Reducing Uncertainty in Mineralization Boundary by Optimally Locating Additional Drill Holes Through Particle Swarm Optimization

Abstract

Reduction in uncertainty at the boundaries of ore deposits plays a critical role in mining projects. Such reductions can be made by choosing an appropriate objective function and using a suitable method for its optimization. In the literature, only a combined variance-based objective function can be found for modeling the problem of optimally locating additional drill holes to reduce uncertainty at the boundaries of mineralization. In this study, new objective functions based on interpolation variance (IV) and information entropy (IE) were developed and optimized through particle swarm optimization. The results of the proposed optimization methodology were compared with those of the combined variance-based objective function. Most of the proposed drill holes of the IV-based, IE-based and combined variance-based optimization methods are similar in many parts of the study area. However, IV-based and IE-based methods gave better results than the combined variance-based method. The proposed drill holes based on the IE-based objective function are better scattered over the area. This issue can be regarded as an advantage for the IE criterion both economically and technically.