Trajectory planning plays a pivotal role in robotic motion control to minimize execution time and suppress joint vibrations while providing stable reference trajectories for controllers. To increase the operational efficiency of collaborative robots while reducing joint vibration, this paper proposes an improved adaptive multi-objective particle swarm optimization (IAMOPSO) algorithm. By integrating dynamic learning factors, adaptive weights, and a mutation operator, the algorithm avoids local optima and improves the diversity of non-dominated solutions while maintaining stable motion. On the generated Pareto front surface, the average optimal solution is chosen using the normalizing function, which can significantly increase the collaborative robots’ joint stability and operational efficiency. Seven-order B-spline curves are used to interpolate the robotic arm’s trajectory in order to guarantee that the position, velocity, acceleration, and jerk of each joint are smooth and continuous while also precisely controlling the movement’s beginning and ending states. Experimental and simulation results show that, compared with conventional optimization algorithms, the proposed IAMOPSO method achieves shorter execution time and lower joint jerk while maintaining a more uniform and continuous Pareto front. The grasping experiment conducted on a collaborative robotic arm further verifies the effectiveness and stability of the proposed method in practical applications.
原文链接
分享
引用
