Abstract: To overcome the drawbacks of the traditional approach to cross-country network design for air- transportation pilot license training, and reduce the cross-country network design cost in total, a cross-country training path connection network was constructed, in which 4 levels including source arc, starting arc, connecting arc and arrival arc were used. Then several sets of feasible training paths for each cross-country training item were obtained by using computer in-depth searching algorithm. Based on these sets, the locations of training bases and the number of aviation transportation pilot license (ATPL) training students flowing on each cross-country training path were considered as decision variables, and several limitations involving the available number of both training bases and total ATPL training students, as well as the training capacity of each cross-country training base, were taken as constraints. A cross-country network design model with objective of minimizing the total cross-country network training cost was established. The results of numeric example show that the proposed approach can decrease the cross-country network training cost by 9.18% in total as opposed to the traditional approach. Additionally, as the airspace and airport capacity range from −30% to 30%, the number of training bases changes and the total network design cost fluctuates from 1.12% to −10.67%. Therefore, the approach is feasible and effective. It can guide the cross-country network design for flight training schools.

Key words: general aviation, cross-country network, connection network, in-depth searching algorithm, cross-country network model, cross-country training path