Improved task-oriented force feedback control methods for humanoid robot

Force feedback control is essential for humanoid robot as it is important for the robot to work in the human-robot mixed environment. This article proposes improved task-oriented force feedback control methods for humanoid robot which are based on the robot control system proposed by Oussama Khatib. The methods are improved in three parts: joint forces optimization, combination of joint forces and motion and posture design, and force error prediction and motion error prediction. Simulation experimental results are carried and this method can largely improve the manipulation results of humanoid robot.