复杂岩性模式识别中测井项目优化选择技术

在利用模式聚类判别方法建立复杂岩性集合判别函数时,需要优选出对岩性集合敏感的测井项目,以有效克服所有测井项目均参与而造成的相互掣肘、矛盾现象。依据构建集合内均匀度参数及集合间距离参数,实现复杂岩性集合判别中测井项目优化选择技术。以海拉尔盆地的4大类岩性集合(一般沉积岩、火山碎屑岩、火山碎屑沉积岩、火山熔岩等)为应用实例,优选后判别符合率明显改善,说明了方法的有效性。

The optimal selection technology of the well logging program of the complex lithology pattern identification

When building the discriminant function of complex lithology set by using the method of pattern cluster discrimination, we need to select the well logging programs which are sensitive to the lithology sets optimally, and then we can overcome the impede and contradiction because of the participation of all the well logging programs efficiently. This article gets the optimal selection technology of well logging programs in the discrimination of the complex lithology sets, which based on building the parameter of the evenness and the distance between the sets in the aggregation. Based on the 4 broad headings of Hailaer basin (general sedimentary rock, the volcanoclastic rock, the volcanoclastic sedimentary rock, lava and etc.), the discrimination coincidence rate is better apparently, and this confirms that the method is effective.