Online parameter identification of a retrofitted hydrogen genset for maximum efficiency

This work presents an electric vehicle range extension system based on a hydrogen genset which is obtained by retrofitting a 5kW gasoline genset equipped with two generators. The power provided by the hydrogen genset is used to extend the battery pack autonomy through a serial topology. This genset has the potential to be less expensive and more robust to extreme low operating temperature than a fuel cell. In addition, it can operate at different hydrogen-air mixtures. To improve the genset energy efficiency, the operating conditions that lead to a maximum efficiency need to be selected online. We propose a two-step method in which the genset efficiency behavior is represented by a parametric smooth surface in the first step whilst the operating conditions for a maximum efficiency is selected during the second step. These smooth surface parameters, identified using the recursive least square method, provided a root mean square error of the deviation between the estimated and the measured efficiency time profiles less than 2.5 ×10⁻³. Using the line search method, the optimal operating condition has been identified and validated.