Machine learning with SHapley additive exPlanations for evaluating mine truck productivity under real-site weather conditions at varying temporal resolutions

Abstract

ABSTRACT The updated abstract, shortened to about 100 words, is shown below: This study built truck productivity prediction models incorporating real-site weather conditions at varying temporal resolutions. The best models were combined with SHapley Additive exPlanations to offer quantitative and qualitative analysis for input variables’ effect on the model outputs. The results showed that mining engineers can make more accurate predictions of truck productivity at the weekly resolution compared with other resolutions. The three most influential input parameters were haul distance, empty speed, and ambient temperature. Extreme weather, such as strong wind speed, heavy precipitation, and extreme relative humidity, had a certain effect on truck-shovel allocation. Meanwhile, a unified graphical user interface was developed to predict mine truck productivity.