Design And Implementation Of Automatic Flight Remote Control System For Drone Based On Android

In recent years,the technology of drone has become more and more mature.Due to its flexible and maneuvering characteristics,it has great application value in scenes such as object search and traffic management.Today’s drones have the following problems in target search applications:they need to manually use the remote control for flight control,shooting and target search,which will consume many human and financial resources;the drone is not effectively supervise during the execution of the mission causing management confusion.If the drone can automatically fly according to the set route and record the flight trajectory and state data during the flight in real time,automatically identify the specified target object in the returned image,automatically return to land,and complete a series of more advanced specific missions.It will greatly expand the functions of the drone,save human resources and improve work efficiency.In order to achieve the above functions,this paper uses the Dajiang consumer-grade drone products and uses the Mobile SDK provided by DJI to design and implement an Android application that enables the drone to fly automatically.The application includes automatic flight module,flight state recording module,real-time image recognition module,and precision landing module.The automatic flight module converts the task data set in the background into a series of instructions recognizable by the drone to realize automatic flight of the drone.The flight state recording module records the flight path and various flight state data of the drone during the automatic flight in real time,so that the actual flight mission execution process of the drone can checked,thereby effectively supervising it.The real-time image recognition module uses the forward computing framework ncnn of mobile terminal and JNI technology.Importing the YOLOv2 network model configuration and weight parameters obtained after training in JNI function that called in Java business class to complete the recognition.The precision landing module utilizes the function of the real-time image recognition module to adjust the attitude of the drone by detecting the relative position of the drone and the apron in real time,so that the drone is constantly approaching the center of the apron.The PID algorithm is used to calculate the adjustment amount of the drone’s attitude adjustment,which makes the landing process more accurate,fast and stable.Firstly,the thesis starts with the relevant industry drone application,and makes the demand analysis and system target description of the autonomous flight application of the drone.Then,according to the requirements analysis,the overall architecture design,functional architecture design and database design of the system are developed.Next,designing and realizing the functions of each module according to the designed system architecture.Finally,the function of each module of the system is tested,and the results show that the system meets the expected goalsThe system developing the application software of the drone control based on the Android system.It is more versatile and easy to operate.It is suitable for enterprises that use the drone to do the target inspection of fixed-point inspection.