Reliability-Based Scheduled Maintenance (SM) for Mining Equipment with Artificial Neural Network (ANN) Model

Prompt and proper maintenance management helps extend the operation lifespan of workplace equipment to achieve production targets without interrupting the production process.In this connection, accurate prediction of the reliability-based scheduled maintenance (SM) time intervals of equipment is essential.The current research aimed to develop a reliability-based model to forecast the maintenance time intervals specifically for Load-Haul-Dumper (LHD) underground mining equipment.The series configuration system of the Reliability Block Diagram (RBD) model was adopted to evaluate the overall system reliability for each LHD machine.The reliability percentage of each sub-system was ascertained through a reliability analysis of a complex repairable system.To build the required Artificial Neural Network (ANN) model for analysis, the "Isograph Reliability Workbench 13.0" software was adopted to estimate the input layers of reliability (R) and the best-fit distribution parameters, such as the scale parameter (), shape parameter (), and location parameter ().The ANN model created was trained using the Levenberg-Marquardt (LM) learning algorithm.The predicted SM values were extremely close to the calculated values as indicated by the optimal R 2 value of 0.9998.The outcome demonstrated that the ANN model could improve the performance of the equipment with a major impact on the initial weight optimization.Suggestions were made for the industry practitioners to enhance the dependability of the equipment with planned maintenance procedures designed by the proposed ANN, with possible potential to be explored by other equipment users.