[Purposes]Aiming at the shortcomings of existing research on the aircraft recovery problem, improvements are made from three aspects: recovery strategy, path optimization, and solution algorithm, to reduce airline losses and improve aircraft recovery efficiency. [Methods] First, based on the traditional discrete spatiotemporal network, a composite discrete network is formed by incorporating the aircraft swapping strategy, and an initial set of aircraft paths is generated. Second, three path reconstruction rules are designed to filter redundant paths, thereby optimizing the initial path set. Then, on the basis of the traditional resource assignment model, an aircraft recovery model is constructed with the objective of minimizing recovery costs, comprehensively considering constraints such as flight coverage, path coverage, and aircraft turn-around support. Finally, in the construction phase of the traditional Greedy Randomized Adaptive Search Procedure (GRASP) algorithm, a greedy evaluation function is designed according to path costs to improve the GRASP algorithm. The proposed method is validated using real flight schedules from an airline as a case study. [Findings] Compared with the initial aircraft path dataset, both the Gurobi solver and the GRASP algorithm take less computation time on the reconstructed path dataset. In terms of both optimization results and solution efficiency, the improved GRASP algorithm outperforms the traditional GRASP algorithm. [Conclusions] The optimization of aircraft paths and the improved GRASP algorithm can effectively improve solution efficiency, providing strong support for airlines to solve aircraft recovery problems.