神经网络在地震层析成像中的应用
|
摘要
Hopfield神经网络是一种单层的反馈型神经网络，它的主要优点是集成并行处理和收敛速度快。本文将该神经网络的原理用于地震层析成像，并导出了联接权系数和外部输入项在层析成像应用中的具体表达式。文中用该方法计算了两个模型实例。计算结果表明，Hopfield神经网络在地震层析成像中是行之有效的，可获取井间地层速度场或慢度场的最优估计。
Hopfield's neural network is a single-layered feedback neural network- Its ma-jor virtues are fast convergence and integrated parallel data processing. We applythis neural network principle to seismic tomography, and deduce the expressions ofconnective weight coefficients and input terms used in tornography.Two model examples are computed in this method. Computation result showsthat Hopfield's neural network is effective in seismic tomography and can be used toestimate optimurn cross-borehole velocity field (or slowness field).
Hopfield's neural network is a single-layered feedback neural network- Its ma-jor virtues are fast convergence and integrated parallel data processing. We applythis neural network principle to seismic tomography, and deduce the expressions ofconnective weight coefficients and input terms used in tornography.Two model examples are computed in this method. Computation result showsthat Hopfield's neural network is effective in seismic tomography and can be used toestimate optimurn cross-borehole velocity field (or slowness field).
引文
1 Hopfield J J. Neurons with graded response have collective computation properties like those of two-stateneurons. Proc Nat'l Acad Sci, 1984: 3088~3092
2 Hopfield J J and Tank D W. Neural computation of decisions in optimization problems. Biological Cybernetcis, 1985, 52, 141~ 152
3 Tank, D W and Hop field J J. Simple "neural"optimization network: An A/D converter, signal decisioncircuit, and a linear programming circuit. IEEE Trans on Circuits and Systems, 1986, 33, 533~ 541
4 Wang L X and Mendel J M. Adaptive minimum prediction-error deconvoIution and source wavelet estima-tion using Hop field neural networks. Geophysics, 1992, 57 (5): 670~679.
5 Langan R T Lerche I and Cutler R T. Tracing of rays through heterogeneous media: An accurate and effi-cient procedure. GeoPhysics, 1985, 50 (9): 1456~ 1465
2 Hopfield J J and Tank D W. Neural computation of decisions in optimization problems. Biological Cybernetcis, 1985, 52, 141~ 152
3 Tank, D W and Hop field J J. Simple "neural"optimization network: An A/D converter, signal decisioncircuit, and a linear programming circuit. IEEE Trans on Circuits and Systems, 1986, 33, 533~ 541
4 Wang L X and Mendel J M. Adaptive minimum prediction-error deconvoIution and source wavelet estima-tion using Hop field neural networks. Geophysics, 1992, 57 (5): 670~679.
5 Langan R T Lerche I and Cutler R T. Tracing of rays through heterogeneous media: An accurate and effi-cient procedure. GeoPhysics, 1985, 50 (9): 1456~ 1465
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心