Sensor-free Variable Impedance Control Using Quadratic Programming and Extended State Observer

Abstract

In collaborative robotic scenarios such as polishing and assembly, achieving accurate and adaptive manufacturing quality relies on force/position synchronized control of the robot. However, measuring contact forces and accurately tracking trajectories can be challenging in the absence of force or torque sensors. This paper presents an impedance control algorithm based on quadratic programming (QP) and extended state observer (ESO) to address these challenges. The proposed approach establishes a new impedance model and transforms the variable impedance law design problem into a QP problem, enabling to consider impedance constraints and adaptively trade off between the tracking accuracy and contact compliance. Additionally, an ESO is designed to estimate and compensate for the contact force. Simulation results demonstrate the effectiveness of the proposed algorithm by applying both constant contact force and time-varying contact force.