Trajectory Tracking Control Of Robotic Manipulator Via Iterative Learning Control

Manipulator trajectory tracking control is an important research content in robot control technology.Industrial robotic arm is a complex system with high nonlinearity,strong coupling and uncertain model,and it often has repetitive characteristics in the process of motion.How to design an effective control strategy for the manipulator system model and motion characteristics,and realize the reference trajectory of the manipulator Fast and accurate tracking has important research significance.Iterative learning control does not require an accurate mathematical model of the controlled system and has excellent control over systems with repetitive motion characteristics.The specific research content of the paper includes:(1)The kinematics and dynamics of the manipulator system are analyzed and a dynamic model of the system is established.First,through the analysis of kinematics,the pose relationship of the system end-effector with respect to the coordinate reference system is established.Then,through the analysis of dynamics,a mathematical model of the degree of freedom of the robotic arm was established using the Lagrangian method.Finally,the problems existing in the modeling process of the manipulator are analyzed.(2)The trajectory tracking control of the manipulator was designed using an exponential gain-based iterative learning method with initial state learning.Based on the original D-type iterative learning law,on the one hand,the learning part of the initial state deviation is introduced to eliminate the interference of the initial state deviation on the tracking performance of the system,and on the other hand,an exponential gain section is added on the basis of the original fixed learning gain to improve the learning speed.Then,an exponential gain-based PD learning algorithm with initial state learning is proposed in combination with the advantages of P-type learning law,and a strict mathematical proof is given using ? norm theory.Finally,a robot trajectory tracking controller is designed based on the improved learning algorithm and compared with the constant gain controller.(3)Design of manipulator tracking controller based on iterative learning with initial suppression error for acceleration suppression.For the random initial state error,a controller is designed to reduce the learning time of initial state deviation with the increase of the number of iterations on the time axis.It is proved that the system has tracking error in the presence of bounded output interference and state interference.Will converge to a bound with state interference and output interference.Finally,the learning algorithm is applied to the trajectory tracking controller design of the manipulator and verified by simulation.