Genetic-Algorithm-Optimized Fuzzy Zeroing Dynamics for Visual Servoing of Robot-Camera Systems Without Modeling and Calibration

Fuzzy control has found extensive application within the realm of robotics. In this study, we introduce a novel control strategy tailored to model-free and calibration-free image-based visual servoing (IBVS) of robot manipulators. This innovation is achieved through the integration of a fuzzy inference system (FIS) that is meticulously tuned using a genetic algorithm (GA). Our proposed approach involves employing two FIS-enhanced zeroing dynamic models: one for data-driven mapping estimation and the other for control purposes. Importantly, this novel framework dispenses with the need for explicit robot and camera models, yet it maintains exceptional precision and prevents error accumulation during task execution. To substantiate the effectiveness of our proposed approach, we conducted extensive simulations and experiments across various robot-camera systems, effectively validating the efficacy, accuracy, and portability across regulation and path tracking IBVS problems. Furthermore, we carried out comparative studies to showcase the superiority of our proposed control scheme.