A prognostic-based dynamic optimization strategy for a degraded solid oxide fuel cell

In actual operation, solid oxide fuel cells are vulnerable to performance degradation. Predicting degradation trends and taking appropriate measures to reduce voltage degradation can prolong the lifetime of fuel cells. In addition, because the influence of related factors on the voltage degradation and efficiency are mutually coupled, it is necessary to take into account the efficiency when decreasing the voltage degradation. Therefore, in this study a prognostic-based dynamic optimization strategy is proposed to quantitatively extend the fuel cell's life without significantly compromising on the generating efficiency. For the lack of learning samples, a prognostic method based on the similarity of phase space trajectory is utilized to estimate the remaining useful life of the fuel cell, and it is compared with a traditional data-driven prognostic method. A multi-objective dynamic optimization method is employed to determine the optimal efficiency-voltage degradation solution through the predicted operating lifetime. Using the proposed prognostic-based dynamic optimization strategy, the results show that the lifetime of the fuel cell can be effectively extended without significant reduction in the generating efficiency for three test degradation trajectories.