An online spatiotemporal temperature model for high temperature polymer electrolyte fuel cells

This paper focuses on the thermal behavior estimation for high temperature polymer electrolyte fuel cells (HT-PEFCs), which can be used for predicting stack temperature and developing a reliable thermal controller. To overcome the complex nonlinear, time-space-distributed properties in fuel cell stacks, an online spatiotemporal temperature model is developed here. This model is based on the least squares support vector machine (LS-SVM), which has the ability to approximate any nonlinear system by a simple model structure. In the proposed method, the spatial correlations across different locations are fully represented by the kernel function, and then the temporal features are further modeled as a function of the inlet flow rate, partial pressure on each side and current density. After that, integrating the spatial kernel function and temporal dynamics function, the spatiotemporal temperature model is then constructed. The proposed method is tested by simulation in Matlab, with a comparison of the experimental data collected from an HT-PEFC test rig. The modeling result demonstrates that the developed model can effectively and precisely predict HT-PEFC thermal behavior. The present study is of key importance and may help as a black box for future development of new optimization and control strategies.