Channel: TECHNOLOGY https://www.ljgczz.com EN-US https://www.ljgczz.com/EN/current.shtml https://www.ljgczz.com 5 r² exceeding 0.7. Piecewise function fitting based on ridge regression and nonlinear regression model y=x/锛�b+kx锛� are suitable for non-linear hydrocarbon loss restoration. Compared with the traditional method of hydrocarbon loss restoration, using the quantitative method to restore hydrocarbons in the study area is more scientific and comprehensive, and has a broad prospect of application.]]> 2/C₃銆侊紙C₁+C₂锛�/锛圕₃+C₄+C₅锛夈�丆₂/锛圕₃+C₄锛夈�丆₁/C₃銆侊紙C₁+C₂+C₃锛�/锛圕₄+C₅锛�, a total of 5 parameters are optimized to form a radar chart. The multiparameter gas-oil ratio and density prediction models established from the geometric model have been verified and applied in 14 layers of 5 wells. The average prediction errors for gas-oil ratio and density are 11.55% and 4.71%, indicating that the established models can realize the rapid and accurate classification of hydrocarbon reservoir types,and the application effect is obvious.]]> TOC (Total Organic Carbon) to study the influencing factors and organic matter enrichment patterns of organic matter enrichment. The results show that the main influencing factors of organic matter enrichment are paleosedimentary environment, paleoproductivity, paleotergenous input, etc., and the enrichment of organic matter is positively correlated with phosphorus, zirconium and other elements. It is believed that the basic conditions for the accumulation of organic matter in this area are strong paleoproductivity and rapid sedimentation of terrigenous inputs, and an oxygen-poor reducing environment is formed under the condition of good overlying cover. Based on the logging elements, the machine learning algorithm is used to obtain the organic carbon content while drilling, and the difference in the distribution of organic matter in the trajectory direction of horizontal wells is explored, which provides more abundant materials for deepening the understanding of shale gas accumulation in this area.]]>