Process
Control
You might say: feedback control
A controller reads a target and measured robot state, computes a correction, and updates actuator commands.
- What it does
- Feedback makes the robot follow a motion or force objective while reducing disturbance and model error.
- Ask
- Is the loop stable, how quickly does it respond, and how large is the tracking error?
- Typical
- A representative robot controller evaluates its law about 1,000 times per second; exact rates depend on the drive and robot.
What to look for
- Target and measured robot state
- Motion or force objective followed
Parts
- Feedback control loop: A closed loop measures the robot, compares the measurement with a target, computes a correction, commands an actuator, and measures again.
- Robot dynamics: Robot dynamics describe how joint forces and torques produce acceleration from link mass, inertia, Coriolis effects, gravity and friction.
- Interaction control: Interaction control shapes the relation between robot motion and contact force when the robot touches an object or environment.
- Model predictive control: At each update, model predictive control optimizes a future command sequence, applies the first command, then repeats with a fresh state estimate.
Sources
- Kevin M. Lynch and Frank C. Park, Modern Robotics: Mechanics, Planning, and Control, Ch. 11, robot control
We can combine the advantages of feedforward control, which commands motion even when there is no error, with the advantages of feedback control, which limits the accumulation of error.
Use it on your project
teardown-apply: hold "Control" (control) from the teardown at https://teardownatlas.com/robotics/ against my project
Paste it into a coding agent that has the teardown skills.
An AI agent wrote this map with the teardown skill. It can contain errors: check the sources before you rely on a part. About and legal