Discrete integral-type zeroing neurodynamics for robust inverse-free and model-free motion control of redundant manipulators.

Nowadays, redundant manipulators have received a lot of attention from industry and academia because of their ability to adapt to complex tasks. Specifically, the motion control is one of the key problems, especially in trajectory tracking. Nevertheless, it is very challenging to achieve precise motion control without relying on their model. In this paper, for the precise and robust motion control of redundant manipulators under different noises, an integral-type design formula combined with zeroing neurodynamics is developed to establish a model-free noise-immune control scheme. Then, an inverse-free version of the control scheme is further proposed to eliminate the matrix inversion operation. At the same time, a filtering method is applied to compensate for the demands for velocity and acceleration information. The robustness and stability of this scheme have been rigorously analyzed, confirming its reliability and effectiveness for high-precision tasks in noisy environments. Simulations and practical experiments on the basis of different manipulators are performed under different noise conditions, and the comparison of different methods shows the superiorities of the proposed scheme with respect to accuracy and robustness in the motion control task. [ABSTRACT FROM AUTHOR]