Gas mixing detection in Solid Oxide Fuel Cells using Machine Learning approach

固体氧化物燃料电池(SOFC)是一种将化学能(氢或含有氧化剂的碳氢化合物)转化为电能的电化学装置。SOFC适用于为重型汽车提供动力或固定发电。SOFC的工作温度为650摄氏度或更高。高温运行环境会产生应力，并可能导致脆性燃料电池出现裂纹。这些裂缝允许气体混合和不受控制的燃烧，最终可能会摧毁燃料电池。该项目致力于及时检测这些裂缝，这对SOFC的安全运行至关重要。该项目有两项计划的研究成果。第一个研究建立了一个模型，该模型基于气体流动和裂纹情况下的热释放来分析SOFC电堆的热分布。第二个研究构建了机器学习分类器来检测裂纹故障。

Solid Oxide Fuel Cells (SOFCs) are electrochemical devices that converts chemical energy (of hydrogen or hydrocarbons with oxidants) into electrical energy. SOFCs are suitable for powering heavy duty automotive vehicles or for stationary power generation. SOFCs operate at 650 degreesC or above. The high temperature operating environment causes stress and can lead to cracks on the brittle fuel cell. The cracks allow gas mixing and uncontrolled combustion that could eventually destroy the fuel cell. This project addresses the timely detection of these cracks that is paramount for the safe operation of SOFCs. The project has two planned research outputs. The first study builds a model that analyse the heat distribution of a SOFC stack based on the gas flow and heat release of crack scenarios. The second study builds machine learning classifiers to detect the crack faults.