论文部分内容阅读
采用Lagrange方法建立了一类新型3P6R平面3-DOF串-并混联拟人臂的标称动力学模型。针对该机构的重复轨迹跟踪问题,考虑其不确定性,充分利用其已知动力学部分,提出了一种集中参数自适应-闭环迭代学习控制器。在每个迭代周期内采用自适应算法学习由未建模动态、外部干扰及摩擦力等多种因素造成的集中不确定性上界,进而逐次补偿由其造成的误差;闭环变系数迭代学习算法保证了该系统在迭代域内收敛,实现了完全轨迹跟踪。严格的证明及仿真结果验证了此控制器的有效性。
A new type of 3P6R plane 3-DOF serial-parallel mixed pseudo-human arm dynamic model was established by Lagrange method. Aiming at the repeated trajectory tracking problem of this institution, taking into account its uncertainty and making full use of its known dynamics, a centralized parameter adaptive-closed-loop iterative learning controller is proposed. In each iteration period, an adaptive algorithm is used to learn the upper limit of concentration uncertainty caused by many factors such as unmodeled dynamics, external disturbance and friction, and then to compensate the errors caused by them. The closed-loop variable coefficient iterative learning algorithm It ensures that the system converges in the iteration domain and realizes complete trajectory tracking. Strict proof and simulation results verify the effectiveness of this controller.