Abstract: A method includes: obtaining current motion state data of the wheel-legged robot, the current motion state data representing motion features of the wheel-legged robot, inputting the current motion state data into a nonlinear controller to obtain a target joint angular acceleration reference value of a target robot joint of the wheel-legged robot, and inputting the target joint angular acceleration reference value into a whole-body dynamics controller to output a joint torque for controlling the wheel-legged robot to perform a control task.

Abstract: Methods, devices, and media for operating a robot arm are provided. In one method, receive a language description for specifying a target implemented by the robot arm; obtain a current state of the robot arm; and determine, according to an action model, an action to be performed by the robot arm based on the language description and the current state. With the example implementation of the present disclosure, the problem of insufficient training data of the robot arm may be alleviated. Further, the pre-trained action model may obtain the basic knowledge about the association relationship between the language description and the person action, may obtain a more accurate action model, and further obtain the action of the robot arm matching the language description in a more efficient manner.

Abstract: A method for controlling a robot to perform a somersault motion enables the robot, when facing an obstacle such as a hole, ditch, or ravine, to cross the obstacle by performing a somersault motion. The robot includes a wheel-leg portion and a base portion connected via a joint. The robot receives a motion instruction and controls a torque of the joint to allow the robot to perform the somersault. During a takeoff phase of the somersault, a center of mass of the wheel-leg portion is lower than a center of mass of the base portion. During a flight phase, the center of mass of the wheel-leg portion is higher than the center of mass of the base portion at some point in time. During the landing phase, the center of mass of the wheel-leg portion is lower than the base portion. This can enable the robot to avoid obstacles.

Abstract: A method includes: obtaining current motion state data of the wheel-legged robot, the current motion state data representing motion features of the wheel-legged robot, inputting the current motion state data into a nonlinear controller to obtain a target joint angular acceleration reference value of a target robot joint of the wheel-legged robot, and inputting the target joint angular acceleration reference value into a whole-body dynamics controller to output a joint torque for controlling the wheel-legged robot to perform a control task.