Research On Membrane Electrode Assembly Degradation And Life Prediction Of Exchange Membrane Fuel Cell

Proton exchange membrane fuel cell is an electrochemical device that can convert chemical energy into electric energy.It has the advantages of high energy conversion rate,high power density and no pollution,and is widely used in the automotive field.However,there are also problems of durability and life span.There is still a long way to go before the fuel cell is applied in a large scale.Membrane electrode is the core component of fuel cell,and the decline of fuel cell is mainly reflected in the decline of membrane electrode.Therefore,it is of positive significance to study the decline of membrane electrode and predict the life of fuel cell stack.The main research contents of this paper are as follows:Starting from the structure of fuel cell,this paper introduces its main components and functions,and analyzes its working principle.Based on the degradation mechanism of fuel cell membrane electrode,the degradation index of membrane electrode was extracted,and the fuel cell model was built by combining electrochemistry,equivalent circuit,proton exchange membrane,catalyst layer and gas diffusion layer model.The relationship between the degradation index of membrane electrode and output performance was established,which provided theoretical support for the simulation study of membrane electrode degradation.Based on the actual fuel cell design parameters,a fuel cell model was built in AMEsim,and the reliability of the model was verified by experiments.The influence of proton exchange membrane thickness,proton conductivity,electrochemical active area of catalyst layer,thickness of catalyst layer,thickness of gas diffusion layer and permeability attenuation of gas diffusion layer on polarization curve of fuel cell is analyzed by model,and the influence of membrane electrode degradation on output performance is obtained.Based on the test data set of fuel cell stack life under stable working conditions published by French Fuel Cell Laboratory,a method of stack life prediction based on the combination of wavelet threshold filtering denoising algorithm and deep learning technology is proposed by selecting voltage as the degradation index of the stack.Life prediction includes the prediction of the future degradation trend and the remaining service life of the stack.Firstly,the data is preprocessed,the wavelet threshold filtering algorithm is used to eliminate the data noise,and the characteristic parameters that affect the life span are selected from the data set in combination with the study of membrane electrode decay.Secondly,a recurrent neural network model with simple model structure is established to predict the life of the stack,and it is verified that feature selection can improve the prediction ability of the model.The complexity of the network model is further improved,and the long short term memory neural network is used to predict the life of the stack.Finally,the prediction results of the two models are compared.The research shows that the long short term memory neural network model has a better prediction effect on the remaining life and the future decline trend of the stack,and the long-short term memory neural network can predict the life of the stack with high accuracy.