工程数字测井自动分层的人工智能技术

根据人工对测井资料分析的思维过程，应用模糊数学理论及统计学参数，建立了岩土性质识别及界面划分的人工智能数学模型，并研制了相应的软件，实现了工程数字测井的自动分层，对离散的数字测井值，通过测井物理参数对地层性质的隶属函数模型，逐一识别，然后进行聚类处理，对地层界面处由于邻层效应产生的模糊数据，分别计算其对上下层岩性的隶属度，采用主因素决定型算法实现界面划分，应用结果表明，该方法简单，快速且较准确。

ARTIFICIAL INTELLIGENCE OF THE LITHOLOGIC AUTO-RECOGNITION FROM DIGITAL WELL-LOGGING DATA IN ENGINEERING

A mathematical model of recognizing lithology and determing lithologic boundary was established,and the software was also developed.According to the method of artificial analysis of the well logging data,and the theory of fuzzy mathematics and statistics parameters,The auto recognizing lithologic formation was realized with a computer.Using the attaching function model of geophysical parameters with the properties of stratum,the discrete data of well logging are recognized datum by datum.The same property data are gathered together as an identical layer.To the fuzziness data of the adjacent boundary caused by the adjacent layers,the values of attaching to the upper and lower layers are calculated,and the boundary is determined by the main factor algorithm. The application results indicate that the method is simple,fast and accurate.