Flowing bottomhole pressure prediction for gas wells based on support vector machine and random samples selection

Dynamic analysis and optimum production strategies of gas wells demand accurate prediction of flowing bottomhole pressure (FBHP). Due to the existence of many uncertain relations between the changeable influence factors and the limitations of existing methods, no single model was found to be applicable over all ranges of variables with suitable accuracy. In this paper, a FBHP prediction method based on support vector machine (SVM) and random samples selection way, named the FBHP-SVM method, was investigated, and a support vector regression model with ɛ-insensitive loss function (ɛ-SVR) based on radial basis function (RBF) was used to predict the FBHP. Compared with the true values, the average absolute and relative prediction errors were 0.20 MPa and 2.62%, respectively. It is worthy to note that a reliable prediction of FBHB can be made when the true value of verification data is in the true values range of training samples.