A multimine mineral value chain (MVC) consists of multiple mines in the upstream supplying raw materials to multiple destinations in the downstream. The short-term mine production scheduling problem of a multimine MVC aims to determine the optimal extraction sequence and destination allocation of blocks from all the mines collaboratively to meet the quality and quantity requirements of all the processing plants in each period, subject to all relevant technical and operational constraints of the mining system. It is carried out at shorter scales of days/weeks/months under the purview of long-term production scheduling. An optimal short-term production schedule for MVC involving multiple mines that share the same infrastructure (processing units, stockpiles and waste dumps) requires coordination across the network of mines. Optimizing the short-term production schedules independently at each mine fails to capture the benefit of collaboratively generating schedules for all the mines, resulting in suboptimal schedules. Additionally, in short-term production scheduling, the operations are modeled in greater detail with additional decision variables and constraints to model the more complex multimine MVC situations. This makes the large-scale instances of the problem computationally intractable for standard mixed-integer programming (MIP) solvers. This work presents a customized genetic algorithm (GA)-based heuristic approach to obtain near-optimal solutions to the industrial-scale instances of the short-term production scheduling problem of multimine MVC within a reasonable computation time.
Full-text paper:
Mining, Metallurgy & Exploration (2022) 39: 1403–1427, https://doi.org/10.1007/s42461-021-00523-7
