Process

Model predictive control

You might say: MPC, receding horizon control

At each update, model predictive control optimizes a future command sequence, applies the first command, then repeats with a fresh state estimate.

What it does
It plans ahead while enforcing model, input and contact constraints.
Ask
Can the optimization finish before the next control update?
Typical
The MIT Cheetah 3 implementation is a convex model-predictive-control study; timing and horizon are implementation details.

Pattern

Predict, then correct

Connections

Sources

  1. Jared Di Carlo, Patrick M. Wensing, Benjamin Katz, Gerardo Bledt and Sangbae Kim, Dynamic Locomotion in the MIT Cheetah 3 Through Convex Model-Predictive Control, AbstractThis paper presents an implementation of model predictive control (MPC) to determine ground reaction forces for a torque-controlled quadruped robot.
  2. Russ Tedrake, Underactuated Robotics, Ch. 10, Receding-horizon MPC(1) measure the current state, (2) optimize a trajectory from the current state, (3) execute the first action from the optimized trajectory, (4) let the dynamics evolve for one step and repeat. This recipe is known as model-predictive control (MPC).

Use it on your project

teardown-apply: hold "Model predictive control" (model-predictive-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