Abstract
The assembly of pyrotechnic grain demands high precision and stability in robotic arm motion control due to the small shell apertures and stringent assembly accuracy requirements. Inverse kinematics is a core technology in robotic arm motion control. This paper constructs a robotic arm inverse kinematics model by reformulating the inverse kinematics problem as a constrained optimization problem and employs a multi-strategy improved Secretary Bird Optimization Algorithm (ISBOA) to achieve high-precision solutions. Aiming at the problems of restricted solution set exploration, easy to fall into local optimization and slow convergence when solving the inverse kinematics of multi-DOF robotic arm by SBOA, this paper introduces the oppositional variational perturbation, golden sine development and evolutionary strategy to optimize the formation of ISBOA, and verifies its effectiveness through numerical experiments. Simulation experiments using 4, 6, and 7-DOF robotic arms are conducted, with inverse solution results analyzed via PCA dimensionality reduction and K-means clustering, demonstrating the superiority of ISBOA in inverse solution diversity. Finally, a MATLAB-CoppeliaSim-UR16e experimental platform is developed to compare ISBOA with traditional analytical and Newton iterative method. Results are analyzed in terms of assembly accuracy, singular position handling, grasping success rate, and assembly success rate, confirming ISBOA's advantages in pyrotechnic grain assembly and its potential for engineering applications.