Research On Target Tracking And Flight Control System Of Small UAV Based On Vision Guidance

In recent years,with the characteristics of small size,light weight,easy to carry and high efficiency cost ratio,micro UAV is widely used in military conflict,counter-terrorism operation,traffic monitoring,power inspection,emergency response and other fields.In many practical applications,searching and tracking specific targets of interest to users has become one of the most important tasks of micro UAV.The micro UAV target tracking system uses the onboard vision sensor to obtain the image,locks the target by using the tracking algorithm,and provides the target position and other information to the guidance and control system,so as to assist the UAV to complete the tasks of attack.Because the size and weight of micro UAV are strictly constrained,it can only carry miniaturized,lightweight and low-power onboard equipment.So the onboard computing resources are very limited,which puts forward higher requirements for system integration and algorithm efficiency.To solve the task of target tracking and flight control based on onboard vision,this paper studies on cooperative target recognition,UAV pose estimation,long-time target tracking,target positioning,formation flight control and other related technologies.The main research contents and results are as follows:A micro UAV onboard vision guidance and target tracking system framework is established to meet the requirements of miniaturization,lightweight and low power consumption.Based on the analysis of the actual application scenario of micro UAV target tracking,a precision attack task based on multi UAV cooperation is proposed.The onboard camera,image processing system,flight controller and other equipment are selected and analyzed,and a software and hardware framework of UAV visual guidance and target tracking system for practical application is proposed.Based on the use of low-cost camera,microcomputer and low-cost open source controller,a complete flight test and simulation environment of micro multi rotor UAV are built.Finally,the feasibility of the system is verified by the simulation of UAV passing through a ring target.A micro UAV autonomous landing system based on cooperative QR code is proposed.According to the actual mission requirements of landing the UAV on the recovery platform,a method to improve the landing accuracy through the optimal design of landing pad is proposed.The position and attitude estimation method by using cooperative QR code target is deeply studied.The flight test system in indoor environment is built,and the autonomous landing on the ground fixed platform is successfully realized.The biased proportional guidance law considering angle of view and angle of fall constraints is studied,and a task flow and framework suitable for autonomous landing of micro UAV are proposed.Parallel target detection and tracking algorithm based on deep learning is proposed.In the process of tracking non-cooperative moving targets for a long time,the traditional target tracking algorithm is vulnerable to background interference.When the target is maneuvering or deformed,it may fail,drift or provide inaccurate target scale estimate.For the long-time target tracking task,the basic principle of kernel correlation filter is deeply studied,and a kernel correlation filter target tracking algorithm based on scale adaptation is established.The actual operation effect of the algorithm is tested through the public data set.The target detection method based on convolutional neural network and large number of sample training is studied,and a non-cooperative target parallel detection and tracking framework based on deep learning is proposed.The feasibility of the algorithm is verified by static target and moving target tracking tests,which improves the tracking effect of micro UAV on non-cooperative targets in complex environment.An autonomous tracking method of aerial moving target based on three-axis stabilized gimbal system and a target positioning method are proposed.Based on the application background of UAV tracking moving targets,the structure,kinematics and dynamic model of onboard visual guidance platform are studied.A micro three-axis stabilized platform is independently designed and built by using low-cost brushless motor,encoder and microcontroller,and the control system of the platform is studied.The UAV target tracking strategy based on pure pursuit method and proportional guidance law is proposed.The autonomous tracking flight test of aerial moving target is successfully carried out,and a target positioning method is proposed.The method is tested and verified in outdoor environment,and has strong engineering practicability.A multi UAV formation control system based on distributed consensus theory and a micro UAV formation maintenance control system are established.The distributed consensus theory is studied.A formation and communication topology suitable for micro UAVs is proposed,and a multi UAV formation flight control algorithm is designed.The simulation results show that the algorithm is able to achieve the expected goal.Aiming at the formation maintenance control of micro UAV formation,the trajectory tracking of UAV is studied based on the total energy control method and L1 guidance law.The formation flight test system is built based on Pixhawk open source flight controller.The autonomous flight test of three and five UAVs is carried out,and the flight data are analyzed.The results show that the system has high formation stability and control accuracy.