油气运移和聚集的人工神经网络模拟

2001年 23卷 第2期
阅读:131
查看详情
ARTIFICIAL NEURAL NETWORK SIMULATION ON HYDROCARBON MIGRATION AND ACCUMULATION
吴冲龙 刘海滨 毛小平 李绍虎 王燮培 吴景富 何大为 张云飞 潘明太
WU Chong-long LIU Hai-bin MAO Xiao-ping LI Shao-hu WANG Xie-pei WU JING fu HE Da-wei ZHANG Yun-fei PAN Ming-tai
中国地质大学, 湖北武汉 430074 2. 中国海洋石油研究中心, 河北高碑店 074010
China University of Geosciences, Wuhan, Hubei 430074 China 2. China National Offshore Petroleum Research Center, Gaobeidian, Hebei 074010, China
盆地演化、油气系统演化以及油气的运移聚集充满了混沌与非线性特征,单纯使用传统的地下流体动力学方程,无法实现油气运聚的模拟和评价。作者探讨了将传统动力学模拟与人工神经网络模拟结合起来的的途径与方法,即在三维构造地层体的动态模拟基础上,采用单元体模型使非均质的复杂通道体系转化为有限个简单均质体后,再利用传统动力学模拟来对相态和驱动力求解,然后运用人工神经网络技术来解决单元体之间的油气运移方向、运移速率和运移量等问题。利用所编制的软件对珠三凹陷的油气二次运移和聚集进行了动态模拟,有效地揭示油气运聚的复杂机理和过程。
There are full of chaos and nonlinear characters in basin evolution, petroleum system evolution and hydrocarbon migration and accumulation, so it is impossible to simulate and evaluate the course of hydrocarbon migration and accumulation by simple using the traditional underground hydrodynamic equation. In this paper, the authors discuss the approaches and methods of how to combine the traditional dynamic simulation and artificial neural network simulation; that is, based on the dynamic simulation of the three dimension structure-sedimentary modeling, the authors use the unit entity model to alter the heterogeneous complex passage system into limited simple homogeneous body, calculate the facies and the drive forces of hydrocarbon through the traditional dynamic simulation, and solve such problems as the direction, velocity and quantity of hydrocarbon migration between the unit body by using the artificial neural network technology. A simulation software had been developed under this way. This software had been used to simulate hydrocarbon migration and accumulation in the Zhusan Depression. We got a good result which could well open out the complex mechanism and course of the secondary hydrocarbon migration and accumulation.
盆地演化; 油气系统; 油气运移; 油气聚集; 非线性过程; 人工智能模拟; 人工神经网络系统;
basin evolution; petroleum system; hydrocarbon migration; hydrocarbon accumulation; nonlinear course; artificial intelligence simulation; artificial neural network system;
国家自然科学基金“九五”重点项目(49732005)
https://doi.org/10.11781/sysydz200102203