Degradation Prediction for Fuel Cell Based on Random Forest Regression and Grey Wolf Optimizer Algorithm

Accurately predicting the degradation of fuel cells and then extending their lifespan through reasonable management measures is crucial. This paper proposes a novel method that combines random forest regression with grey wolf optimization algorithm to predict the degradation of fuel cells. First, the measurement data is reconstructed using a Gaussian average weighted smoothing method to reduce the measurement noise. Then, a degradation model of fuel cells is established using random forest regression. Finally, the hyperparameters of the random forest are optimized using the grey wolf optimization algorithm to improve the accuracy of degradation prediction. The effectiveness of this method is verified through degradation experiments conducted at two different load currents. The test results demonstrate that proposed method can significantly improve the accuracy of degradation prediction and outperform other methods in terms of precision.