| 基于IGA算法的电阻率神经网络反演成像研究 |
| |
| 引用本文： | 高明亮,于生宝,郑建波,徐畅,刘伟宇,栾卉.基于IGA算法的电阻率神经网络反演成像研究[J].地球物理学报,2016,59(11):4372-4382. |
| |
| 作者姓名： | 高明亮 于生宝 郑建波 徐畅 刘伟宇 栾卉 |
| |
| 作者单位： | 吉林大学仪器科学与电气工程学院, 长春 130026 |
| |
| 基金项目： | 地面电磁探测（SEP）系统研制-野外试验研究（201311193-05（SinoProbe-09-02-05））资助 |
| |
| 摘 要： | 为满足地球物理资料反演解释的高精度、快速、稳定的要求,本文结合免疫遗传算法寻优速度快和BP神经网络反演不依赖初始模型等优点,设计了一种将BP神经网络和免疫遗传算法进行有机结合的全局优化反演策略,并将该策略成功地应用于二维高密度电法数据反演.利用免疫遗传算法(Immune Genetic Algorithm,简称IGA)对神经网络的反演参数进行同步优化,提高了电阻率反演的精度.仿真和实验结果验证设计的全局优化反演策略取得了较好的效果,通过与线性反演方法和BP法以及遗传神经网络法等反演方法进行比较,得出该方法具有反演精度更高,反演时间更短等显著优势的结论.
|
| 关 键 词： | 免疫遗传算法 BP神经网络 高密度电阻率法 反演精度 |
| 收稿时间： | 2015-09-02 |
Research of resistivity imaging using neural network based on immune genetic algorithm |
| |
| GAO Ming-Liang,YU Sheng-Bao,ZHENG Jian-Bo,XU Chang,LIU Wei-Yu,LUAN Hui.Research of resistivity imaging using neural network based on immune genetic algorithm[J].Chinese Journal of Geophysics,2016,59(11):4372-4382. |
| |
| Authors: | GAO Ming-Liang YU Sheng-Bao ZHENG Jian-Bo XU Chang LIU Wei-Yu LUAN Hui |
| |
| Institution: | College of Instrumentation and Electrical Engineering, Jilin University, Changchun 130026, China |
| |
| Abstract: | In order to meet the requirements of high precision, high speed and stability in geophysical inversion, we design a global optimization inversion strategy based on BP neural network and immune genetic algorithm, considering the fast searching speed of immune genetic algorithm and the independence of the initial model in BP neural network inversion. The strategy is successfully applied to the two-dimensional high density resistivity inversion. By using Immune Genetic Algorithm (IGA) for synchronous optimization of the neural network inversion parameters, the precision of the resistivity inversion can be improved. The results of experiment and simulation verify that the global optimization strategy achieves great results. Comparing with the linear inversion method, BP method, and genetic neural network method, our method has higher precision and shorter time of inversion. |
| |
| Keywords: | Immune Genetic Algorithm BP neural network High density resistivity method Inversion precision |
| 本文献已被 CNKI 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 |
|
