论文部分内容阅读
流动单元的划分方法直接影响着储层性质的评价及油藏后期开发。针对常用的流动单元评价方法——流动带指标法容易出现聚类指标优劣交叉的问题,提出了基于模糊C均值聚类模型划分流动单元的方法。该方法通过分析多种储层属性,寻找各类的最优聚类中心,并给出样本点属于每一类的隶属度,更准确地标定出每个样品的聚类域。以克拉玛依油田五3中区克下组油藏为例,通过在聚类过程中有效地实施人工干预,解决了流动单元划分指标不匹配的问题,在生产实际中取得了较好的应用效果。
The division method of flow units has a direct impact on the evaluation of reservoir properties and later development of reservoirs. Aiming at the commonly used method of flow unit evaluation, the method of flow band index is prone to crossover, and a method of dividing flow units based on fuzzy C-means clustering model is proposed. The method searches for the optimal cluster centers by analyzing a variety of reservoir properties, and gives the membership degree of each sample belonging to the sample points. The clustering domain of each sample is more accurately calibrated. Taking Kexia Formation reservoir in the middle of Karamay Oilfield as an example, the problem of mismatch of flow unit division index was solved through effective human intervention in the clustering process, and a good application effect was obtained in production practice.