Trajectory Tracking Method for Grasping Robotic Arm Using Swarm Intelligence Perception Algorithm

Abstract

Translate Article

This paper proposes a trajectory tracking method based on swarm intelligence perception algorithm to address the problems of insufficient trajectory tracking accuracy and limited anti-interference ability of grasping robotic arms in dynamic environments. This method constructs an integrated framework of “perception planning control”, which integrates the target pose data of visual sensors and the contact force information of force sensors through swarm intelligence algorithms to achieve real-time estimation of environmental dynamic information; Generate a smooth trajectory that satisfies the constraint conditions based on the estimation results, and design an adaptive controller to achieve precise trajectory tracking. The experimental results show that the tracking error of this method converges to within 0.3 mm in static scenes, the response time in dynamic scenes reaches 45 ms, and the error amplification in interference scenes is controlled at 12%. The comprehensive performance is significantly better than traditional PID control and robust control. The deep integration of swarm intelligence and multi-source perception has solved the problem of disconnection between perception and control in traditional methods, providing technical support for complex grasping tasks.