Flying Ad Hoc Network(FANET)is a robust and flexible wireless communication method that involves key factors such as signal coverage,throughput,and battery consumption.However,there are still practical challenges to optimizing these parameters.In this paper,recurrent neural network(RNN)is used to model the influencing factors,and the adaptive predictive control(APC)method is introduced to construct the control system,and the comprehensive performance of the airborne Wi-Fi network and the FANET experimental system is deeply studied.Through the intelligent Ad Hoc network and AP control method,the research team successfully optimizes the communication system capacity and coverage of the UAV,which has significant advantages to the aircraft control.