Open AccessModeling the Electrical Conductivity of Anode for Solid Oxide Fuel Cell using Support Vector Regression Machine
Author(s) -
Junying Tang,
Ping Huang
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/562/1/012095
Subject(s) - anode , electrical resistivity and conductivity , solid oxide fuel cell , mean absolute percentage error , support vector machine , conductivity , oxide , fuel cells , materials science , regression analysis , generalization , linear regression , approximation error , regression , analytical chemistry (journal) , biological system , computer science , algorithm , mean squared error , chemistry , chemical engineering , mathematics , statistics , electrical engineering , metallurgy , machine learning , engineering , electrode , mathematical analysis , biology , chromatography
The electrical conductivity of Solid Oxide Fuel Cell (SOFC) anode is one of the most important indexes effect the efficiency of SOFC. In order to improve performance of fuel cell systems, it is necessary to have model which modeling the electrical conductivity. In this paper, a model using Support Vector Regression Machine (SVRM) was established to modeling the electrical conductivity of La 0.75 Sr 0.25 Cr 0.5 Mn 0.5 O 3- δ - x CuO (LSCM- x Cu) composite anode under two influence factors, including operating temperature ( T ) and Cu content ( x ) in LSCM- x Cu composites anode. The test result by SVRM support that the generalization ability of SVRM model is with high accuracy. The mean absolute error ( MAE ) of 4 test samples is 0.32, mean absolute percentage error ( MAPE ) is 1.05%, multiple correlation coefficients ( R 2 ) is 1.00, which is quite satisfied with the engineering demand. This investigation suggests that SVRM is a powerful tool to be used for optimal designing or controlling the technological process of fuel cell system.