An improved particle swarm optimizer for inverse kinematics of manipulator

Abstract

Abstract Swarm intelligence optimization algorithms have been proven to significantly improve the efficiency and accuracy of solving inverse kinematics problems for manipulators. This is of great importance for research in the fields of manipulator trajectory tracking control and path optimization. Distinguished from conventional swarm intelligence algorithms, the improved particle swarm algorithm proposed in this study uses adaptive inertia weight and asynchronous learning factors to update the particle speed during particle evolution, better balancing global and local search capabilities. At the same time, the idea of generating new offspring by mutation is introduced into particles to enhance the diversity of the population and reduce the possibility of particles falling into local optima. The algorithm's performance was assessed through its application to 13 classical benchmark functions. The obtained results were subsequently compared with the outcomes of genetic algorithms, particle swarm algorithms, and their respective variations. The results suggest that the improved particle swarm algorithm demonstrates significantly better convergence speed and accuracy on most functions in comparison to alternative algorithms (Wilcoxon rank-sum test, α=0.05). Finally, a collaborative manipulator experimental platform was established to assess the efficacy of the improved particle swarm algorithm in solving the inverse kinematics of the manipulator. The findings suggest that the improved particle swarm algorithm demonstrated superior accuracy in solving the inverse kinematics of the manipulator in a shorter duration, yielding errors of less than 0.1 mm, which lays an important foundation for improving the trajectory tracking control accuracy and path planning efficiency of the manipulator.