| 电阻率二维神经网络反演 |
| |
| 引用本文： | 徐海浪,吴小平.电阻率二维神经网络反演[J].地球物理学报,2006,49(2):584-589. |
| |
| 作者姓名： | 徐海浪 吴小平 |
| |
| 作者单位： | 中国科学院壳幔物质与环境实验室，中国科学技术大学地球与空间科学学院，合肥230026 |
| |
| 基金项目： | 国家自然科学基金项目（40374025）和新世纪优秀人才支持计划资助. |
| |
| 摘 要： | 由于非线性特性地球物理反演一直以来都是一个比较困难的问题. 近十年来，非线性反演方法如人工神经网络、遗传算法在地球物理数据解释中得到越来越多的应用，但目前基本仍限于一维反演问题. 对于二维反问题，反演参数较多，神经网络反演运用较少. 本文利用BP神经网络优化方法，实现了电阻率二维非线性反演. 与传统线性化的迭代反演比较，神经网络反演能够克服传统方法的不足、获得更好的反演结果.
|
| 关 键 词： | 电阻率 二维反演 反向传播网络 |
| 文章编号： | 0001-5733（2006）02-0584-06 |
| 收稿时间： | 2004-12-02 |
| 修稿时间： | 2004-12-022005-11-14 |
2_D resistivity inversion using the neural network method |
| |
| XU Hai-Lang,WU Xiao-Ping.2_D resistivity inversion using the neural network method[J].Chinese Journal of Geophysics,2006,49(2):584-589. |
| |
| Authors: | XU Hai-Lang WU Xiao-Ping |
| |
| Institution: | CAS Key Laboratory of Crust_Mantle Materials and Environments, School of Earth and Space Sciences, University of Science and Technology of China, Hefei 230026, China |
| |
| Abstract: | Geophysical inversion is a very difficult problem due to its non_linear nature. In the past decade, non_ linear inversion algorithms such as artificial neural network (NN) and genetic algorithm (GA) are increasingly used for the interpretation of geophysical data. However, until now, most of geophysical inversions using NN are limited to one_dimensional (1_D) models. As to 2_D inverse problems, NN inversion is hardly used because of a large number of parameters. In this paper, 2_D resistivity non_linear inversion is developed using the Back_Propagation (BP) neural network method. Compared to the traditional iterative inversion method through linearization, the neural network inversion is able to overcome disadvantages of the traditional inversion and obtain better results. |
| |
| Keywords: | Resistivity 2-D inversion Back-Propagation (BP) network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《地球物理学报》浏览原始摘要信息 |
| 点击此处可从《地球物理学报》下载免费的PDF全文 |
