【摘要】 本文认为，从地层的真实映象到我们获得的地震资料是一种非线性映射。对地震资料的种种处理实际上是对这种非线性映射的“逆过程”的模拟和逼近。本文提出了一种用神经网络解决这种“模拟和逼近”的新方法。利用神经网络能较好地解决非线性问题和柔性的特点，以得到真正的“高分辨率”和“高信噪比”的输出资料。通过对一组人工资料的处理，表明神经网络方法能很好地解决“反褶积”处理问题。更多还原

【Abstract】 In this paper, the seismic records are regarded as the nonlinear mapping of the real geophysics condition. The seismic signal processing is the simulation of the inverse mapping of the nonlinear mapping. We present a deconvolution method to solve this problem with ANN in order to get the high resolution and high siganl-noise ratio because the nonlinear problem can be solved better with ANN. The results show that this method is better than the conventional deconvolution methods for some artifical noisy seismic records.更多还原