论文部分内容阅读
夏克文胡兴卫等：用神经网络识别水泥胶结质量，测井技术，１９９６（３）２０，２０７～２０９。采用双发双收补偿声系可剔除常规水泥胶结测井（ＣＢＬ）曲线的误差。对于水泥胶结质量的识别，人工神经网络优于统计方法。采用带有非线性连接权的二层前馈神经网络能实时完成水泥胶结质量的识别。
Xia Kewen Hu Xingwei et al .: Neural network to identify the quality of cement bonding, logging technology, 1996 (3) 20,207 ~ The use of double double compensation sound system can eliminate the error of conventional cement cementing logging (CBL) curve. Artificial neural network is superior to statistical methods for cement bond quality identification. The two-layer feedforward neural network with nonlinear connection right can be used to identify the cement bond quality in real time.