🤖 AI Summary
This study addresses the issue of instability or potential damage in UR10 robotic arms caused by kinematic singularities during path planning. To mitigate this, the authors propose an adaptive obstacle-avoidance method that integrates fuzzy logic with reinforcement learning. Singular configurations are detected in real time using manipulability measures and condition numbers, and a fuzzy decision mechanism combined with a stable reinforcement learning policy dynamically generates safe trajectories. Notably, this work is the first to embed a fuzzy logic–based safety mechanism within a reinforcement learning framework, enabling efficient avoidance of singular postures. Experiments conducted via PyBullet simulation and the URSim interface demonstrate a 90% success rate in reaching target positions while consistently maintaining a safe distance from singular configurations.
📝 Abstract
This paper presents a comprehensive approach to singularity detection and avoidance in UR10 robotic arm path planning through the integration of fuzzy logic safety systems and reinforcement learning algorithms. The proposed system addresses critical challenges in robotic manipulation where singularities can cause loss of control and potential equipment damage. Our hybrid approach combines real-time singularity detection using manipulability measures, condition number analysis, and fuzzy logic decision-making with a stable reinforcement learning framework for adaptive path planning. Experimental results demonstrate a 90% success rate in reaching target positions while maintaining safe distances from singular configurations. The system integrates PyBullet simulation for training data collection and URSim connectivity for real-world deployment.