Fuel cell residual life prediction method based on deep learning

Abstract

The invention relates to a fuel cell residual service life prediction method based on deep learning. By adopting a deep neural network model, the accurate prediction of the RUL of the fuel cell is realized, and the method belongs to the technical field of proton exchange membrane fuel cell health state monitoring. The method comprises the following specific steps: S1, acquiring monitoring data of the fuel cell, and carrying out noise reduction processing; s2, fitting the training data after noise reduction processing, and establishing a nonlinear mapping relation between the current input data and the target data; s3, establishing a loss function and optimizing neural network model parameters through a feedback derivation method to obtain an optimal neural network model; and S4, taking the input data of the prediction starting point as the input of the optimal DNN so as to realize iterative rolling prediction. And S4, simply summarizing as follows: iteratively using the prediction output of the neural network at the current moment as the prediction input of the next moment, and iteratively predicting the voltage or power output in a rolling manner, thereby realizing the RUL prediction of the fuel cell.

Description

technical field [0001] The invention relates to a method for predicting the remaining life of a fuel cell based on deep learning, and belongs to the technical field of proton exchange membrane fuel cells. Background technique [0002] As an environmentally friendly and clean energy, fuel cells have a good prospect in the application of new energy fuel cell electric vehicles by virtue of their huge advantages in power conversion efficiency. However, compared with other energy devices such as lithium-ion batteries, high operating costs, short life cycle, and safe and reliable operation are still the main reasons that restrict the wide application of fuel cells. For example, the most popular proton exchange membrane fuel cell system (PEMFC), its main characteristics are good quick start performance, high power density (3.8 ~ 6.5kW / m3) and low operating temperature (50 ~ 80 ℃). System prediction and health management (PHM) is an emerging field of scientific and technological de...

AI Technical Summary

Problems solved by technology

However, the current fusion algorithm largely depends on whether the exponential or logarithmic model it constructs can accurately describe the attenuation of battery life.

Images