This study endeavored to develop an algorithm named Quality Constrained Scheduling of Mining Operations (QCSMO) for underground coal mining. This algorithm uses dynamic data structures and mixed linear-integer programming (MLIP) technique to develop computer software that generates a mining sequence such that the mining units and areas are scheduled within a given time horizon to meet both coal production and quality requirements. QCSMO guarantees that the scheduled sequence is optimal within each period and at least "near optimal" throughout the whole process, under the spatial and operational precedent constraints.
QCSMO dynamically builds MLIP models and uses LINDO TM, a linear optimization software, to optimally assign mining units to areas within each period. The final output of QCSMO can help the tasks of generating a mining plan, estimating the quality of production or evaluating a mining layout. The dynamic nature of QCSMO allows users to adjust production requirements at any period within the time horizon and gives them much flexibility to fit the needs in the real world.
The implementation of this algorithm required inter-language calls between C and FORTRAN routines. The method used by this research may be applied to other engineering fields where operational process control and optimization are needed and when productivity and quality are concerned simultaneously under various precedent constraints of activities.