This thesis enhances robotic arm control in intricate polishing environments by using impedance control for consistent force regulation between the arm and workpiece. An adaptive impedance control algorithm accommodates contact force variations, boosting system robustness. The PSO-BP variable impedance controller integrates particle swarm optimization and BP neural network for real-time optimization, overcoming limitations of fixed-parameter impedance control on surfaces with curvature changes. Experimental results show a significant reduction in polishing force fluctuations and improved force tracking. These findings support efficient robotic arm utilization in polishing tasks.