NA-OR: A path optimization method for manipulators via node attraction and obstacle repulsion.

Area of Science:

• Robotics
• Artificial Intelligence
• Mechanical Engineering

Background:

• Path planning for robot manipulators in complex environments faces challenges with sampling-based algorithms, often resulting in high path curvature and insufficient safety margins.
• Existing methods struggle to balance path smoothness and obstacle avoidance effectively.

Purpose of the Study:

• To propose and evaluate a novel path optimization method, NA-OR (Node Attraction-Obstacle Repulsion), for robot manipulators.
• To address the limitations of sampling-based algorithms by improving path curvature and safety margins.

Main Methods:

• The NA-OR method iteratively refines paths using Node Attraction (NA) to reduce curvature and Obstacle Repulsion (OR) to enhance safety margins.
• NA pulls path nodes towards their neighbors, promoting smoothness.
• OR generates repulsive torques to move nodes away from potential collisions.

Main Results:

• The NA-OR method demonstrated significant improvements in path curvature and safety margin compared to the initial Bi-RRT path planning.
• Experimental results on a 6-DOF manipulator confirmed the method's effectiveness in diverse scenarios.
• The optimized paths showed reduced path cost and enhanced smoothness.

Conclusions:

• The NA-OR method effectively optimizes manipulator paths, enhancing operational safety and smoothness.
• This approach is particularly beneficial for applications demanding high security and reliable motion planning.
• The proposed method offers a superior alternative for manipulator path optimization in multi-obstacle environments.