@inproceedings{aa69c1444636487e8ab3e24575766647,
title = "PEMFC power prediction based on deep auto-encoder and LS-SVMR",
abstract = "A hybrid method based on deep auto-encoder and Least Squares Support Vector Machines Regression(LS-SVMR) is proposed in this paper to anticipate failures in proton exchange membrane fuel cell (PEMFC). Deep learning to deal with high-dimensional data and Support Vector Machines (SVM) to classification accuracy are combined. Through training, a Deep Neural Network of Auto-Encoder (DNNAE) is used to reduce the training data without losing essential information and PEMFC power is then predicted by LS-SVMR using the reduced data. Experimental results show that PEMFC power can be predicted with high accuracy using health monitoring data. Compared with LS-SVMR without using DNNAE, the result indicates that a training data size same as 25\% original dimension allows to improve the forecasting accuracy.",
keywords = "Auto-encoder, Deep learning, LS-SVMR, PEMFC power prediction, Prognosties and health management",
author = "Sheng Hong and Longlong Sun and Yin, \{Jia Wei\} and Ting Yu and Yifan Wang and Wei Zhu",
note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 3rd IEEE International Conference on Big Data Analysis, ICBDA 2018 ; Conference date: 09-03-2018 Through 12-03-2018",
year = "2018",
month = may,
day = "25",
doi = "10.1109/ICBDA.2018.8367714",
language = "英语",
series = "2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "391--396",
booktitle = "2018 IEEE 3rd International Conference on Big Data Analysis, ICBDA 2018",
address = "美国",
}