Joint Torque Detection Based on Motor Current and Singular Perturbation Control for Cleaning Room Manipulator

This paper presents a method of joint torque detection based on motor current and singular perturbation control for the cleaning room manipulator. In order to improve the robot control performance, state feedback controller with position feedback and joint torque feedback had been widely used, however the joint torque is often obtained by the joint torque sensor, whose price is always high, and the torque sensor increase the joint flexibility, which limits its widely application in the field of robotics. To solve the above problems, this paper presents a method of joint torque detection based on motor current. Combining with the mathematical model of permanent magnet synchronous motor (PMSM) and the dynamic model of robot joint, joint torque model can be obtained. Because of the whole system model with the motor model and robot model has the characteristics of high order, severe nonlinear and strong coupling, which is not convenient to design the controller, the singular perturbation method is introduced to design the controller. The singular perturbation method divides the system into a fast subsystem and a slow subsystem, with the controller separately designed, and then the composite controller of the original system is obtained. With the torque calculation method plus PD control applied to the quasi steady state model, and the torque feedback control used in the inner loop boundary layer model, at the end of this paper, system experiments are carried out on the cleaning room manipulator, and key theoretical results are verified by position tracking experiments.