The use of machine learning (ML) is commonplace in the petroleum industry and covers practically all aspects from upstream to downstream. Prediction of measured reservoir fluid properties has extensively benefited from this technology and numerous papers reported the use of ML to predict different properties; e.g., saturation pressure, oil formation factor, solution gas/oil ratio and viscosity.

中文翻译：

---

使用机器学习对碳氢化合物流体压力-体积-温度实验室数据进行整体预测

机器学习（ML）的使用在石油行业很常见，几乎涵盖了从上游到下游的各个方面。测量储层流体性质的预测已广泛受益于该技术，并且许多论文报道了使用机器学习来预测不同的性质；例如，饱和压力、油形成因子、溶解气/油比和粘度。