End-Effector Force Estimation for Robotic Manipulators from Motor Current Measurements

Abstract

End-effector force of the robotic manipulator plays an important role in force control. Generally, conventional end-effector force measurement approaches need additional sensors, e.g., the force/torque (F/T) sensor and the joint torque sensor, which raise the cost and introduce complexity into manipulators’ mechanical and electrical design. Therefore, we propose an approach to estimate the end-effector contact force of a manipulator utilizing the motor current measurement without any additional sensors. By building the dynamics model of the manipulator with external wrench, the proposed approach utilizes the friction torque model of the joints to improve the precise of the wrench estimation. Additionally, In order to identify the parameters of motors and the friction torque model, a parameter identification method based on non-linear least squares regression is introduced. The experiments are implemented on the robot IRIC-I which is developed in our lab. Experimental results indicate that the end-effector force can be reliably estimated without any extra sensors, and show the potential of utilizing the estimated end-effector force in force control of the robotic manipulator.