Inertial and Vision Sensor Based End-Effector Sensing and Control for Robot Manipulators

Abstract

In the application of industrial robot manipulators, it is often desirable to obtain accurate position and velocity information regarding the end-effector. Estimations based on motor-side encoders alone are often inaccurate due to joint flexibilities and errors in the robot link kinematics. A vision based approach may also be insufficient due to its low sampling rate and image processing and transportation delay. However, with additional accelerometer measurements, a kinematic Kalman filter (KKF) can be formulated to estimate the end-effector motion accurately without encoder signals. The estimation results can be utilized for real time tracking control effectively. In this paper a multirate kinematic Kalman filter (KKF) scheme is formulated using vision and acceleration measurements from the end-effector. Estimations based on the scheme are utilized as feedback signals for tracking control. The effectiveness of the proposed approach is demonstrated by experiments on a single joint direct drive setup.