Research On Precise Positioning Technology Of SCARA Robot Based On Vision Servo

With the rapid development of China’s economy and the urgent needs of the industrial structure upgrading, the fierce competition in industrial market makes all the enterprises strive to improve the production efficiency to reduce product defect rate. These factors increase the demand of high precision industrial robots in the field of assembly market especially. The SCARA robot is suitable for the repeated and high precision assembly occasion because of its structural characteristics. However, high precision positioning is the precondition of high precision assembly. The study of precision position in SCARA robot with visual servo system consists of the following parts:For the coordinated operation problem of SCARA robot vision servo precision positioning system, Researcher need to establish mathematical model for each part of the system, such as the intelligent vision camera model, the kinematics and dynamic models of the SCARA robot, the transformation relationship between the camera coordinate system and the robot coordinate system.For the SCARA robot control precision, against the disturbance to fixed point-to-point control and the mechanical dither caused by extremely torque mutation in SCARA robot’s movement, Researcher present a controller design method based on fuzzy internal model to improve its control precision. Firstly, the SCARA robot dynamic model is used as controlled object, and using the internal model control principle for the estimation model as the controlled system. The appropriate internal model filter f(s) is chosen with that the steady-state error of the system is zero. And then the whole control law is designed, the fuzzy control rule table is formulated by using the tracking error and the error rate, and the Mamdani method is used as the fuzzy inference rules, and then the gravity method is used to make the fuzzy decision. Finally, the experiment shows that the system works well against external disturbance, it can eliminate the robot trembling caused by sudden change of the output torque and the peak reduce 99.2% compared to single internal model control.To improve SCARA robot plane positioning inaccuracy, we present a compensation strategy based on Grid Model and minimum error approximation principle. Firstly, the paper established a simplified model for SCARA robot and summarized the Grid Model principle. Secondly, the paper constructed an initial Grid Model by analyzing the location relationship between first-time reached actual point and desired point and then selected next starting granularity point of established Grid Model by using minimum error approximation principle. Finally, the convergence and update direction can be determined by the location of desired point in Grid Model. The experiment result shows that the proposed method offset the factor caused by non-accurate model, with positioning accuracy of 0.031 mm is much better than previous value and the proposed method has brief parameter regulation strategy, clearly mobile times and strong industrial application.Building a SCARA robot visual servo precision positioning system based on LABVIEW experiment platform, To design the SCARA robot kinematics inverse solution module, straight line trajectory planning module, motor control module and the error compensation module. Control SCARA robot to complete the specified action and realize accurate positioning function by using NI- 7842 board to achieve the data collection and output.