基于云变换的混合计算模型在水淹层识别中的应用

Application of Hybrid Computing Neural Network Model in Water Flooded Layer Recognition Based on Reverse Cloud Transformation

目前为止,现有的自动判别方法难以反映定量指标和定性指标相结合的混合信息对水淹层识别的影响.因此,为提高水淹层判别的准确度,本文提出基于云变换的定量与定性混合计算神经网络模型来实现水淹层判别.一方面,利用云模型将提取测井相数据中的定性信息,保证了原始数据的完整性与客观性;另一方面,将输入信息中的定性概念通过正向标准云变换转换为量化的数值信息,保证了数据的科学性;最终将混合信息输入混合计算神经网络模型中进行判别,从而得出结论.实验证明采用基于云变换的混合计算神经网络模型对水淹层进行识别,具有精度高、速度快的特点,是水淹层识别的一种比较实用的方法.

So far, the existing method for automatically discriminating information is difficult to reflect the impact of mixing quantitative and qualitative indicators of the combination of the water layer recognition. Therefore, in order to improve the accuracy of determining flooded layer, this paper proposes neural network model to calculate quantitative and qualitative transformation hybrid cloud-based discrimination to achieve flooded layer. On the one hand, qualitative information is extracted logs by cloud model, to ensure the integrity and objectivity of the original data; on the other hand, the information in the qualitative concept forward through the normal cloud is converted to quantized transform numerical information, ensures the scientific data; the eventual establishment of correspondence between the characteristics of the system input and results. Experimental results show the high accuracy of the calculation method for water layer recognition neural network based on hybrid cloud transformation, it has the characteristics of fast, flooded layer identification is a more practical approach.