Control algorithm based on an experimental approach for PEM fuel cell systems efficiency optimization

Nowadays, PEM fuel cells are considered as one of the most promising electric energy production technology using hydrogen. For this reason, tracking the efficiency of the PEM fuel cell system is an important research topic. Most of the control algorithms developed to aim this goal are based on models of the PEMFC although to date there is no complete model which takes into account all the phenomenon related to the PEMFC. In this paper, we propose a control algorithm based on an experimental approach which searches for operation parameters (stack temperature, air relative humidity and air stoichiometric ratio) to optimize the efficiency of the PEMFC system. Indeed, an experimental study done beforehand showed that by acting on these parameters, the efficiency of the system which depends highly on the electric power generated by the stack, on the power lost in the auxiliaries and on the hydrogen flow rate, can be optimized. The developed algorithm is based on a local optimization method derived from the line search method. The validity of this method has been proven with different classic functions and the control algorithm has been implemented experimentally in the control interface of our PEMFC system. The experiments analysis showed promising results.