Research On Attitude Control Of Unmanned Aerial Vehicle Based On Optical Flow And Data Fusion

UAV needs to be better aware of their position and velocity of information to achieve their autonomous flight.Obtaining real-time motion state information of UAV and performing accurate pose estimation are the foundation for the UAV’s steady flight and some complicated tasks.The Global Position System(GPS)is the dominant tool for UAV positioning.Global Positioning System commonly used in small rotor UAV,however,the strength of GPS signals is subject to environmental influences,is relatively unstable,and may even disappear in some environments.Some features of a small rotor UAV,which to a certain extent,limited the use of many positioning methods.For the above-mentioned issues,this thesis uses visual technology to assist the control of UAV to make up for the existing features of UAV.Subsequently,the date of the sensor is fused by a complementary filter,and the date obtained by the fusion is used to control the UAV to improve the performance of the UAV.The following is an overview of the main contents of this article:First of all,visual techniques are used to obtain the motion information of UAV.The sensor has the advantages of low cost and light weight,free from the interference of radiation,and it has no limitation of geographical environment.In addition,the visual sensor can also provide the distance between the UAV and the object,but the GPS can only provide the location of the UAV,so it cannot adapt to the complex environment.Optical flow is a visual method that very suitable for the control of UAV.UAV with optical flow algorithm is suitable for indoor and small space and is not affected by the external environment signal.Another advantage of optical flow is that the depth of the actual object can be calculated through the feature points,which is beneficial to the avoidance of UAV.UAV are made up of a variety of sensors that interact with each other.The information features provided by each sensor are different.In order to achieve a more accurate and smooth control of the UAV,it is very important to optimize the combination of information.The mutual cooperation and complementary advantages of multiple sensors are the basis for autonomous flight of UAV.Therefore,this thesis uses complementary filters to integrate the visual and GPS information,which can improve the efficiency of the entire sensor system and make the UAV more accurately sense its own position and velocity information,so as to improve the performance of UAV.This is the premise of the UAV to achieve autonomous missions,target tracking and other complex tasks.Aiming at the system of UAV with visual assistant control,the back-stepping controller is designed to realize UAV hovering.Then for the UAV control system with complementary filter,a nested PID controller is designed to realize the flight control of UAV.Finally,the feasibility of all the proposed algorithms is verified by Matlab/Simulink,and the simulation results are analyzed in detail.The analysis results show that the algorithm mentioned in this thesis can not only realize the expected function of UAV,but also optimize the control system of UAV.And most importantly,the proposed algorithm also improves the performance and anti-disturbance ability of UAV.