Network-based two-stage robust scheduling strategy for the aircraft assembly system with uncertain duration

Improving the flexibility and robustness of scheduling strategies is crucial for timely delivery of aircraft in dynamic environments. However, with the large scale and frequent disturbances of aircraft assembly operations, conventional dynamic scheduling methods face more difficulty in solving the aircraft assembly scheduling problem with duration uncertainty, especially in analyzing the job duration error transfer characteristics and achieving inverse scheduling optimization. Therefore, this paper proposes an efficient network-based two-stage robust scheduling strategy. Firstly, with the help of network theory, we study the heuristic information about global propagation characteristics of job duration fluctuation and identify the key jobs with high influence of duration error diffusion. Secondly, the dynamic scheduling of aircraft assembly is transformed into a two-stage scheduling with resource-constrained project scheduling and inverse scheduling with resource reallocation for flexible durations. The network-based two-stage robust scheduling strategy achieves scheduling jobs to comply with precedence and resource constraints based on an improved genetic algorithm in the first stage, and the inverse optimization of resource reallocation based on network-based hybrid non-dominated sorting genetic algorithm-II after disturbances in the second stage. Finally, the effectiveness of the strategy is verified on the benchmark from the project scheduling problem library and aircraft assembly data. The experiments show a significant improvement in the performance of key job identification and scheduling in aircraft assembly. Our study provides an efficient scheduling solution for implementing complex aircraft assembly systems with uncertain durations.