Research On Precise Landing Control System Of Multi-rotor UAV

Multi-rotor unmanned aerial vehicles(UAVs) are widely used in the field of power inspection because of their simple structure and flexibility.However,when the UAV descends in the classical control system mode,due to the limited positioning accuracy of GPS signals and the air flow influence of "ground effect",the error of accurately landing the drone on the target is relatively large.Therefore,this paper integrates visual positioning technology into the existing classic control system of UAV,and studies the autonomous landing control system of UAV,designs a precise landing control system integrating visual positioning on the embedded platform of the airborne end of UAV.This paper was supported by the Guangzhou science and technology project "smart UAV power patrol application demonstration project(project number: 2020070400004)".The research work of this paper mainly includes the following contents:1.Acquisition of landing area targets in visual images.In the process of collecting visual images,airborne cameras will be interfered by environmental factors such as too complex background of landing points,uneven illumination and so on,resulting in unsatisfactory image collection effect.For this reason,this paper firstly carries out preprocessing operations such as gray level homogenization,image enhancement and filtering on the landing area target image,then uses OTSU algorithm to segment the target of the landing area in the image,and then calculates the center point coordinate of the target by drawing the minimum circumscribed rectangle of the contour of the target.Combined with the position coordinates of the UAV itself,the position deviation of the target relative to the landing area of the UAV in the geographic coordinate system is obtained after coordinate conversion.2.Design of precise landing control system.The classical control system of UAV adopts the control mode of "outer ring position + inner ring attitude",but there is a large landing error.For this reason,this paper proposes a control method that integrates visual positioning.The position deviation of UAV relative to the target is obtained b y vision,and then a speed control link is added to the output of the outer loop position control module,which consists of a speed segmentation module and a speed controller,thus forming an accurate landing control system for the UAV.In this paper,a speed segmentation module is designed to determine the speed setting of the UAV according to the position deviation of the UAV relative to the target and the upper limit of the flight speed.Then,P-type,PI-type,PD-type and PID-type speed controllers are designed for experimental comparison,and the speed controller with better control effect is selected to control the UAV to reach the set speed output by the speed segmentation module.3.Self-tuning of speed controller parameters.According to the comparative experiments of four types of speed controllers,the speed PD controller is selected in this paper,and on this basis,a fuzzy inference engine is designed to realize the automatic adjustment of controller parameters,so as to improve the efficiency of parameter setting.4.Experiment and result analysis.In this paper,a six-rotor drone is taken as the research object.First,the visual image processing experiment of the drone is performed,and then six simulation experiments and two physical test experiments are performed.The experimental results show that:(1)When the UAV lands autonomously with the classical control system,there is a large overshoot and steady-state error,and serious oscillation occurs in the deceleration process,with the transition time re aching 10 s.(2)When the UAV lands autonomously using the precision landing control system,the overshoot and steady-state error of the system can be effectively reduced and the oscillation phenomenon can be eliminated.(3)Among the four types of speed controllers designed,when the precision landing control system adopts the mode of position control + speed PD controller,the UAV landing control effect is better;And when the position control + parameter self-tuning speed PD controller is adopted,the stability of the system can be further improved to a certain extent,the control precision can be improved,and the better control of the UAV can be realized.(4)When the six-rotor UAV lands autonomously under the action of classical control system,the average landing error is 0.74 m,while when it lands autonomously under the action of position control + speed PD controller,the average landing error is 0.25 m.(5)When the UAV lands autonomously using the position control + parameter self-tuning speed PD controller,the average landing error is 0.17 m,and the setting time of controller parameters is reduced from the original time above 1200 s to 90 s.It can be seen that the precision landing control system integrated visual positioning designed in this paper can impr ove the landing accuracy of UAV to a certain extent and reduce the setting time of controller parameters,which is of great value to the application of UAV in the field of power inspection.