Comparative Study of Iterative Methods for Inverse Kinematics of Redundant Serial Robots with Increasing Degrees of Freedom

Abstract

This paper presents a comparative study of five state-of-the-art iterative methods - Particle Swarm Optimization, Quantum Particle Swarm Optimization (PSO), Genetic Algorithms (GA), Damped Least Squares (DLS), and Forward and Backward Reaching Inverse Kinematics (FABRIK) - used to solve the challenging inverse kinematics problem in robots with increasing degrees of freedom in their kinematic chains. The analysis includes 7, 9, 11, 13, and 15 degrees of freedom redundant serial robots with hinge and pivot joints. Performance indicators, such as execution time, iteration count, and final error, were evaluated for each method across 500 randomly generated target poses for the five robotic chains, providing valuable insights into the algorithms' behavior as the degrees of freedom in the kinematic chains increase.