Research on Coordinated Scheduling Optimization of Coal Blending and Ship Loading

This paper addresses the coordinated scheduling optimization of coal blending and ship loading in push-type coal ports, a critical bottleneck in coal supply chain efficiency. The problem involves complex, interdependent decisions on blending schemes, yard allocation, equipment scheduling, and operation sequencing under multiple constraints. We propose a hybrid optimization approach combining a Constraint Programming (CP) model with an improved Variable Neighborhood Search (VNS) algorithm. For small-scale instances, the CP model employs a Select-TR variable grouping strategy (blending selection first, then timing/equipment allocation) with customized variable/value selection heuristics, reducing solution time by 69.37% and improving solution quality by 70.0% compared to conventional methods. For large-scale problems, the improved VNS algorithm adopts a divide-and-conquer framework with Rolling Time Method and simulated annealing acceptance criteria, utilizing global and local neighborhood structures. This achieves a 100% feasible solution rate and reduces total departure time by 16.3% on average versus pure CP. A case study at a 200-million-ton annual throughput port demonstrates the method reduces ship departure time by 28.4%, eliminates operational conflicts, and increases reclaimer utilization by 25.5% compared to manual scheduling, providing a scientific basis for intelligent port scheduling.