Research on Formation Pressure Prediction Based on Neural Network System Identification Theory

Abstract

The traditional formation pressure calculation mostly uses the system identification method of empirical formula and fitting formula. It can not get satisfactory identification results. In this paper, neural network is applied to system identification. A neural network identification model and improved error back propagation algorithm are researched. It is applied to the formation pressure prediction of injection production system. The process of system identification is the process of directly learning the input and output data of the system. The experimental results show that the neural network is effective in system identification.