Energy optimization of fuel cell system by using global extremum seeking algorithm

This paper presents a real-time optimization method and demonstrates its application to a Proton Exchange Membrane Fuel Cell (PEMFC) system. The optimization function was defined as mix of two performance indicators, the FC net power and the Fuel Consumption Efficiency, using the appropriate weighting coefficients. The weighting coefficients will modify the optimization surface and many extreme could appear on plateau of the optimization surface. The Global Extremum Seeking (GES) algorithm proposed here as real-time optimization method will locate and track the global maximum point and related to this will be established the optimal fueling rates for the PEMFC system under a given load. In this study four strategies will be tested, including the Static Feed-Forward (sFF) control strategy as reference. The optimal operating conditions were sought at different levels of load and the gaps between these four strategies were estimated. For example, in comparison with the PEMFC system controlled by the sFF strategy, the GES operation of the PEMFC system could increase the energy efficiency with 1-2.1%, depending on the FC current level and values used for the weighting coefficients. If the PEMFC system operates under variable load profile, the Fuel Consumption Efficiency could also increase with more than 0.54 W/lpm for GES & LF-based optimization strategy in comparison with the sFF strategy. The effectiveness of GES & LF-based optimization strategy was shown considering a FC Hybrid Power Source under variable load profile.