ANN层状介质结构识别方法的鲁棒性研究
Study on the Robustness of Neural Network for Layered Media Structure Recognition
-
摘要: 通过对理论信号的实测信号的分析,研究了人工神经网络对层状介质结构识别的 鲁棒性,分析了层状介质物理参数的变化对神经网络识别效果的影响.实验结果表明.各介 质参数在一定范围内变化时,所得神经网络具有较强的鲁棒性.该研究结果反映出利用神经 网络进行层状介质结构识别具有较强的实用价值.Abstract: The robustness of artificial neural network is a key to the recognition using artificial network for layered media structure recognition. In this paper, the robustness of neural network for layered media structure recognition is studied by analyzing the theoretical simulative signal and measured signal from real layered media. The study shows that the media parameter does not affect the recognition of the neural network, and that the neural network scheme is practical.
-
Key words:
- Neural network /
- layered media structure recognition /
- robustness
-
计量
- 文章访问数: 2381
- HTML全文浏览量: 57
- PDF下载量: 948
- 被引次数: 0
下载:
