Application of machine learning to determine the shear stress and filtration loss properties of nano-based drilling fluid

Abstract A detailed understanding of the drilling fluid rheology and filtration properties is essential to assuring reduced loss during transport process. As per literature review, silica nanoparticle an exceptional additive enhance enhancement. However, a correlation based on nano-SiO 2 -water-based that can quantify nanofluids not available. Thus, two data-driven machine learning approaches are proposed for prediction, i.e. artificial-neural-network least-square-support-vector-machine (LSSVM). Parameters involved prediction shear stress SiO concentration, temperature, rate, whereas time inputs simulate volume. feed-forward multilayer perceptron constructed optimised using Levenberg–Marquardt algorithm. The parameters LSSVM Couple Simulated Annealing. performance each model evaluated several statistical parameters. predicted results achieved R (coefficient determination) value higher than 0.99 MAE (mean absolute error) MAPE percentage below 7% both models. developed models further validated with experimental data reveals excellent agreement between data.