Four-rotor UAV Control System And Visual Positioning Design

Four-rotor UAV has many advantages such as its symmetrical structure,strong maneuverability,good stability,vertical take-off and landing,and so on,which makes the research on drone become a hot issue in academic field and military field.Quadrotors is a multivariable,strongly coupled,nonlinear,underactuated complex system that is susceptible to external factors such as wind and temperature,making it difficult to control its attitude.The design puts forward very high requirements for the entire flight control system,and makes the research of flight control algorithm and attitude algorithm become the key technology of quadrotors control.Normal drones use GPS for positioning,but in a room with poor or missing GPS signals,the drone cannot be accurately positioned.Therefore,a new technology has emerged—visual-based UAV positioning,and become the current research hotspot.This topic has done the following work from the quadrotors control and the visual positioning of drones:(1)Study the basic mechanism and flight principle of the quadrotors,and the drone rack is designed.According to Newton's second law,the external force balance equation and kinematics equation of quadrotors are constructed.Under a reasonable assumption,a simplified mathematical model of the quadrotors is obtained and a theoretical study of the quadrotor flight control system is conducted.(2)Study the drone control algorithm and attitude solution algorithm.The digital filter is used to process the sensor raw data.The quaternion-based complementary filter algorithm is used to solve the attitude.The double closed-loop PID control algorithm is used to control the attitude and position of the UAV.(3)A quadrotors control system was designed,including hardware circuit design and software system design.The hardware circuit is designed with STM32F407 as the main controller,including sensor data acquisition circuit,voltage regulator circuit,current and voltage acquisition circuit,flight control remote communication circuit,PC host computer and flight control communication circuit,microcomputer and flight control communication circuit,Motor control circuit and auxiliary function circuit.Using AD software to complete the design of flight control PCB and remote control PCB;design of flight control software system,use C language to design flight control system on hardware system,custom communication protocol to complete flight control and remote control and PC host communication Periodically collect and process data to control the brushless motor to achieve a double closed-loop PID control algorithm and attitude algorithm.(4)Research the ORB-SLAM visual localization algorithm.The visual positioning hardware system is set up,and the ORB-SLAM algorithm is transplanted on Ubuntu.The source code of ORB-SLAM is modified and debugged in C++ language.The camera pose information is extracted and sent to the flight control system to realize the drone vision.Positioning.