Estimating Pressure-Volume-Temperature Properties of Crude Oil Systems Using Boosted Decision Tree Regression

نویسندگان

چکیده

貯留層流体特性の推定において,機械学習方式の性能は相関関係に基づく従来方式に匹敵する。機械学習方式のなかで,人工ニューラルネットワークモデルが最も一般的に使用されているモデルの一つであるが,ブラックボックス問題なる課題を有することが知られている。そこで本研究では,この課題を解決すべく新たなモデル化方式を検討する。正確な推定を実現すると同時に,推定結果に対する入力パラメーターとその重要度に関する考察を可能とするため,提案方式ではブースト決定木回帰(BDTR)を用いて,石油とガスの比重,石油と溶解ガスの比,貯留層温度の関数として,沸点圧力(Pb)ならびに沸点圧力における石油容積係数(Bob)を推定する。構築されたBDTRモデルは,既存の機械学習モデルや最も一般的に使用されている相関関係に基づく従来手法と比べて,PbおよびBobをより高精度に推定可能である。特にPbの推定においては,欠損値補完の前処理を施すことにより,複数の入力特徴量が欠損しているデータを用いた評価においても相関関係に基づく既存方式よりも優れた有効性が示された。これは提案方式の優れた予測性能および汎用性を示すものである。

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Modeling of Pressure Dependence of Interfacial Tension Behaviors of Supercritical CO2 + Crude Oil Systems Using a Basic Parachor Expression

Parachor based expressions (basic and mechanistic) are often used to model the experimentally observed pressure dependence of interfacial tension (IFT) behaviors of complex supercritical carbon dioxide (sc-CO2) and crude oil mixtures at elevated temperatures. However, such modeling requires various input data (e.g. compositions and densities of the equilibrium liquid and vapor phases, and molec...

متن کامل

Prediction of Plant Species Boibiversity using Generalized Linear Model (GLM) and Boosted Regression Tree (BRT) in Eastern Rangelands of Mazandaran

Abstract Background and objectives: Prediction of species richness and diversity patterns are used to develop conservation strategies for biodiversity under regional and global environmental changes. Since modeling the distribution of plant species can provide useful and important information about identifying and introducing potential habitats of plant species, and also few studies have been ...

متن کامل

Boosted Decision Tree for Q-matrix Refinement

In recent years, substantial improvements were obtained in the effectiveness of data driven algorithms to validate the mapping of items to skills, or the Q-matrix. In the current study we use ensemble algorithms on top of existing Qmatrix refinement algorithms to improve their performance. We combine the boosting technique with a decision tree. The results show that the improvements from both t...

متن کامل

Predicting volume of distribution with decision tree-based regression methods using predicted tissue:plasma partition coefficients

BACKGROUND Volume of distribution is an important pharmacokinetic property that indicates the extent of a drug's distribution in the body tissues. This paper addresses the problem of how to estimate the apparent volume of distribution at steady state (Vss) of chemical compounds in the human body using decision tree-based regression methods from the area of data mining (or machine learning). Hen...

متن کامل

Prediction of the physicochemical properties of quince puree during thermal treatment using M5 decision tree

Thermal pasteurization is known as the most common method in food processing. In this study, the effectiveness of come up time on inactivation of heat resistance isoform of polyphenol oxidase (PPO), as a pasteurization index, was investigated during heating of the quince puree in a water bath at 60 to 90 . M5 decision tree, a novel statistical method, was applied to understand the simple relati...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Journal of The Japan Petroleum Institute

سال: 2022

ISSN: ['1346-8804', '1349-273X']

DOI: https://doi.org/10.1627/jpi.65.221