Generalized Regression Neural Network Based Meta-Heuristic Algorithms for Parameter Identification of Proton Exchange Membrane Fuel Cell

Abstract

An accurate parameter extraction of the proton exchange membrane fuel cell (PEMFC) is crucial for establishing a reliable cell model, which is also of great significance for subsequent research on the PEMFC. However, because the parameter identification of the PEMFC is a nonlinear optimization problem with multiple variables, peaks, and a strong coupling, it is difficult to solve this problem using traditional numerical methods. Furthermore, because of insufficient current and voltage data measured by the PEMFC, the precision rate of cell parameter extraction is also very low. The study proposes a parameter extraction method using a generalized regression neural network (GRNN) and meta-heuristic algorithms (MhAs). First of all, a GRNN is used to de-noise and predict the data to solve the problems in the field of PEMFC, which include insufficient data and excessive noise data of the measured data. After that, six typical algorithms are used to extract the parameters of the PEMFC under three operating conditions, namely high temperature and low pressure (HTLP), medium temperature and medium pressure (MTMP), and low temperature and high pressure (LTHP). The last results demonstrate that the application of GRNN can prominently decrease the influence of data noise on parameter identification, and after data prediction, it can greatly enhance the precision rate and reliability of MhAs parameter identification, specifically, under HTLP conditions, the V-I fitting accuracy achieved 99.39%, the fitting accuracy was 99.07% on MTMP, and the fitting accuracy was 98.70%.