Design and validation of a fuzzy logic controller for multi-section continuum robots
Jing Liu,Tianyi Zeng,Abdelkhalick Mohammad,Xin Dong,Dragos Axinte
2024-09-30
Abstract:The rise of multi-section continuum robots (CRs) has captivated researchers and practitioners across diverse industries and medical fields. Accurate modeling of these dexterous manipulators continues to be a significant challenge. This complexity stems primarily from many nonlinearities that plague their behavior, including hysteresis and cable elongation. Researchers have devised a spectrum of model-based and learning-based strategies to navigate this intricate landscape, aiming to conquer the modeling problem and elevate control performance. Despite the advancements in these approaches, they encounter challenges stemming from their complex design and intricate learning processes, impairing versatility and hindering robust closed-loop control. This paper introduces a simple-structured, model-less fuzzy logic controller for the closed-loop control of continuum robots. Unlike traditional methods relying on complex models and numerous sensors, this controller boasts a built-in shape reconstruction algorithm. This algorithm allows it to achieve robust control using only the feedback of end position and orientation, significantly reducing sensor dependence. It efficiently adapts to various nonlinearities like hysteresis, cable elongation, and unexpected external disturbances. The experimental results conclusively demonstrate the accuracy and robustness of the proposed fuzzy controller. On a three-section, six-degree-of-freedom continuum robot, it achieved a miniscule trajectory tracking Root Mean Square Error (RMSE) from 0.28 to 0.54 mm, representing just 0.17 to 0.32% of the robot's length. Additionally, the controller demonstrates robustness by successfully handling an unexpected external disturbance of 100g during the trajectory tracking.
Systems and Control