Method for predicting service life of fuel cell

Abstract

The invention discloses a fuel cell life prediction method comprising the following steps: S1, reading original life degradation data of a fuel cell, and preprocessing the original life degradation data to obtain preprocessed data; s2, carrying out intrinsic mode decomposition on the preprocessed data by utilizing an empirical mode decomposition algorithm to obtain a series of intrinsic mode components; s3, randomly initializing weight parameters of the recurrent neural network; s4, determining the training duration, the size of a storage pool and the spectral radius; s5, inputting the intrinsic mode components into the recurrent neural network for training to obtain a training model; s6, performing prediction of a prediction step length on the intrinsic mode component through the training model to obtain prediction data of the intrinsic mode component corresponding to the prediction step length; and S7, fusing all the prediction data in the step S6 to obtain the prediction data of the original life of the fuel cell. The method has high prediction precision and is suitable for online real-time prediction.

Description

technical field [0001] The present application belongs to the technical field of fuel cells, and in particular relates to a life prediction method for fuel cells. Background technique [0002] A fuel cell is a power generation device that directly converts the chemical energy in the fuel into electrical energy through an electrochemical reaction. Compared with traditional energy sources, fuel cells are an efficient and clean electrochemical power generation device, which has received widespread attention at home and abroad in recent years. [0003] Among them, proton exchange membrane fuel cell power generation technology is a technology with great application potential and industrial prospects. Because of its non-polluting, high efficiency, low noise and other characteristics, it is widely used in automotive power, distributed power generation and portable power supplies. At present, the operating life of fuel cells is one of the main factors restricting their further comm...

AI Technical Summary

Problems solved by technology

[0005] Based on this, it is necessary to address the above technical problems and provide a fuel cell life prediction method to solve the problems that the existing prediction methods can only be single-step prediction or multi-step prediction accuracy is not high.

This application uses recurrent neural network combined with empirical mode decomposition technology to mathematically model fuel cells, uses recurrent neural network to learn known time series data, and performs multi-step prediction of unknown performance degradation trends, and can achieve higher prediction Accuracy, which effectively solves the problem of low accuracy of multi-step life prediction of fuel cells at present

Images