���ڻ�ɫGM(0��N)���IJ⾮Ԥ����ʯ�������о�

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。由于回归分析直接利用原始数据或对原始数据作了标准化处理、对数变换后进行建模,以致其模型精度不高。而灰色GM(0,N)模型通过对原始数据进行累加生成所得的灰色模块来建立模型,这在一定程度上弱化了原始数据的随机性。为此,文章提出了一种基于灰色GM(0,N)理论利用测井资料求取岩石可钻性的实用方法。该方法从测井信息与岩石可钻性的内在关系出发,选用与岩石可钻性密切相关的多个测井参数,通过灰色GM(0,N)技术建立岩石可钻性的预测模型。室内试验表明,该模型能进一步提高岩石可钻性的预测精度,能满足钻井工程的实际需要,可以很方便地建立某地区的岩石可钻性剖面并掌握其分布规律。将该法应用于SC油田Du4井的测井资料处理中,为邻井及时地提供了较为准确的地层岩石可钻性剖面,也为该区的新井钻头选型提供了科学依据。

Research on Rock Drillability Predicted by Log Data on the Basis of Gray[GM(0,N)]Algorithm

Conventional log prediction models of rock drillability were all set up on the basis of regression analysis. The model accuracy was small, because the modeling was carried out by applying directly the original data or applying the normalized and logarithemically transformed data in regression analysis. However, the gray[GM(0,N)]model canbe established by the gray sequence obtained through the cumulative generation of the original data, which weakens the randomness of the original data to a certain extent. A method of predicting rock drillability by log data in light of gray [GM(0,N)]theory is proposed in the paper. In cosideration of the internal relation between log information and rock drillability and through selecting several logging parameters closely related to the drillability, the log prediction model of rock drillability may be set up by applying the gray[GM(0, N)] technique. Through laboratory tests, it was indicated that,by the model the prediction accuracy of rock drillability could be further raised;the actual needs of drilling engineering could be met,and a rock drillability profile could be convenient to establish and its distribution law could be well grasped. This method was used for processing the log data of well Du-4 in SC oil field, thus providing a relatively accurate rock drillability profile for its adjacent wells and a scientific basis for the bit type selection of new wells in the field.