将模拟退火算法用于神经网络,对AB5型储氢合金的初始放电容量性能进行预测.通过实验确定了冷却进度表中各项参数并讨论了其对网络预测性能的影响,提出了实用的冷却制度;将模拟退火算法与传统的梯度下降法结合,用于人工神经网络的优化与预测,得到了预测效果等与迭代时间性能更好的神经网络.
参考文献
[1] | Pasturel A;Chatilllon-Conlinet C .ThermodynamicandStructuralPropertiesofLaNi4MCompoundsandTheirRelatedHydrides[J].Journal of the Less-Common Metals,1983,83:73. |
[2] | Sakai T;Qguro K;Miyamura H et al.Some Factors Affecting the Cycle Lives of LaNi5-based Alloy Electrodes of Hydrogen Batteries[J].Journal of the Less-Common Metals,1990,161:193. |
[3] | Guo J;Li C H;Liu H L et al.Computer-Aided Materials Design of La-Rich MH Electrode[J].Materials Science and Engineering B,1997,47:96. |
[4] | Guo J.;Liu HL.;Chen NY.;Li CH. .EFFECT OF RARE EARTH COMPOSITIONS ON THE HYDROGEN STORAGE PROPERTIES OF AB(5)-TYPE ALLOYS AND PATTERN RECOGNITION ANALYSIS[J].Journal of the Electrochemical Society,1997(7):2276-2278. |
[5] | Kok L H;Hong M J;Yi L.Computer Aided Design of Ni/MH Electrodes[J].Journal of Materials Chemistry,1999(09):937. |
[6] | 阎平凡;张水.人工神经网络与模拟进化计算[M].北京:清华大学出版社,2000:300. |
[7] | 康立山;谢云;尢矢勇.非并行计算[M].北京:科学出版社,1994:136. |
[8] | 部茂祖,姜俊峰,李静梅.模拟退火算法中冷却调度选取方法的研究[J].计算机工程,2000(09):63-64,66. |
上一张
下一张
上一张
下一张
计量
- 下载量()
- 访问量()
文章评分
- 您的评分:
-
10%
-
20%
-
30%
-
40%
-
50%