Multi-objective optimization of an open-pit mining system to determine safety buffer using the modified NBI method and the meta-model approach

Abstract

Background

The open-pit mining industry plays a crucial role in the extraction of valuable minerals and resources, contributing significantly to global economies. However, the increasing complexity of mining operations, necessitates the adoption of advanced optimization techniques. A myriad of engineering problems include multiple conflicting objectives, which today are often solved by expensive simulation computations. Methods based on surrogate models are one of the approaches to solving this type of problem.

Method

This paper presents the multi-objective optimization in the extraction system of a copper mining complex using the normal boundary intersection (NBI) method and a meta-regression model to determine the economic lot-sizing. In this paper, three objective functions were considered including: (i) maximizing the amount of sulfide rock extraction, (ii) minimizing the total cost of the haulage system, and (iii) maximizing the total rocks loaded on trucks. The central composite design (CCD) method was used to develop the design of experiments (DOE) for the design variables.

Results

according to obtained findings, the considered design variables were the number of trucks of 120 tons, 240 tons, 35 tons, and 100 tons. The values of objectives considered in each combination of experiments were considered the response surface. A quadratic nonlinear regression model was determined for the objectives of maximizing the amount of sulfide rock extraction and minimizing the total costs of the haulage system and a linear regression model for the objective of maximizing the total rock loaded on trucks. The accuracy of the models was checked using the predicted residual error sum of squares (PRESS) and $R^2$. Moreover, the most common PRESS error was employed to validate the Meta-models. Subsequently, the multi-objective optimization problem was solved using the NBI method. Finally, Pareto solutions were provided using this approach, and they were discussed.

Conclusion

Implementing multi-objective optimization in open-pit mining using the modified NBI method and meta-model approach enhances decision-making by balancing safety buffers with operational efficiency. This strategic framework enables managers to minimize risks while maximizing resource extraction, ultimately leading to more sustainable mining practices.