论文部分内容阅读
将遗传算法、Fisher最优分割和边缘检测算法相结合,设计了一种适用于多参数测井地层划分的遗传最优分割算法.该算法以测井测点号为基因,以有序测点号组为染色体,由测点的边缘隶属度和段间/段内离差总和的比值共同构成综合优化指标(适应度),采用锦标赛选择机制,通过MonteCarlo抽样进行交叉(通过两两染色体比较设定抽样区间),采用灾变变异和精英保持策略,利用多参数测井数据求取最优化地层分界线.在济阳坳陷中生界地层划分中的实际应用,表明该方法的自动分层结果符合地质实际,且计算速度很快.该方法对于没有露头、岩心稀少、地震勘探分辨率低和地球物理测井数据量大情况下的深部隐伏地层划分有很好的适用性.
The genetic algorithm, Fisher optimal segmentation and edge detection algorithm are combined to design a genetic algorithm for multi-parameter logging stratigraphic division of the optimal genetic algorithm. The number group is a chromosome, which is composed of the ratio of the edge membership of the measuring point and the sum of the segment / segment deviations to form a comprehensive optimization index (fitness), and is cross-checked by a Monte Carlo sampling using a tournament selection mechanism Set the sampling interval), using the catastrophic variation and elitist retention strategy, using multi-parameter logging data to get the optimal stratigraphic boundary.The practical application of the Mesozoic stratigraphic division in the Jiyang depression shows that the method of automatic stratification Which is in accordance with the geological conditions and has a fast calculation speed.This method has good applicability for the deep buried stratigraphic subdivision under the conditions of low outburst, low core resolution, low resolution of seismic exploration and large amount of geophysical logging data.