节点文献
应用人工神经网络计算孤立波爬高
Using Artificial Neural Network to Calculate Solitary Run-up
【摘要】 运用附加动量法和自适应学习速率对具有一层隐含层的 BP 网络进行改进,并由此建立神经网络模型来计算孤立波的爬高。将试验值分别与 Synolakis 公式以及本文结果进行比较,证实神经网络模型计算方法对于计算孤立波的爬高是合适的。
【Abstract】 By the use of the additional momentum method and the adaptive learning rate, BP network with one hidden layer is modified, and the established Neural Network model is used to calculate solitary run-up on plane slope. The comparison between the experiment value and the calculated results by Synolakis formula and the aforementioned method shows that the present model can be suitable to predict solitary run-up.
【关键词】 孤立波;
爬高;
人工神经网络;
BP;
网络;
附加动量法;
自适应学习速率;
【Key words】 solitary wave; run-up; artificial Neural Network; BP network; additional momentum method; adaptive learning rate;
【Key words】 solitary wave; run-up; artificial Neural Network; BP network; additional momentum method; adaptive learning rate;
- 【文献出处】 海洋通报 ,Marine Science Bulletin , 编辑部邮箱 ,2003年01期
- 【分类号】P731
- 【被引频次】1
- 【下载频次】64