نوع مقاله : مقاله پژوهشی
نویسندگان
گروه هوش مصنوعی دانشکده مهندسی کامپیوتر دانشگاه یزد، یزد، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Drilling and blasting operations are among the most critical and cost-intensive stages in open-pit limestone mining. Minimizing these costs while controlling adverse impacts such as fragmentation size, flyrock, and backbreak is a fundamental challenge in mining engineering. Aiming to extend and comparatively evaluate previous models, the present study conducts a comparative analysis on six metaheuristic algorithms: Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Imperialist Competitive Algorithm (ICA), Differential Evolution (DE), Grey Wolf Optimizer (GWO), and Covariance Matrix Adaptation Evolution Strategy (CMA-ES). To constrain violations of geomechanical and safety limits, a penalty function technique was utilized in the modeling, and the performance of the algorithms was evaluated over 15 independent runs. Numerical simulation results indicate that the PSO, DE, and CMA-ES algorithms, in their best runs, achieved the minimum objective function value, reducing the operational cost to 6104.74 Rials per ton. The optimal values for the design parameters in this state were determined as spacing of 3.68 meters, 459 blastholes, and a hole length of 13.10 meters. Nevertheless, the stability analysis of the runs indicates the complete stability of the PSO algorithm in finding the global optimum across all iterations. In contrast, the DE algorithm exhibited a slower convergence rate, and the advanced CMA-ES algorithm, despite finding the global optimum in some instances, suffered from severe instability in average runs when facing the rigid boundaries of the penalty function. Furthermore, the GWO and GA algorithms experienced premature convergence, recording higher costs (6188.23 and 6149.91 Rials per ton, respectively). Overall, these results demonstrate that in complex and highly constrained mining search spaces, the PSO algorithm is a reliable option for the economic design of blasting patterns due to its excellent balance between accuracy and stability.
کلیدواژهها [English]