Analysis of Computational Efficiency of Artificial Intelligence based Search Techniques in Trajectory Planning of Industrial Manipulator

Abstract

this paper presents implementation and analysis of computational efficiency of artificial intelligence based search techniques for motion planning of a two degree of freedom revolute-revolute industrial manipulator (2-DOF RR). Depth First Search (DFS), Breadth First Search (BFS), Uniform Cost Search (UCS) and A* Search have been implemented in MATLAB environment for trajectory planning of manipulator from an initial position to final position. A comparative analysis for utilization of computational resources by these search techniques is then presented based on sizes of their frontier and explored sets. Simulation results show that A* is faster and its memory usage is 4 to 9 times less as compared to DFS, BFS, and UCS in trajectory planning on industrial manipulator.