A multi-strategy enhanced moth-flame optimization algorithm for complex inverse kinematics problems in series robots

• Published in: Cluster computing Vol. 28; no. 5; p. 290
• Main Authors: Liu, Jianlin, Huang, Haisong, Fan, Qingsong, Ma, Chi
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2025; Springer Nature B.V
• Subjects: Accuracy; Algorithms; Complexity; Computer Communication Networks; Computer Science; Degrees of freedom; Dynamic characteristics; Equations of motion; Exact solutions; Experiments; Heuristic methods; Inverse kinematics; Kinematics; Manipulators; Methods; Neural networks; Operating Systems; Optimization; Optimization algorithms; Processor Architectures; Redundancy; Robot arms; Robot dynamics; Robotics; Robots; Simulation; Optimization algorithm; The robotic arm UR16e; Inverse kinematics; Joint limits
• ISSN: 1386-7857; 1573-7543
• DOI: 10.1007/s10586-024-04985-4

In this study of optimizing the dynamic characteristics of the manipulator, with the increase of motion redundancy, the complexity of solving the inverse motion problem increases significantly, making it extremely challenging to develop a general analytical solution algorithm. To address this challenge, this paper introduces a multi-strategy enhanced moth-flame optimization algorithm, DMFO. Firstly, the uneven initialization of MFO limits its comprehensive initial search capability and leads to slow convergence speed. Consequently, it lacks competitiveness in specific engineering problems. The mutation chaos strategy, quantum evolution strategy and precise elimination strategy are used to improve the formation of DMFO. Secondly, DMFO and eight meta-heuristic algorithms (GA, PSO, ABC, MFO, DE, WO, BKA, COA) are tested on CEC2017, demonstrating its superior local and global exploitation and exploration capabilities. Then these kinematics models of 4,6,7-DOF manipulators is established based on DH method, and the objective function is derived from the forward motion equation. The effectiveness of DMFO in solving the inverse kinematics of robot arms is validated through its application. Finally, the engineering verification is carried out on the manipulator UR16e, which highlights the success of DMFO in addressing the inverse motion problem of multi-DOF serial robots.