| As China’s aging population continues to deepen and labor costs continue to rise,replacing manual labor with industrial manipulator will be inevitable.Most of these operations can be completed by industrial manipulator,and their completion quality depends mainly on the accuracy of industrial manipulator trajectory tracking.Hence,the research on industrial manipulator trajectory control is very important.Iterative learning predictive control combines the advantages of iterative learning control and predictive control,which can improve the control effect by using the historical information generated by the repetitive operation of industrial manipulator and also restrain external disturbance,and control the trajectory tracking of industrial manipulator in both iterative axis and time axis.The main research contents of this dissertation are as follows:(1)The inverse dynamic model of two-degree-of-freedom industrial manipulator is established by analyzing the dynamics of a two-degree-of-freedom industrial manipulator and solving the dynamical equation.Then,the principle of iterative learning control and predictive control is deeply analyzed.Iterative learning control and predictive control are carried out on the iterative axis and the time axis respectively,and an iterative learning predictive control method is obtained.(2)Aiming at the problem that the initial state of the system changes in the actual operation of the industrial manipulator,and the traditional iterative learning predictive control cannot be applied,a new stochastic initial state error iterative learning control algorithm is proposed,and the stochastic initial state error iterative learning predictive control is realized by combining the iterative learning predictive control framework.Through MATLAB simulation,it is proved that the proposed algorithm can realize the trajectory tracking of industrial manipulator under random initial condition.(3)The two-stage iterative learning predictive control with joint velocity parameters is proposed and its convergence is demonstrated.Using MATLAB simulation experiments,the proposed algorithm has higher control accuracy in the industrial manipulators trajectory tracking problem compared with the classic iterative learning predictive control results.Even under the influence of model mismatch and environmental noise,iterative learning predictive control can well complete the trajectory tracking of industrial manipulators. |
