Twin Support Vector Regression for Prediction of Natural Gas Hydrate Formation Conditions

Abstract

Natural gas hydrates have become a threat to natural gas companies and oil industries for the formation of gas hydrates in transfer pipelines can lead to pipe blockage. Natural gas hydrate formation is favored by low temperature and high pressure; thus, if the right combination of temperature and pressure are well investigated, it is possible to solve the pipe blockage problems. Therefore, developing a precise, easy-to-use method to predict natural gas hydrate formation temperature (HFT) is very important. In this paper, we are intended to develop a novel machine learning model called twin support vector regression (TSVR) for predicting HFT in a wide range of natural gas mixtures. For constructing the TSVR model, 513 experimental data in the system methane, ethane, propane, butane, pentane, nitrogen, carbon dioxide/water, methane, carbon dioxide, propane, methanol, sodium chloride, and calcium chloride have been collected from open literature. We compared the performance of the TSVR model with least squares support vector regression (LSSVR) and three widely used correlations in the system methane, ethane, propane, butane, pentane, nitrogen, and carbon dioxide, which show that the TSVR model has fewer deviations than the LSSVR model and three correlations. The values of root-mean-square error (RMSE), mean absolute percentage error (MAPE), and correlation factor (R 2) are 2.4010, 0.0061, and 0.9108, respectively. In addition, statistical parameters show that the TSVR model can also surpass the LSSVR model in predicting HFT in the system water, methane, carbon dioxide, propane, methanol, sodium chloride, and calcium chloride.

Enhanced coalbed methane well production prediction framework utilizing the CNN-BL-MHA approach

“…Nonetheless, its applicability becomes limited in the context of Guizhou’s highly variable coal reservoirs due to their complexity. On the other hand, the mathematical statistics-based methods, specifically including multiple linear regression and support vector regression, facilitate predictions by aligning field data with trend lines 15 , 18 – 20 . These methods demand less extensive data and operate with simpler models.…”

Machine learning models for fast selection of amino acids as green thermodynamic inhibitors for natural gas hydrate

“…Statistical thermodynamic models have been developed to predict the gas hydrate phase equilibrium [13][14][15][16][17][18]. These models often require many parameters and need to find proper values for these parameters [19,20]. Moreover, thermodynamic models often assume a specific form of mathematical equation and statistical regression is used to determine the unknown parameters, but it is hard to assume an empirical equation due to the lack knowledge on the relationship between amino acids and HFT.…”

“…It can be used to distinguish the given data based on their different patterns, extract useful information from the data and output predictions with new input variables [21]. Recently, a number of ML algorithms such as artificial neural networks, decision trees, k-nearest neighbor, gradient boosting regression, adaptive neuro Fuzzy interference system, gene expression programming, random forest (RF), support vector machines (SVM) have been used to predict gas hydrate formation and dissociation [19,20,[22][23][24][25][26][27][28][29]. These studies mainly focused on the hydrate phase transition in aqueous solutions with or without traditional hydrate inhibitor such as monoethylene glycol, Luvicap 55 W and ionic liquids which might have negative environmental impacts as aforementioned.…”