Optimisation of a real-time multi-period truck dispatching system in mining operations

Abstract

In recent years, there has been considerable research on optimisation of long-term and short-term open pit mine plans to maximise the net present value of a given mining project. The typical goal of truck allocation models in the literature is either minimisation of the number of trucks required to pursue the production target for a given operational period or maximisation of the production based on the available trucks and shovels. With this framework, there is no guarantee of meeting the main goal of a mining project, which is maximisation of the net present value. The objective of this research is to develop a linear programming model for truck allocation, and to utilise it in conjunction with the existing mixed integer programming model for short and long-term mine planning. We test a linear programming model and run it with a mixed integer-programming model concurrently and interactively for solving truck allocation problem. The proposed approach of truck allocation overcomes the shortcomings of the existing models by taking into account economic parameters, multi time periods and the uncertainty of load and travel times and ore grades. We provide a case study of McLaughlin gold mine to show the implementation of our model.