質子交換膜燃料電池性能之穩健參數設計與分析
博士 === 國防大學中正理工學院 === 國防科學研究所 === 97 === The dissertation aims at investigating the performance of proton exchange membrane fuel cell (PEMFC) by robust parameter design (RPD) for ensuring lower variance and optimum of electric power. There are numerous operating parameters affecting the PEMFC operat...
Main Authors: | , |
---|---|
Other Authors: | |
Language: | zh-TW |
Published: |
2008
|
Online Access: | http://ndltd.ncl.edu.tw/handle/30113339473957229842 |
id |
ndltd-TW-097CCIT0584001 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-097CCIT05840012017-09-15T04:39:38Z http://ndltd.ncl.edu.tw/handle/30113339473957229842 質子交換膜燃料電池性能之穩健參數設計與分析 Yu Wei Lung 余威龍 博士 國防大學中正理工學院 國防科學研究所 97 The dissertation aims at investigating the performance of proton exchange membrane fuel cell (PEMFC) by robust parameter design (RPD) for ensuring lower variance and optimum of electric power. There are numerous operating parameters affecting the PEMFC operating performance, such as fuel cell operating temperature, anode and cathode humidification temperature, operating pressure and reactant flow rate. This study adopted a fractional factorial design of the design of experiments (DOE) with Taguchi method to screen first whether these factors have significant effects on a response and the interactions between various parameters. The orthogonal array of the Taguchi method is then utilized to determine efficiently the optimal combination of factors for a fuel cell. This dissertation also presents an integrated approach that combines the Taguchi method with neural networks to prevent discrete parameter levels in the Taguchi method from the estimation of the real optimum. The Taguchi method including the statical and the dynamic characteristics first acquires the primary optimums of the operating parameters in the PEMFC. Each treatment in a row of the orthogonal array together with its relative responses was used to establish a set of training patterns (input/target pair) to the neural network. The neural network can then construct relations between the control factors and responses in the PEMFC. The actual optimums of the operating parameters in the PEMFC were obtained by the trained neural network. Experimental results are presented for identifying the proposed approach. In the experiment, a single PEMFC (1-cell PEMFC) and a collection of three PEM fuel cells in series (3-cell stack) are adopted for experiments. The analyses by the results of the experiments reveal that the operating temperature, the operating pressure and the interaction between operating temperature and operating pressure have a significant effect on the fuel cell performance for the 1-cell and the 3-cell stack. In addition, the cathode humidification temperature has also a significant effect on the 3-cell stack. While both the operating temperature and pressure increase simultaneously with that, the other factors are at appropriate conditions, it is possible to improve the performance of PEMFC. Wu Sheng Ju Shiah Sheau Wen 吳聖儒 夏曉文 2008 學位論文 ; thesis zh-TW |
collection |
NDLTD |
language |
zh-TW |
sources |
NDLTD |
description |
博士 === 國防大學中正理工學院 === 國防科學研究所 === 97 === The dissertation aims at investigating the performance of proton exchange membrane fuel cell (PEMFC) by robust parameter design (RPD) for ensuring lower variance and optimum of electric power. There are numerous operating parameters affecting the PEMFC operating performance, such as fuel cell operating temperature, anode and cathode humidification temperature, operating pressure and reactant flow rate. This study adopted a fractional factorial design of the design of experiments (DOE) with Taguchi method to screen first whether these factors have significant effects on a response and the interactions between various parameters. The orthogonal array of the Taguchi method is then utilized to determine efficiently the optimal combination of factors for a fuel cell.
This dissertation also presents an integrated approach that combines the Taguchi method with neural networks to prevent discrete parameter levels in the Taguchi method from the estimation of the real optimum. The Taguchi method including the statical and the dynamic characteristics first acquires the primary optimums of the operating parameters in the PEMFC. Each treatment in a row of the orthogonal array together with its relative responses was used to establish a set of training patterns (input/target pair) to the neural network. The neural network can then construct relations between the control factors and responses in the PEMFC. The actual optimums of the operating parameters in the PEMFC were obtained by the trained neural network. Experimental results are presented for identifying the proposed approach.
In the experiment, a single PEMFC (1-cell PEMFC) and a collection of three PEM fuel cells in series (3-cell stack) are adopted for experiments. The analyses by the results of the experiments reveal that the operating temperature, the operating pressure and the interaction between operating temperature and operating pressure have a significant effect on the fuel cell performance for the 1-cell and the 3-cell stack. In addition, the cathode humidification temperature has also a significant effect on the 3-cell stack. While both the operating temperature and pressure increase simultaneously with that, the other factors are at appropriate conditions, it is possible to improve the performance of PEMFC.
|
author2 |
Wu Sheng Ju |
author_facet |
Wu Sheng Ju Yu Wei Lung 余威龍 |
author |
Yu Wei Lung 余威龍 |
spellingShingle |
Yu Wei Lung 余威龍 質子交換膜燃料電池性能之穩健參數設計與分析 |
author_sort |
Yu Wei Lung |
title |
質子交換膜燃料電池性能之穩健參數設計與分析 |
title_short |
質子交換膜燃料電池性能之穩健參數設計與分析 |
title_full |
質子交換膜燃料電池性能之穩健參數設計與分析 |
title_fullStr |
質子交換膜燃料電池性能之穩健參數設計與分析 |
title_full_unstemmed |
質子交換膜燃料電池性能之穩健參數設計與分析 |
title_sort |
質子交換膜燃料電池性能之穩健參數設計與分析 |
publishDate |
2008 |
url |
http://ndltd.ncl.edu.tw/handle/30113339473957229842 |
work_keys_str_mv |
AT yuweilung zhìzijiāohuànmóránliàodiànchíxìngnéngzhīwěnjiàncānshùshèjìyǔfēnxī AT yúwēilóng zhìzijiāohuànmóránliàodiànchíxìngnéngzhīwěnjiàncānshùshèjìyǔfēnxī |
_version_ |
1718533645338148864 |