This paper proposes a decoupling control algorithm based on dual force sensors to address the coupling between force and motion of end-effector, industrial robots, and end-effector series connection, friction, and coupling forces caused by nonlinear factors. Install a force sensor for decoupling between the end-effector and the industrial robotic arm and another between the end-effector and the polishing head to obtain the interaction force in contact with the workpiece and feedback to the control circuit. In order to verify the effectiveness of decoupling, the decoupling control algorithm was integrated into impedance and admittance control. The goal of eliminating coupling forces was achieved by separating the effects of various inherent dynamics in the system (such as inertial effects and friction) from the required control inputs. Decoupling frees the control system from these dynamic entanglements, enabling it to respond more accurately and adaptively to external forces and environmental disturbances. Experimental results have shown that the bandwidth of the system is increased after decoupling, which can improve the system’s anti-interference ability and tracking performance.