A 0-1 GOAL PROGRAMMING-MODEL FOR SCHEDULING MULTIPLE MAINTENANCE PROJECTS AT A COPPER MINE

This paper describes the application of 0-1 goal programming (GP) in multi-project resource-constrained scheduling for the maintenance of mineral-processing equipment at a copper mine in China. This involves four projects, 40 jobs and nine types of resources. We first develop a generalized 0-1 GP model. Because of the limited computing capacity at the mine, an effective two-phase hybrid solution approach is then proposed. In Phase I, under resource constraints, a feasible schedule is obtained by a heuristic. Based on this feasible schedule and a model-simplification procedure, many decision variables in the generalized model are eliminated and redundant constraints are removed. (Although this procedure is particularly designed for this mine, we also discuss its extension to general problems.) In Phase II, the streamlined model is solved by implicit enumeration. This approach eliminates 90% of the variables and 96% of the constraints in the problem we studied. We are thus able to solve a multi-project scheduling problem of medium size on a PC. In comparison with the heuristic method that the mine was using, our hybrid approach reduced project duration by 13%, total project cost by 30%, cost for the critical project by 15%, and overall workload by 59%.