Research On Visual Servo Control Of Parallel Robot

In this paper,a visual servo control method of 3-dof parallel robot is studied by closely combining the requirements of industrial robots in sorting,assembly,navigation and other fields.Based on image processing and homography matrix solution and decomposition,the hybrid visual servo control method is studied in this paper.The image processing and control strategy was verified through simulation and experiment.The contributions of this paper are as follows:In the aspect of image processing based on homography matrix,firstly,the LM method and the SVD method are used to solve the expression of the homography matrix,and the homography matrix is decomposed to obtain the information of the mapping relationship between two images,such as the rotation matrix and the translation vector.Then,this paper puts forwards two methods for the acquisition of image feature points required in the homography matrix solution.The feature point matching method uses the SURF algorithm to complete the feature point detection,and the feature point matching is completed by the FLANN method,which has a longer processing time but a higher accuracy.We use Shi and Tomasi algorithm to complete the feature point detection,and use LK optical flow method to perform feature point tracking to complete feature point acquisition.The method can greatly quicken the image processing speed,laying the foundation for parallel robot control.In the design of hybrid visual servo controller,firstly,the model of parallel robot is completed,and the simulation system of three-loop control model of parallel robot drive PMSM is built.Then,based on the homography matrix,the motion control algorithm is designed to control the parallel robot with the error between the images becoming zero.In order to make up for the time spent in image processing,KF algorithm is introduced to predict and feed forward the target motion so as to enhance the real-time performance of the system.Finally,the above three parts were simulated and verified,and the results were analyzed to prove the correctness of the parallel robot and motor models,the feasibility of the hybrid visual servo control algorithm,and the effectiveness of KF algorithm in real-time feed forward of the system.In the design of fuzzy immune controller,in order to compensate for the shortcomings of system robustness caused by the fixed parameters in the above controller,the fuzzy controller is completed by combining fuzzy control and parallel robot visual servo control.In order to improve the response speed of the fuzzy controller,the immune feedback is added on the base of the fuzzy controller,and the fuzzy immune controller design is completed.Finally,it is found that the fuzzy controller improves the accuracy of the system,but cause higher speed fluctuations.The fuzzy immune controller improves the accuracy of the system and speed up the system response.In the verification of experiment,the prototype of hybrid visual servo control system is built with Industrial Personal Computer,parallel robot,Dheng camera and other software and hardware.Then,the reliability and accuracy of image processing and solution and decomposition algorithm of homography matrix are verified by the experimental platform.Finally,the experimental verification of three kinds of controllers is completed.Among them,the fuzzy immune controller has better effect,the smallest average error and the fastest response,which proves the stability of the designed controller,the speed and accuracy of the fuzzy immune controller.