Study Of Gesture And Accuracy Hovering Control Of The Quadrotor Under Complex Environment

Currently,the market of drone is developing rapidly.Compared with fixed-wing drones,rotary-wing drones have the advantages of easy-to-use,hovering flight and low cost,therefore the rotary-wing drones are widely used in diversity applications,for instance: inspecting at the factory,express delivery,forest fire fighting,aerial photography,plant protection on the farmland and also the police security monitoring,etc.However,the attitude and hovering control accuracy of rotorcraft are important performance indicators affecting the above application scenarios.Inaccurate attitude and hovering control could cause the UAV to fluctuate,drift,and even get out of control.As a result,deviations from the factory inspection route may occurred,the field crop protection insecticide spraying is not in place,the goods delivery may fail and the aerial photography quality is poor.Meanwhile,the coupling of wind interfering and the noise of drone itself will also have a negative impact on attitude control and fixed-point hovering.That is the reason why we need to study the gesture and accuracy hovering control of the rotary-wing drones.In the control of rotorcraft drones,aircraft attitude control and hovering in the air are achieved through the fusion of various sensor data.Due to accuracy of the sensors,noise and the applicability of data fusion algorithm,the accuracy of attitude and fixed point hovering control in the air is always low.For example,the data fusion of IMU and barometer,due to the sensor noise and the fusion algorithm bugs,there will be big drift when drone hovering in the air;in the height more than 10 meters or complex lighting environment,the fusion of the laser ranging sensor and the optical flow sensor will have poor performance.At the same time,due to the sudden wind or external force,rotorcraft unmanned aerial vehicle hovered from the original spatial location point during hovering and positioning in the air,it was difficult to re-adjust the accurate and rapid return to the original space location hovering point.In the rotary wing dronemarket,since the quadrotor is the mainstream,the quadrotor drone(four-axis drone)is studied as the research object in this thesis.This thesis analyzes the attitude control principle of unmanned four-rotor,the UAV modeling and quantitatively calculates the negative impact of wind power on the hovering of four-rotor UAV.Four types of application are designed for the fixed point hovering mode.There are two types of IMU and barometer-based hovering control schemes are proposed for highaltitude UAV application scenes.Two types low-altitude application scenarios are proposed based on optical flow sensor data fusion.Among the four types of applications,there are two optimal schemes are selected.And also the advantages of serial PID,extended Kalman filter,complementary filter and optical flow method for UAV attitude information data fusion are used in quadrotor attitude control and fixed-point hovering control under complex environments.Therefore,in the two applications of the attitude control and precise fixed-point hovering control,the height range is from 0.5 meters to 100 meters above the ground,and also the applicability and stability are robust.The research focuses on the following aspects:(1)In this paper,the principle of attitude control of UAV is deeply analyzed.Based on this,the model of quadrotor is established;the comparison of the UAV attitude signal detection is carried out.Complementary filters,Kalman filters and extended Kalman filters are derived and compared.(2)A quantitative analysis of the interference factors of drones,such as wind disturbances,was performed to analyze and filter different brands of MEMS performance.(3)Based on the data fusion of MEMS sensors,two hovering solutions for UAVs are proposed in this paper.One is a serial PID control scheme with a small amount of data calculation,and the other is a cascading extended Kalman filter and complementary filter,capable of accurate prediction and compensation.(4)Based on optical flow technology,two hovering schemes with zero position and zero velocity are proposed in this paper and compared.The system design of this article is based on the C/C++ development of the open-source embedded operating system FREERTOS.The STM32F409 MCU is used as the main controller,and the 2.4G HZ wireless data transmission is combined to make the four-axis drone could be remote controlled.The data can be grabbed easily from the PC console and analysis.After repeated tests,the four-axis drones performed well in gust of wind,and the two best solutions performed well.Meanwhile,for a variety of indoor and outdoor quadrotor solution,this paper offer a reference.