Extended active observer for force estimation and disturbance rejection of robotic manipulators

• An extended active observer is proposed to estimate force in nonlinear robot systems.
• The observer estimates inertial parameters in real time to ensure the performance.
• The observer considers the effect of measurement noise and rejects it successfully.
• The observer uses concepts behind the Kalman filter to optimally design observer gain.
• The observer is applied to a typical 2DOF nonlinear robot system for verification.

The current control methods applied to robotic manipulators either require full state and force measurements, or use the state and force estimation in the absence of any kind of disturbance. As an alternative approach, a new adaptive motion control approach for robotic manipulators extending the existing active observer for simultaneous inertial parameters and force estimation is proposed. The scheme provides accurate force and full state estimation in the presence of robot inertial parameter variations and measurement noise, both subsequently used in the design of a controller. Since the proposed method relies mainly on the position of the plant, it significantly reduces the difficulty and cost of implementation. The velocity, parameter and force signals are estimated from the position. The approach is applied to a typical two-degree-of-freedom (2DOF) robotic manipulator through computer simulation. The results are encouraging and demonstrate the noise rejection ability of the scheme.