Lazy Chaos A*: A novel chaos-inspired algorithm for efficient 3R robot arm path planning based on semi-known exploration behavior

Abstract

This paper introduces Lazy Chaos A* (LCA*) that is a new chaos-based path-planning algorithm designed to generate efficient motion of a 3R robotic arm, through semi-known exploration behavior. Based on the concepts of Chaos A*, the proposed approach presents two key innovations: (1) a dynamic re-chaotization mechanism where candidates among neighbors are reshuffled at each expansion step to guarantee a wide exploration and avoid premature convergence; and (2) a lazy evaluation mechanism where inverse-kinematics calculations and collision checks are not performed until required, eliminating unnecessary feasibility calculations. In order to assess performance, LCA* was compared to Chaos A* through four benchmark scenarios, where obstacle density and geometric complexity were gradually increased. The findings demonstrate that LCA* is always able to generate shorter and more practical paths and has considerable decreases in the computational overhead. Those results indicate that the dynamic and lazy approach of LCA* has been able to improve the quality and efficiency of paths it generates, which makes it a promising strategy to manipulator path planning in known static environments that are simulated by semi-known exploration behavior.