|
|||||
|
|
| 基于神经网络的空间流场智能模拟 | |
| 引用本文： | 顾正华,唐洪武,高柱,李云,肖洋.基于神经网络的空间流场智能模拟[J].系统仿真学报,2004,16(7):1372-1375. |
| 作者姓名： | 顾正华 唐洪武 高柱 李云 肖洋 |
| 作者单位： | 1. 河海大学,水利水电工程学院,南京,210098;南京水利科学研究院,水工研究所,南京,210029 2. 河海大学,水利水电工程学院,南京,210098 3. 南京水利科学研究院,水工研究所,南京,210029 |
| 基金项目： | 河海大学科技创新基金(2002406443) |
| 摘 要： | 基于作者提出的水流智能模型思想,运用人工神经网络方法对透水框架式四面体的尾流场进行智能模拟,为复杂水流的空间流场仿真提供一种新的研究途径。通过水槽实验,借助声学多普勒流速仪(ADV)自动测量分析系统测得四面体九种工况下的空间尾流场,作为四面体水流智能模型的学习样本和测试样本,经过训练获得智能模型的工作表达式,预测流场与实测流场基本吻合。研究结果表明,基于神经网络的水流智能模型,仿真速度快,外延泛化能力和容错能力强,且精度较高,因此具有广阔的应用前景。 |
| 关 键 词： | 流场 水流智能模型 智能模拟 透水框架式四面体 人工神经网络 |
| 文章编号： | 1004-731X(2004)07-1372-04 |
| 修稿时间： | 2003年5月31日 |
ANN-based Intelligent Simulation on Flow Field |
|
| GU Zheng-hua,TANG Hong-wu,GAO Zhu,Li Yun,Xiao Yang.ANN-based Intelligent Simulation on Flow Field[J].Journal of System Simulation,2004,16(7):1372-1375. | |
| Authors: | GU Zheng-hua TANG Hong-wu GAO Zhu Li Yun Xiao Yang |
| Institution: | GU Zheng-hua1,2,TANG Hong-wu1,GAO Zhu1,Li Yun2,Xiao Yang1 |
| Abstract: | On the basis of Flow Intelligent Model (FIM) conception from Gu Zheng-hua and Tang Hong-wu, penetrating frame-tetrahedrons wake field was simulated through artificial neural networks method, and a new idea was provided on flow fields simulation. Through flume experiment with Acoustic Doppler Velocimeter (ADV) system, nine kinds of penetrating frame-tetrahedrons wake fields on the different conditions were measured as learning and testing samples of penetrating frame-tetrahedrons FIM. The expression of the wake fields intelligent model was obtained by training, and its simulation result was consonant with prototype. Conclusions illustrate that FIM-based ANN will be prosperous for its rapid simulation speed, good extension generalization capability and tolerance, and high precision. |
| Keywords: | flow field flow intelligent model intelligent simulation penetrating frame-tetrahedron artificial neural networks |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |