The Surprising Effectiveness of Linear Models for Whole-Body Model-Predictive Control
By: Arun L. Bishop , Juan Alvarez-Padilla , Sam Schoedel and more
Potential Business Impact:
Robots walk and move without complex math.
When do locomotion controllers require reasoning about nonlinearities? In this work, we show that a whole-body model-predictive controller using a simple linear time-invariant approximation of the whole-body dynamics is able to execute basic locomotion tasks on complex legged robots. The formulation requires no online nonlinear dynamics evaluations or matrix inversions. We demonstrate walking, disturbance rejection, and even navigation to a goal position without a separate footstep planner on a quadrupedal robot. In addition, we demonstrate dynamic walking on a hydraulic humanoid, a robot with significant limb inertia, complex actuator dynamics, and large sim-to-real gap.
Similar Papers
Whole-Body Model-Predictive Control of Legged Robots with MuJoCo
Robotics
Robots walk and move more like humans.
Whole-Body Inverse Dynamics MPC for Legged Loco-Manipulation
Robotics
Robot dog with arm does complex tasks.
Real-time Whole-body Model Predictive Control for Bipedal Locomotion with a Novel Kino-dynamic Model and Warm-start Method
Robotics
Makes robots walk smoothly and quickly.