Surrogate Adaptive Controller Tuning Based on DE in a 3R Serial Robot: A Comparative Analysis

The controller tuning based on bio-inspired opti-mization is a highly time-consuming task. This work evaluates the trade-off between controller performance and the computational burden of the novel surrogate indirect adaptive controller tuning approach based on the response surface method (SIACTA-RSM) by implementing the differential evolution algorithm. This work shows an overview of the benefits of SIACTA-RSM when applied to the 3R planar serial robot under parametric uncertainties. Comparative results concerning the well-established indirect adaptive controller tuning approach based on the system's general dynamic model (IACTA-GDM) are done in both the regulation and the tracking control tasks. The outcomes show that the SIACTA-RSM outperforms the computational burden of IACTA-GDM by reducing the average time required to solve the adaptive controller tuning process by 25 times. Furthermore, SIACTA-RSM yields the controller's best performance in the tracking problem by reducing up to 5% IACTA-GDM's mean result. Also, in the regulation problem, SIACTA-RSM achieves a competitive performance by increasing only 2.53% IACTA- GDM's mean outcome. The results demonstrate that the SIACTA- RSM is the most suitable option for online controller tuning ap-plications, keeping a competitive performance while significantly reducing the computational burden.