A Hybrid Fuzzy Collaborative Neural Dynamics for Manipulability Optimization of Redundant Manipulators

Manipulability is a key performance indicator for redundant manipulators, but its optimization is a nonconvex problem with multiple local optima. Most existing motion planning methods either ignore this indicator, simplify it to a convex form, or directly apply nonconvex optimization techniques that are prone to local optima. A hybrid fuzzy collaborative neural dynamics (HF-CoND) algorithm is proposed by integrating neural dynamics with heuristic optimization principles for singularity-free motion planning and control of robots. The proposed framework features a dynamic fuzzy logic system for adaptive parameter modulation along with a robust collaborative mechanism. By coupling elite learning strategies with an adaptive restart mechanism to refine the search, the algorithm enhances global exploration while mitigating the risk of premature convergence. Within this architecture, the collaborative mechanism facilitates high-fidelity information exchange between neural units and significantly diminishes the probability of encountering kinematic singularities. Based on the Franka Emika Panda manipulator, static self-motion and trajectory tracking tasks are conducted. The results show that, compared with existing algorithms, the proposed HF-CoND algorithm achieves higher manipulability.