With the continuous growth of air transportation volume, the joint scheduling optimization of various ground service resources at airports greatly affects airport operation costs. Considering the correlation among remote aircraft positions, remote gate allocation and ferry vehicles scheduling, a joint optimization model was established with the goal of minimizing the total travel time of ferry vehicles. MIGA (multi-stage improved genetic algorithm) was designed, and local optimization was avoided through three-stage cooperative optimization. Case experiments show that the joint optimization model can significantly reduce the travel time of ferry vehicles, and MIGA can balance optimization accuracy and real-time requirements, verifying the effectiveness of the model and algorithm. For large-scale problem of 40 flights, the accuracy and stability of MIGA outperforms other intelligent algorithms. Ablation experiments confirm that the multi-stage and multi-island evolution mechanism effectively improve the global search ability of the algorithm. Comparison experiments reveal that the remote aircraft position allocation result is the core factor leading to the increase in the travel time of ferry vehicles, Which provides a basis for differentiated scheduling strategies.