Prediction of rock penetration rate using a novel antlion optimized ANN and statistical modelling

Abstract

The prediction of penetration rate (PR) is a complex task which relies on several variables such as nature of rock and rock mass properties as well as operating parameters of the drilling equipment. In the recent times, different studies have used soft computing approaches such as neural network and fuzzy modelling for these kinds of complex problems. However, the performance evaluation and comparative analyses of different optimization tools with the traditional multiple regression to predict the PR of granite is yet to be examined. In this study, antlion optimized ANN (ALO-ANN), ordinary artificial neural network (ANN), multiple linear statistical model (MLSM) and multiple non-linear statistical model (MNLSM) were used to predict the PR with density (ρ), porosity (μ), and point load index (Is(50)) as input parameters. The performance of the proposed models were compared using statistical indicators such as the coefficient of determination (CoD), average absolute percentage error (AAE), mean absolute error (MAE), and root-mean squared error (RMSE). The result shows that ALO-ANN provides a better predictive model and proves the effectiveness, robustness, and reliability of the ALO-ANN model over other predictive models. Sensitivity analysis was performed and it was found that ρ and Is(50) have more influence on the PR than the μ. This study has proved the application of a new and hybrid artificial intelligent method (ALO-ANN) in predicting the PR of granite rock to enhance drilling and blasting designs.