Model identification of proton-exchange membrane fuel cells based on a hybrid convolutional neural network and extreme learning machine optimized by improved honey badger algorithm

In this research, a new optimized deep learning-based methodology is proposed for optimal and efficient modeling of the Proton-exchange membrane fuel cells. Here, a hybrid method based on Convolutional Neural Network (CNN) and Extreme Learning Machine (ELM) network is used to provide the purpose. The model is then optimized based on a new improved metaheuristic, called Improved Honey Badger Algorithm (IHBA) to get optimal results. The reason for using the improved version of the IHBA is for improving the model results in terms of accuracy and to provide a high-speed convergence for the algorithm. The designed model is then performed to a model to verify its effectiveness. The results indicate that the proposed model has a promising confirmation with the experimental training data where the maximum error rate is 0.039. The results of the proposed model are then compared with a CNN-based model estimator to validate its higher efficiency. Final results show that although both model estimators have good confirmation with the experimental data, the proposed model provides more satisfying results with less error value.