Design Of Quadrotor UAV Tracking System Based On Machine Vision

Small quadrotor UAVs have many advantages,such as light weight,small size,easy handling and good stability.In recent years,with the improvement of the performance of hardware equipment,the functions realized based on the UAV platform have become more and more abundant.In particular,the combination of UAVs and machine vision modules enables UAVs to independently identify and track predetermined targets,which requires UAVs to have good stability during flight for photographing and identification.By combining the machine vision module with the UAV platform,it greatly expands the application scenarios of UAVs and has certain engineering application value and practical significance.This paper mainly studies the flight stability of UAVs and the related work of identifying and tracking predetermined targets with machine vision.The contents are as follows:1.Modeling of quadrotor UAV: Through the introduction of the selection of hardware for each module of the UAV,the mathematical model of the power module is established,and the flight performance of the built UAV is calculated and analyzed,and the model can stall and fly for about 14 min.Explain the UAV flight principle and establish the corresponding coordinate system.And according to the specific structure of the UAV,the dynamics and mechanical models of the UAV are derived by using the Newton-Euler equation.The overall mathematical model of the quadrotor UAV is obtained by analyzing the overall system model and linearizing it,and the corresponding simulation model is established for the simulation verification of the controller.The overall mathematical model of the quadrotor UAV is obtained by analyzing the overall system model and linearizing it,and the corresponding simulation model is established for the simulation verification of the controller.2.Active Disturbance Rejection Control Algorithm:By researching the controller of UAV,aiming at the shortcomings of traditional PID control in UAV controller,an Active Disturbances Rejection Controller(ADRC)algorithm is proposed.Because UAV is easily disturbed by environmental factors during flight,ADRC has good anti-interference ability and is very suitable for UAV platform.And the simulation performance analysis of ADRC and PID algorithm is carried out through MATLAB/Simulink.In the same environment,it is concluded that ADRC has smaller fluctuation amplitude,faster adjustment speed and better stability effect than PID control.3.Flight control based on machine vision:Using the improved UAV platform to combine with the machine vision module,a flight controller based on machine vision is constructed to complete the accurate identification of the predetermined target and the tracking of the moving target.At the same time,the improved flight platform has better flight stability and can identify and track targets stably for a long time,which verifies the reliability and stability of the experiment.