Sensitivity analysis of the fuel economy strategy based on load-following control of the fuel cell hybrid power system

In this paper, a real-time strategy for a fuel cell hybrid power system based on fuel optimization and load following is analyzed to identify the critical parameters of the optimization problem. Because the dither's frequency dictates the searching speed of the optimum in the perturbed extremum seeking algorithm, this parameter has been selected from multiple parameters involved in the optimization problem to improve the fuel economy based on sensitivity analysis. The sensitivity analysis reveals the multimodal behavior of the fuel economy, but also the best choice for the dither. The fuel cell hybrid power system is a highly dynamic system, so the optimization and control loops have been designed for a robust but efficient operation of the system based on new perturbed extremum seeking algorithm to track the optimum. Considering the recommended 100 Hz sinusoidal dither, the improvement in fuel economy is of 47.9 L for an 8 kW/12 s load cycle. Also, the electrical efficiency of the fuel cell system increases with 7.83%. Both improvements were estimated compared to a commercial strategy, called in the literature as static feed-forward strategy.