The Research And Implementation On Autonomous Hovering Of Micro Quadrotor Aircraft

Multi-rotor aircraft has been applied to many fields because of its simple structure,outstanding maneuverability,motility and low-speed flight characteristics such as vertical takeoff landing and hovering.Meanwhile low-cost micro quadrotor aircraft has also gradually attracted much attention because it is more portable,safe and durable.Due to the limitation of volume and cost,it is difficult to autonomous flight stability based on airborne sensors because of its weak hardware performance,simple software system and function,light load,limited sensor function and quantity,so its application is limited.In this thesis,it will improve the efficiency of the system by porting lightweight systems and optimizing the design of multi-task software architecture.Add optical flow module and position evaluation control method,use optical flow velocity measurement to achieve stable autonomous hovering of micro quadrotor aircraft based on low-cost visual recognition.It also studies in depth the main contents about flight control system.First,studies the working principle and background knowledge of micro quadrotor aircraft.Complete the secondary development of software and hardware based on open source projects.According to the requirement,optimizes the hardware design and makes the experimental prototype.Studies the working principle and application characteristics of the main inertial measurement unit and optical flow modules,and designs the optical flow control program based on SAD(Sum of Absolute Differences)block matching algorithm.According to the hardware characteristics,selects the software system,and builds the compiling environment to port system and design software architecture and function module.Secondly,studies in depth the principle of position attitude evaluation and control of the basic flight function based on IMU after system optimization.Adjusts the parameters of controller and stability of system to realize altitude hold mode.On this basis,installs the optical flow module and completes the related work of sensor system expansion.Referring to the open source position evaluation and control method,designs the Kalman filter position evaluation corrected by optical flow velocity measurement and the process of horizontal and vertical position control based on cascaded PID(Proportion Integration Differentiation)controller,and ports them to the related task.Adjusts the parameters of controller and stability of system to realize stable autonomous hovering based on optical flow.Finally,designs the flight test to analyze and compare the flight data recorded by external camera and collected by ground station,it is calculated that in hover mode based on optical flow velocity measurement the aircraft's horizontal position error is about 3 cm,velocity error is about 2 cm/s,vertical position error is about 1 cm,flight time is more than 5 minutes.It is verified that micro quadrotor aircraft can implement long high-precision stabilized autonomous hovering based on low-cost visual recognition by porting and optimizing flight control system and installing validly optical flow module.