【摘要】 测井解释的一项主要任务是区分油气水层 ,目前采用的方法主要是基于统计学理论的经验公式 ,因而或多或少存在一些不足。而人工神经网络方法具有高度自学、自适应和抗干扰等特点 ,能够有效地区分油水层。文章根据油水层物理特性 ,进行了BP算法改进。实际应用表明 ,其识别率远高于传统方法 ,效果令人满意。更多还原

【Abstract】 An important task of well logging interpretation is to determine the oil & water formation. The present methods are based on statistics theory,and have many shortcomings. The method of neural network has the characteristics of high self study,self adaptation and interference resisting,and can determine the oil & water formation. According to the physical characteristics of oil & water formation, the back propagation algorithm(BP)is improved in this paper. Actual application shows that the coincident ratio is much higher than that of the statistics method and the interpretation effect is satisfactory.更多还原