CAJ | 학술논문

目前采用地震属性预测储层参数的方法层出不穷，但是这些方法多数是基于单变量、线性的机器学习算法，在已知样本较少的情况下精度得不到保证。为了获取高精度的储层参数，指导油气的勘探开发，迫切需要寻求一种新的方法最大限度地挖掘地震地质信息。支持向量机是以结构风险最小化原则为核心的新型机器学习算法，与传统的机器学习算法相比，其具有基于多变量、小样本、非线性和预测精度高的优点。以渤海湾SZ-361油田Ⅰ油组顶部储层参数预测为例，采用支持向量机算法，得到了较高精度的储层预测结果，证实了支持向量机算法可以应用于油气勘探领域。
목전채용지진속성예측저층삼수적방법층출불궁，단시저사방법다수시기우단변량、선성적궤기학습산법，재이지양본교소적정황하정도득불도보증。위료획취고정도적저층삼수，지도유기적감탐개발，박절수요심구일충신적방법최대한도지알굴지진지질신식。지지향량궤시이결구풍험최소화원칙위핵심적신형궤기학습산법，여전통적궤기학습산법상비，기구유기우다변량、소양본、비선성화예측정도고적우점。이발해만SZ-361유전Ⅰ유조정부저층삼수예측위례，채용지지향량궤산법，득도료교고정도적저층예측결과，증실료지지향량궤산법가이응용우유기감탐영역。
The methods of predicting reservoir parameters based on the seismic attributes emerge in an endless stream at present .However , these methods are almost based on single variable and linear algorithm ,and the accuracy could not be ensured because there are not enough samples .In order to obtain reservoir parameters with high accuracy which are used to guide the exploration and development of oil and gas , it is necessary to find a new method to get maximum seismic and geological information .SVM is a new machine learning algorithm which is based on the principle of minimum structural risk .Compared with traditional machine learning algorithm ,SVM have several advantages such as multi-variable- based ,small sample ,nonlinear and high accuracy of prediction .This paper presents an example of predicting reser-voir parameters at the top ofⅠoil group in SZ-361 oilfield in Bohai bay and obtains a more rigorous result by SVM .This research indicated that SVM can be applied to the exploration and development of oil and gas .