Research On Tracking Aiming And Vision Positioning System Of Space Target Laser Drive Experimental Device

With the increasing frequency of human space activities,the number of space debris has increased dramatically,posing a great threat to the safety of spacecraft and astronauts,requiring manual removal.The "laser removal debris" technology has many advantages such as long distance of action,large range,and reusability,and is considered to be the most effective measure for removing debris between 1 and 10 mm in diameter.At present,the scheme based on this technology is still in the stage of theoretical analysis and establishment of the fragment dynamics response model.The development of the space target laser-driven experimental device with space target tracking,aiming,vision positioning and laser ablation is the key to the experimental analysis and verification of the laser ablation driving debris dynamic response model.The main research centered on the two subsystems of the experimental device's tracking,aiming and visual positioning and its key technologies.The target is to establish a moving target tracking system and a target pose calculation model based on monocular vision to achieve dynamic tracking aiming and spatial pose calculation for the cube target.In this thesis,the control scheme of the tracking system is designed.In order to meet the low delay requirement of the tracking and aiming,the real-time communication mode between the image processing module and the servo control module is adopted.Since the application environment of the device is an indoor environment whose background can be controlled,the Otsu method capable of adaptive threshold segmentation is used to segment the image,identify the target,and obtain the position information of the target relative to the turntable.The mathematical model of the servo drive system is established,and the fuzzy control and PID control are implemented.The fuzzy setting of the position loop PD parameters is completed and simulated by simulink.The simulation results show that the fuzzy tracking of the position tracking controller has a smaller overshoot and shorter response time than the controller without fuzzy tuning.The time-varying sinusoidal signal also has Better tracking results.In the visual positioning system,the mathematical model of camera imaging is established.The model parameters are calibrated by the Zhang calibration method,and the mathematical correspondence between the space point and its image point is established.The two feature points of the corner point and the circular center in the monocular visual positioning are compared,and the conclusion that the circular center has better recognition stability and higher positioning accuracy is obtained.In order to solve the problem that can not be distinguished between the circular shapes during multiple circular imaging,the target surface pattern of the square is designed in combination with the Marker code,which realizes the stable recognition of different circular surfaces of different surfaces and the same surface of the target.The monocular vision localization method based on single image and coplanar feature points is used to solve the pose,and the reprojection error of the pose calculation result is iteratively optimized by the Levenberg-Marquardt method.The optimized position and attitude are finally used.The calculation results verify the feasibility of the method through simulation.The tracking performance of the tracking system and the positioning accuracy of the visual positioning system were verified by experiments.The experimental results show that the steady-state tracking accuracy of the tracking system meets the requirements and the dynamic performance is good.The visual positioning system will appear the jump of the pose when the absolute values of the RX and RY angle values of the target surface and the camera imaging plane are less than 5°,and analyzed the causes of this “fuzzy phenomenon”,the accuracy meets the requirements in the other case.