Optimization Modeling and Analysis of Truck Allocation System in Surface Mining Operations
In the surface mining industry, trucks and shovels are the essential components in the whole operation system. Optimal allocation of trucks/shovels resources is critically necessary in the surface mining operations not only for satisfying the mining production profit goal, but also for minimizing the mining operation cost. Also, many factors in the operation process make the truck allocation decision making plagued with uncertainties, vagueness and complication. In the past, Interval-parameter Linear Programming (ILP) has been widely used to deal with uncertainties and to assist optimal decision making in a variety of system planning and management problems. However, the existing ILP solution algorithms, i.e., best-worst case algorithm and 2-step algorithm, are found to be ineffective through a validity checking process. Moreover, the results from ILP cannot reflect the linkage between decision risks and the system return. In this study, a Risk Explicit Interval-Parameter Linear Programming (REILP) model and a fuzzy-REILP model are developed to generate the least cost strategies while minimizing the decision risks. The developed methods are then applied to the optimal planning of the truck allocation system in an open-pit mine case, and this is the first attempt of using ILP-based optimization techniques to the surface mining industry. This method is specifically designed to deal with extensive uncertainties existed in the truck allocation system and to provide decision supports to the surface mining operators and planners. In the developed methods, the ILP is used to reflect uncertainties existed in both objective function and constraints. Based on the basic ILP, a risk function is defined to assist in finding solutions with minimum system cost while minimizing the decision risk, under certain aspiration levels. The aspiration level could be conservative, medium or aggressive, and can thus be presented as a fuzzy set to reflect the preference of decision makers. Three sets of solutions are obtained accordingly. Besides, the model was also solved under the aspiration level from 0 to 1, with a step of 0.1, for providing a comprehensive decision support. This approach can effectively reflect dynamic, interactive, uncertain characteristics, as well as the interactions between overall cost and risk level of the mining truck allocation system. The results can effectively reflect the tradeoff between decision risks and the system return, and thus provide valuable information to support the decision-making process related to the planning of the truck allocation, and timing and routing of the mine-hauling activities.