Intelligent Isochronal Test Analysis

摘要

An artificial neural network (ANN) is a powerful tool in data processing and pattern recognition that mimics a biological neural network. A new approach of isochronal test analysis based on ANNs is presented. The results from ANNs are in good agreement with the results from empirically achieved deliverability curve and pre-required characteristics of Horner and drawdown plots. It is shown that application of ANN in isochronal tests provides a novel approach to speed up calculations and escape tedious and time consuming plots. Also details presented can be used to automate other types of well test. Therefore ANN can perform as a robust tool in well test analysis.