摘要
从系统的角度研究了信息融合技术的基础理论,在介绍了信息融合系统的融合层次、功能模型和结构模型后,重点研究了信息融合的常用算法的优缺点。小波分析、遗传算法、神经网络与混沌分析进行信息融合时可采用辅助式、合作式、松散耦合式、紧致耦合式等。结果表明:在应用混沌预测前,小波消噪是有效的,遗传算法和神经网络相结合取长补短为混沌预测提供支持也是有效的。
This paper studied the fundamental theory of information integrated technology from the systematic perspective. After the introduction of integration hierarchy, function model and framework model, this paper mainly focused on the research of advantages and disadvantages of common algorithms used in information Integration. Patterns such as auxiliary mode, cooperation pattern, loose coupling mode and compact coupling mode were taken when wavelets analysis, genetic algorithms, neural networks and chaos analysis were applied in information integration. The results showed that wavelet de-noising is effective before applying the chaos prediction. The advantages of genetic algorithms and neural networks can complement each other, when they are combined used together. It is effective that this combination provides support for chaos predietion.
出处
《水资源与水工程学报》
2009年第5期91-96,共6页
Journal of Water Resources and Water Engineering
关键词
神经网络
小波消噪
遗传算法
混沌预测
neural networks
wavelet de-noising
genetic algorithms
chaos prediction
