Prediction model of CBM content based on logging data and optimized general vector machine

In order to achieve high-precision evaluation of CBM gas content, reasonable production layout and efficient exploration and development, taking the gas content of No.3 coal seam in Shizhuang south area of Qinshui Coalfield as a sample, an improved quantum particle swarm optimization general vector machine hybrid prediction model（ＥＮ－ＩＱＰＳＯ－ＧＶＭ） based on elastic network optimization logging curve is proposed in this paper. Firstly, based on the response characteristics and sensitivity analysis of coal seam gas content logging, the Elastic Network（ＥＮ） is used to optimize the feature input parameters of the general vector machine model. Then, an improved quantum particle swarm optimization（ＩＱＰＳＯ） algorithm is proposed to optimize the GVM network weight threshold, and a general vector machine coal seam gas content prediction model based on elastic network and improved quantum particle swarm optimization algorithm is constructed. Finally, the model is used to predict the coal seam gas content of blind wells in the target area, and compared with the support vector machine, BP neural network model and the traditional multiple regression model under the same optimization strategy to analyze the performance and adaptability of the model. The results show that the blind well prediction accuracy of the new model is reduced from 21.83% to 4.25%, which has stronger generalization ability and is more suitable for the high-precision evaluation of gas content in highly heterogeneous coal reservoirs, and can lay a geological foundation for the exploration and development of coalbed methane target areas.