Petroleum Engineering Department, Australian College of Kuwait, West Mishref, Kuwait.
Fouman Faculty of Engineering, College of Engineering, University of Tehran, Fouman, Iran.
Sci Rep. 2021 Mar 29;11(1):7033. doi: 10.1038/s41598-021-86264-5.
The present study evaluates the drilling fluid density of oil fields at enhanced temperatures and pressures. The main objective of this work is to introduce a set of modeling and experimental techniques for forecasting the drilling fluid density via various intelligent models. Three models were assessed, including PSO-LSSVM, ICA-LSSVM, and GA-LSSVM. The PSO-LSSVM technique outperformed the other models in light of the smallest deviation factor, reflecting the responses of the largest accuracy. The experimental and modeled regression diagrams of the coefficient of determination (R) were plotted. In the GA-LSSVM approach, R was calculated to be 0.998, 0.996 and 0.996 for the training, testing and validation datasets, respectively. R was obtained to be 0.999, 0.999 and 0.998 for the training, testing and validation datasets, respectively, in the ICA-LSSVM approach. Finally, it was found to be 0.999, 0.999 and 0.999 for the training, testing and validation datasets in the PSO-LSSVM method, respectively. In addition, a sensitivity analysis was performed to explore the impacts of several variables. It was observed that the initial density had the largest impact on the drilling fluid density, yielding a 0.98 relevancy factor.
本研究评估了高温高压油田的钻井液密度。这项工作的主要目的是通过各种智能模型引入一套用于预测钻井液密度的建模和实验技术。评估了三种模型，包括 PSO-LSSVM、ICA-LSSVM 和 GA-LSSVM。PSO-LSSVM 技术的偏差因子最小，反映了最高的响应精度，因此表现优于其他模型。绘制了实验和建模的决定系数 (R) 回归图。在 GA-LSSVM 方法中，训练、测试和验证数据集的 R 分别计算为 0.998、0.996 和 0.996。在 ICA-LSSVM 方法中，训练、测试和验证数据集的 R 分别为 0.999、0.999 和 0.998。最后，在 PSO-LSSVM 方法中，训练、测试和验证数据集的 R 分别为 0.999、0.999 和 0.999。此外，还进行了敏感性分析以探索几个变量的影响。结果表明，初始密度对钻井液密度的影响最大，相关性因子为 0.98。
