An Advanced Long Short-Term Memory (LSTM) Neural Network Method for Predicting Rate of Penetration (ROP)
Hui Ji, Yishan Lou, Shuting Cheng, Zelong Xie, Liang Zhu
Abstract
of 0.978 and RMSE and MAPE are 0.287 and 12.862, respectively, hence overperforming the existing methods. The average accuracy of the optimized LSTM model is also improved by 44.2%, indicating that the prediction accuracy of the optimized model is higher. This proposed method can help to drill engineers and decision makers to better plan the drilling operation scheme and reduce the drilling cycle.
Topics & Concepts
Mean squared errorMean absolute percentage errorComputer scienceParticle swarm optimizationArtificial neural networkCorrelation coefficientPearson product-moment correlation coefficientRate of penetrationAlgorithmArtificial intelligenceDrillingMachine learningStatisticsMathematicsEngineeringMechanical engineeringDrilling and Well EngineeringHydraulic Fracturing and Reservoir AnalysisOil and Gas Production Techniques