【摘要】 岩石可钻性是钻头选型和提高钻井速度的基础,利用石油测井资料计算岩石的可钻性是一种行之有效的途径。常规的岩石可钻性测井预测模型都是基于回归分析而建立起来的,形式简单,精度不高。阐述了自适应模糊推理系统(Adaptive Network-based Fuzzy Inference System,AN-FIS)网络,提出了岩石可钻性测井模型的ANFIS网络预测模型,选取与岩石可钻性密切相关的参数对ANFIS网络进行训练。同BP神经网络模型比较,此模型具有更高的准确性。更多还原

【Abstract】 Rock drillability is the basis for selecting the drill-bit and enhancing the drilling velocity.It is a feasible and effective approach to calculate rock drillability from well-log data.The traditional prediction model of rock drillability is based on regression analysis and the accuracy is not so good for practical application.This paper introduces the adaptive network-based fuzzy inference system(ANFIS),and then presents a forecast model of ANFIS network.We select parameters related to rock drillability,the network is been trained.The results show that the forecast model of ANFIS is more accurate than the forecast model of BP.更多还原