Robot control
A method to control a robot to perform at least one Cartesian or joint space task comprises using quadratic programming to determine joint forces, in particular joint torques, and/or joint accelerations of said robot based on at least one cost function which depends on said task.
1. A method for controlling a robot to perform at least one Cartesian or joint space task, the robot comprising a plurality of links connected by joints, the method comprising:
determining with a computer, at least one of joint forces or joint accelerations of the robot using quadratic programming and based on at least one cost function which depends on at least one first Cartesian or joint space task; and
controlling the robot with the computer to perform the at least one first task based on the determined joint forces or joint accelerations.
2. The method of claim 1 , wherein the joint forces are joint torques.
3. The method of claim 1 , wherein at least one of:
the robot is redundant with respect to the at least one first task;
controlling the robot to perform the at least one first task comprises controlling the robot by force control; or
controlling the robot to perform the at least one first task comprises controlling the robot by compliance control.
4. The method of claim 3 , wherein controlling movement of the robot by force control comprises controlling movement of the robot by torque control.
5. The method of claim 1 , wherein at least one of:
the robot is redundant with respect to the at least one first task; or
controlling the robot to perform the at least one first task comprises controlling the robot by pose control.
6. The method of claim 1 , further comprising:
controlling the robot to perform at least one second Cartesian or joint space task simultaneously with the at least one first task;
wherein quadratic programming is used to determine at least one of joint forces or joint accelerations of the robot based on the at least one first task and the at least one second task; and
wherein at least one quadratic problem is solved to determine the joint forces or joint accelerations.
7. The method of claim 6 , wherein the joint forces comprise joint torques.
8. The method of claim 6 , further comprising:
prioritizing the at least one first task and the at least one second task among each other.
9. The method of claim 8 , wherein prioritizing the at least one first task and the at least one second task comprises selectively prioritizing the tasks by a soft hierarchy or a strict hierarchy.
10. The method of claim 9 , wherein the at least one first task and the at least one second task are prioritized by a soft hierarchy by assigning a weight for each task depending on a relative importance of a given task to other tasks.
11. The method of claim 9 , wherein the at least one first task and the at least one second task are prioritized by a strict hierarchy, wherein priority guaranties the prioritized scheme and lower priority tasks are scarified if they conflict with higher priority tasks.
12. The method of claim 1 , wherein using quadratic programming to determine at least one of joint forces or joint accelerations of the robot is based on at least one of:
at least one inequality constraint in Cartesian and/or joint space;
at least one velocity limit in Cartesian and/or joint space; or
at least one acceleration limit in Cartesian and/or joint space.
13. The method of claim 12 , wherein at least one of:
the joint forces comprise joint torques; or
the at least one inequality constraint comprises at least one of:
at least one pose limit,
at least one position limit, or
at least one orientation limit.
14. The method of claim 12 , wherein at least one of:
one or more of at least one Cartesian constraint, at least one joint velocity constraint, or at least one acceleration constraint is shaped such that a velocity limit becomes zero at a pose limit; or
an orientation of a robot-fixed reference with respect to at least one predetermined direction in Cartesian space is limited.
15. The method of claim 1 , wherein controlling the robot to perform the at least one task comprises controlling the robot while avoiding at least one static or dynamic obstacle.
16. The method of claim 15 , wherein at least one of:
avoiding at least one obstacle is performed based on a Jacobian representing a constraint; or
the at least one obstacle is determined based on at least one of a model of the environment surrounding the robot or on computer vision.
17. A system for controlling a robot to perform at least one Cartesian or joint space task, the system comprising:
means including a microprocessor for determining at least one of joint forces or joint accelerations of the robot using quadratic programming and based on at least one cost function which depends on the at least one first task; and
means including a microprocessor for controlling the robot to perform the at least one first task based on the at least one determined joint forces or joint accelerations.
18. A computer program product for controlling a robot to perform at least one Cartesian or joint space task, the computer program product comprising program code stored on a non-transitory, computer-readable medium, the program code, when executed by a computer, causing the computer to:
determine at least one of joint forces or joint accelerations of the robot using quadratic programming and based on at least one cost function which depends on the at least one first task; and
control the robot to perform the at least one first task based on the at least one determined joint forces or joint accelerations.