Abstract
To enhance cross-country training capability and reduce operational costs, this paper proposes a centralized scheduling method for multiple flight schools based on network sharing. By constructing a connection network model and utilizing an in-depth search algorithm alongside linear programming, the study identifies feasible itineraries that meet regulatory requirements while minimizing total costs. Empirical results demonstrate an 8% reduction in training costs and improved airspace management, with benefits scaling as training demand increases.
Introduction
Training a pilot requires specific "cross-country" missions—flying between multiple airports to build navigation and endurance skills. However, flight schools face two major hurdles: skyrocketing operating costs and a severe shortage of available training airspace. Traditionally, flight schools operate as islands, scheduling their own routes independently. This leads to "traffic jams" in certain corridors while others remain underutilized.
Furthermore, meeting the strict regulatory requirements of the Civil Aviation Administration of China (CAAC)—specifically the CCAR Part 141 regulations—is a complex logistical puzzle. Each training "subject" has unique constraints, such as minimum distances, specific numbers of landings, and day/night requirements. This research aims to solve these issues by treating multiple schools as a single, shared network.
Methodology
The researchers proposed a two-stage centralized scheduling method: network construction and mathematical optimization.
Network Construction and Search
Airports are modeled as nodes and flight paths as arcs. To navigate complex regulatory requirements, an In-depth Search Algorithm (DFS) is used to traverse the network and identify every possible "feasible itinerary" for eight distinct training subjects, ensuring each path meets the specific distance and landing requirements mandated by law.
Mathematical Optimization
Once feasible paths are identified, a Linear Programming (LP) model is established. The objective function minimizes the total operating cost across all participating schools, constrained by the demand for training subjects and the maximum hourly aircraft capacity of each flight segment.
Key Results
The model was tested using an empirical case study involving 11 airports and 72 flight subjects across three schools.
- Cost Savings: The centralized model achieved a total cost of approximately 920,587 yuan, compared to 1,001,676 yuan under independent scheduling—an 8% improvement.
- Feasibility and Scalability: The model successfully generated valid itineraries for all subjects while respecting airspace limits. Notably, as training demand increases, the cost-reduction rate climbs from 8% to nearly 18%.
Discussion and Conclusion
The core insight of this study is that network sharing acts as a force multiplier. When schools share a network, the system finds efficiencies that individual schools cannot see; a route that is inefficient for one school may be the perfect fit for another. While the current model uses static variables, it provides a robust framework for reducing costs and managing crowded airspace. Future work could integrate real-time variables like weather and fuel price fluctuations to further enhance practical utility.
Summary updated