期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
生成对抗网络及其图像处理应用研究进展 被引量:8
1
作者 王晋宇 杨海涛 +2 位作者 李高源 张长弓 冯博迪 《计算机工程与应用》 CSCD 北大核心 2021年第8期26-35,共10页
生成对抗网络(GAN)是一种基于对抗思想的架构体系。作为人工智能大发展背景下诞生的前沿算法,GAN已经在图像处理的多个领域取得了显著的成果。从传统GAN的算法入手,对其模型架构、数学机理、优缺点进行剖析。总结了具有代表性的GAN变体... 生成对抗网络(GAN)是一种基于对抗思想的架构体系。作为人工智能大发展背景下诞生的前沿算法,GAN已经在图像处理的多个领域取得了显著的成果。从传统GAN的算法入手,对其模型架构、数学机理、优缺点进行剖析。总结了具有代表性的GAN变体,并对GAN在图像处理方面的前沿应用进行介绍。结合现有GAN发展依然存在的问题,对GAN的发展趋势进行了展望。 展开更多
关键词 生成对抗网络 纳什均衡 结构 损失变体
下载PDF
Research on Wet Steam Spontaneous Condensing Flows Considering Phase Transition and Slip 被引量:1
2
作者 Ke CUI Huan-long CHEN +1 位作者 Yan-ping SONG Hiroharu OYAMA 《Journal of Thermal Science》 SCIE EI CAS CSCD 2013年第4期320-326,共7页
A new dual-fluid model considering phase ansition and velocity slip was proposed in this paper and the Cunningham correction was used in the droplet resistance calculation. This dual-fluid model was applied to the num... A new dual-fluid model considering phase ansition and velocity slip was proposed in this paper and the Cunningham correction was used in the droplet resistance calculation. This dual-fluid model was applied to the numerical simulations of wet steam flow in a 2D LAVAL nozzle and in the White cascade respectively. The results of two simulations demonstrate that the model is reliable. Meanwhile, the spontaneous condensing flow in White cascade was analyzed and it infers that the irreversible loss caused by condensation accounts for the largest share (about 8.78% of inlet total pressure) in total pressure loss while the loss caused by velocity slip takes the smallest share (nearly 0.42%), and another part of total pressure loss caused by pneumatic factors contributes a less share than condensation, i.e. almost 3.95% of inlet total pressure. 展开更多
关键词 dual-fluid model wet steam spontaneous condensing flow LAVAL nozzle White cascade.
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部