Research On Six-joint Robot Tracking Algorithm On Linear Conveyor Belt Based On Interactive Multi-model Kalman Filter

At present,industrial robots have been introduced into many industrial sites to achieve different degrees of automation.However,the contour processing technology of fixed platform cannot meet the requirements of the process or is too inefficient,so a better solution is urgently needed.In order to make up for the low efficiency caused by the fixed processing scheme and the inability to meet the process requirements,this subject needs to rely on selfresearch and development to meet this demand.Trajectory tracking is often used in various situations,such as fetching,sorting,etc.The tracking function can greatly improve work efficiency and optimize the overall structure of the production line.Through the encoder displacement and velocity information obtained from the straight line conveyor belt,using a uniformly accelerated motion and uniformly retarded motion two motion model to simulate and forecast the movement of straight line conveyor belt,with the theory of markov chain of probability calculation the actual probability of two models and complete probability transformation,using adaptive kalman filter model more mix two motion model,hybrid forecast in the end.The algorithm implementation and simulation are completed,and the results of verification are analyzed and compared to verify the improvement of accuracy.The machining execution coordinate increment after interpolation and before inverse kinematics solution is obtained from the interpolator of IMICNC,and the increment is superimposed with the displacement increment predicted by the filter,then the inverse kinematics solution is carried out,and finally transferred to the servo driver,so as to complete the function of contour tracking control.After building the experimental platform and writing the software,the contour tracking control system has been verified by the functional verification experiment.