An integrated architecture for fixed truck assignment optimization in open-pit mines: Synergistic optimization, hybrid modeling, and multi-objective optimization

Fixed truck assignment (FTA) is a prevalent method of truck dispatching in open-pit mining. To improve equipment utilization and reduce energy consumption, this paper proposes an integrated architecture for optimizing FTA. The innovations of this architecture include three aspects: (1) The concept of synergistic optimization is introduced into FTA for the first time, with a proposed approach of “haul distance–transport grouping–fleet speed.” This approach establishes the relationship between equipment utilization, energy consumption, and their influencing factors. (2) A hybrid modeling method is designed, integrating Kernel Principal Component Analysis and Artificial Neural Network (KPCA-ANN) into the programming model to represent the NP-hard nature of the FTA optimization. (3) An adaptive reference point-based Nondominated Sorting Genetic Algorithm III (ARP-NSGA-III) is introduced, incorporating both letter and real-valued encoding to solve the 3 × b -objective ( ∀ b ∈ Z + ) optimization problem across various route topologies. Finally, the integrated architecture is evaluated using historical data from a case mine. The results show that the optimal FTA solution reduces the shovel idle time (SIT) by 679 h, truck waiting time (TWT) by 1.65 × 10 5 h, and truck fuel consumption (TFC) by 6.2 × 10 6 L. The reduction in fuel consumption (FC) is equivalent to reducing 1.65 × 10 4 tons of CO₂ emissions, accounting for 8.3 % of the total CO₂ emissions from haul activities. • Optimization of transport grouping and fleet speed for varying haul distances. • Integration of Kernel PCA-ANN into the Mixed Integer Programming model. • Adaptive reference point-based NSGA-III incorporates letter and real-valued encoding. • Nonworking time for shovels and trucks is reduced by 679 h and 1.65 × 10 5 h. • Fuel consumption is reduced by 6.2 × 10 6 L, equivalent to 1.65 × 10 4 tons (8.3 %) of CO₂.