Motion Planning with Obstacle Avoidance within Complex Environment for Redundant Manipulators

Navigating in complex environments is a critical challenge for robotic systems, especially for manipulator robots operating in dynamic and partially unknown spaces. Traditional motion planning methods rely on predefined maps and deterministic strategies. They are limited when adaptive real-time navigation is required. This paper analyzes various motion planning approaches based on graph algorithms which are suitable for unknown environments containing obstacles. Such methods are probe-based algorithms, like Rapidly-exploring Random Trees (RRT). A comparative evaluation of RRT with classical graph traversal algorithms is performed for a planar manipulator robot with additional degrees of freedom, considering computational efficiency and trajectory optimization. Experimental results indicate that RRT and RRT* are efficient choices for the given robotic setup.