摘要
油气田的模式识别方法包括地震多参数统计模式识别方法和神经网络方法，它可提取和应用自相关、自回归、功率谱和均方根振幅等多种地震信息特征参数，并用这些地震参数的两种模式识别方法，对有井和无井地区进行油气预测。对南海神孤鼻状构造带进行油气预测，在4个局部构造上发现好的特征异常显示，并且它们与构造圈闭及砂岩分布基本吻合。
The pattern recognition system for oil and gas fields contains seismic multiparameter statistic pattern recognition and artificial nervous network. With the software system, many kinds of seismic feature parameters based on the theories of the autocorrelation, autorecur, power-spectrum and mean square root amplitude are obtained and used. With these seismic parameters and the two methods of pattern recognition as stated above, one can make prededictions for oil and gas fields in areas with or without wells drilled. Using this procedure, four local structures in Shenhu Nose Structure Zone have been predicted with good feature anomalies, which are consistent with the structure closures and the sand distribution.
出处
《中国海上油气（地质）》
1994年第2期121-129,共9页
China Offshore Oil and Gas(Geology)
