[1]覃志武,谢晋雄,蔡伊娜,等.基于环境与神经网络的软件自适应建模方法[J].深圳大学学报理工版,2017,34(No.6(551-660)):604-610.
 Qin Zhiwu,Xie Jinxiong,Cai Yina,et al.Software adaptive modeling method based on environment and neural network[J].Journal of Shenzhen University Science and Engineering,2017,34(No.6(551-660)):604-610.
点击复制

基于环境与神经网络的软件自适应建模方法()
分享到:

《深圳大学学报理工版》[ISSN:1000-2618/CN:44-1401/N]

卷:
第34卷
期数:
2017年No.6(551-660)
页码:
604-610
栏目:
电子与信息科学
出版日期:
2017-11-30

文章信息/Info

Title:
Software adaptive modeling method based on environment and neural network
作者:
覃志武1谢晋雄1蔡伊娜1闫毅宣12
1) 深圳市检验检疫科学研究院,广东深圳 518010;2) 河北师范大学数学与信息科学学院,河北石家庄050024
Author(s):
Qin Zhiwu1 Xie Jinxiong1 Cai Yina1 and Yan Yixuan12
1) Shenzhen Academy of Inspection and Quarantine, Shenzhen 518045, Guangdong Province, P.R.China 2) College of Mathematics and Information Science, Hebei Normal University, Shijiazhuang 050024, Hebei Province, P.R.China
关键词:
软件自适应环境需求软件建模神经网络环境用例预测
Keywords:
software adaptive environment requirements software modeling neural network environment use case prediction
文献标志码:
A
摘要:
软件需求模型的建模是保证软件可靠运行的基础.传统方法在建模过程中没有对环境需求加以区分,对环境的变化无法有效识别和合理应对,导致软件生命周期缩短.现有自适应建模过程属于被动感知需求,没有有效对未来的需求加以预测和应对,同样无法延长软件生命周期.为了尽可能延长软件生命周期,同时提高重构开发效率,提出一种新颖的针对环境变化的软件自适应建模方法.该方法将软件运行所处的环境作为需求单独分析处理,首先识别环境用例,其次构建环境用例并将功能指标进行量化处理,最后用BP神经网络预测环境需求变化并作出应对策略.案例研究表明该方法不仅可有效提高建模效率,而且突出了对环境良好的适应性.
Abstract:
The requirement modeling of software is the basis for improving the efficiency of development and ensuring the reliable operation. However, traditional methods neither can distinguish the environmental requirements during the modeling process, nor can be effectively identified and give reasonable response to the changes in the environment. Meanwhile, the existing software adaptive modeling process belongs to passive sensing requirements and does not effectively predict and deal with future requirements. According to the above problems, a novel software adaptive modeling method is proposed for environment change. The method utilizes the environment in the software as a separate analysis of requirements. The environment use case is identified, constructed, and the function index is quantified. Finally, the BP neural network is used to predict the change of the environment requirements and make the corresponding strategy. The case study shows that this method not only can effectively improve the modeling efficiency, but also highlights the good adaptability to the environment.
更新日期/Last Update: 2017-10-10