Prediction and Analysis of Shale Gas Well Pressure based on Multimodal Elman

Abstract

In the actual shale gas exploitation process, there are strong nonlinear, multi-characteristic, time series problems among the production data, which make the hidden information among variables not fully expressed, resulting in low pressure prediction accuracy. Therefore, a prediction method based on multimodal Elman is proposed to analyze the pressure of shale gas wells. The main contributions are as follows: (1) constructing the maximum covariance feature set as the direction axis for projection, so as to maximize the degree of heterogeneity between the original data. (2) Based on this, a multimodal Elman model is constructed for prediction and analysis. Compared with Elman algorithm, it can obtain more potential variables, thus increasing the prediction accuracy of the model. The experimental results show that the application in shale gas production proves the effectiveness and feasibility of the proposed method.