Research On The LB Assembly Modeling And Control System Based On Image

In the process of TV production and assembly,the traditional manual assembly restricts the assembly efficiency.The current mainstream image matching algorithms such as SIFT and ORB cannot meet the real-time accuracy requirements of intelligent assembly process.Aiming at the requirements of TV production and assembly process,this paper designed an intelligent assembly platform with Cartesian coordinate robot as the main body,realizing the intelligent assembly of TV backplane and TV Light Bar(LB),which is of great significance to the development of industrial automation.The specific research contents are as follows:LB intelligent assembly platform and adaptive LB assembly manipulator are designed.The manipulator can adaptively grab different types of backplane and LB.The finite element strength analysis of the manipulator is carried out,which verifies that the manipulator meets the requirements of process assembly and the rationality of structure.A visual servo control system composed of visual system,control method and robot is established.The mainstream SIFT and ORB image matching algorithms are studied.To overcome the problems of slow matching speed and low robustness of SIFT and ORB algorithms in real-time image feature matching.The AFK algorithm integrating ATo FAST and r BRIEF-BRISK fusion feature descriptor is proposed.At the same time,the three image matching algorithms of SIFT,ORB and AFK are compared and analyzed.The proposed AFK algorithm can meet the requirements of accurate real-time matching.A dynamic texture recognition method GLMLBP-TOP is proposed.Contrast information is added on the basis of texture features.Dynamic texture classification is performed on the intelligent assembly process of LB,so as to monitor the real-time assembly status of LB and ensure the smooth cooperation between LB and the inner hole and groove of the TV backplane.The problem of low robustness and lack of texture information of LBP method are solved.The model of RBF neural network adaptive control system based on local model approximation is established.Dynamics modeling of LB intelligent assembly robot is established by Lagrange method.To overcome the shortcomings of BP neural network,such as slow learning speed and poor network performance,the RBF neural network adaptive control system based on local model approximation is established.The controller design and stability analysis of the system are carried out,the joint position tracking and velocity tracking effects are simulated.At the same time,the hardware platform and software platform of the system are built.The host computer interface is written by using the Qt software platform,and the import of image processing algorithms based on Open CV and Python is completed.Data transmission with PLC through TCP/IP protocol is realized.