ABSTRACT
In this chapter, we will first briefly introduce the method of transmitting human impedance information to the robot. The transfer of human impedance to the robot enables the robot to perform flexible operations and complete more complex tasks. Then, we will introduce several common methods for modeling human skills, which are used to model the teaching trajectories so that the robot can reproduce and generalize the teaching trajectories. The trajectory segmentation technique is also introduced, which is used for complex multistep tasks. Then, we will introduce the design of controllers, such as admittance control, which is used in human–robot interaction; variable impedance control, which can enable the robot to adapt the force profiles to different task situations with a natural regulation process; controller based on neural networks, which can avoid the large trajectory tracking error caused by various interferences; and the inaccuracy of robot dynamics model.
