelica which is an open source modelling and simulation platform that supports multiphysics system modelling using Modelica’s evolutionary and equation-based approach. This thesis presents the development of a Proton Exchange Membrane (PEM) fuel cell system model using the OpenModelica open-source modelling platform. Individual cell components such as the membrane, catalyst layers and flow fields are modeled using circuit analogies and transport equations. The parameterized component models are assembled to build a modular fuel cell stack unit. This stack is supported by a Balance of Plant (BoP) that contains auxiliary subsystems such as Anode, Cathode and Coolant. The sensitivity analysis in the OpenModelica model is performed by evaluating the effects of demand inputs, such as fuel utilisation and temperature, on stack performance. Feedback control techniques such as PI control are applied to the system model to maximise power output while serving as a base for optimization of parameters and operation strategies. The model is designed to be a base model, which can be equipped with different data from different manufacturers. This allows for efficient evaluation of design modifications and optimisation of model parameters for improved fuel cell performance and energy efficiency. The results obtained reveal that PEMFCS is able to accurately predict the desired voltage/power based on the demand input with a relatively low error. This model can be integrated into the Dymola environment along with existing vehicle models to act as a power source for a comprehensive analysis and testing.