CS-MPC: Contraction-Certified Sampling-Based MPC for Nonlinear Systems Under Uncertainty
Organizations: The author is with the Daniel Guggenheim School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA 30332 USA
Abstract
We present contraction-certified sampling-based model predictive control (CS-MPC) for tracking and safety of nonlinear robotic systems under parametric uncertainty, disturbances, and safety constraints that are revealed online. CS-MPC uses a contraction-certified tracking feedback inside every sampled rollout, which concentrates the rollouts around the reference, and an execution layer that verifies the applied input using a quantile-tightened discrete-time control barrier function condition. We derive a verification procedure for nonlinear contraction of the feedback-closed dynamics, and we prove a finite-sample stochastic stability bound and a mission-level probabilistic safety guarantee. We validate CS-MPC in simulation on a rover, a quadrotor, and an asteroid landing. When compared with standard model predictive path integral (MPPI) control, CS-MPC attains the same tracking error with 7 to 20 times fewer rollouts for the rover, and a 2 to 4 times lower tracking error for the quadrotor, with no safety violations in 200 Monte Carlo missions.
Figures & tables
| Method | RMS | Clr. | Coll. | Step | |
|---|---|---|---|---|---|
| CS-MPC | 5 | 1.437 | 0.100 | 0/5 | 27.0 |
| CS-MPC | 10 | 1.331 | 0.101 | 0/5 | 27.5 |
| CS-MPC | 15 | 0.521 | 0.090 | 0/5 | 25.1 |
| CS-MPC | 30 | 0.545 | 0.109 | 0/5 | 25.4 |
| CS-MPC | 100 | 0.534 | 0.112 | 0/5 | 27.9 |
| MPPI b | 5 | 2.436 | 5/5 | 7.8 |
| Method | RMS | Clr. | Coll. | Step | |
|---|---|---|---|---|---|
| CS-MPC | 10 | 0.334 | 0.188 | 0/5 | 4.1 |
| CS-MPC | 30 | 0.382 | 0.225 | 0/5 | 4.3 |
| CS-MPC | 300 | 0.263 | 0.146 | 0/5 | 6.9 |
| CS-MPC | 1000 | 0.250 | 0.208 | 0/5 | 13.7 |
| MPPI b | 30 | 1.594 | 0.379 | 0/5 | 2.7 |
| MPPI b | 300 | 0.718 | 0.210 | 0/5 | 5.5 |
| Method | Land | TD | Viol. | Clr. | Fir. | |
|---|---|---|---|---|---|---|
| CS-MPC | 0.14 (0.31) | 1.5 | 0/10 | 6.3 | 740 | 4.20 |
| MPPI b | 2.59 (4.28) | 37.2 | 0/10 | 5.1 | 1004 | 11.79 |
| MPPI o | 2.44 | – | 10/10 | – | – | |
| Feedback only | 0.15 (0.33) | 1.3 | 10/10 | 689 | 2.14 |
| Example | Rate | Samp. | Bkp. | Inf. | |||
|---|---|---|---|---|---|---|---|
| Rover | 50 | 0 | 0.00 | 0.05 | 30.5% | 0.5% | 0.0% |
| Quadrotor | 100 | 0 | 0.00 | 0.05 | 14.9% | 3.2% | 0.1% |
| Asteroid | 50 | 0 | 0.00 | 0.05 | 7.4% | 0.0% | 0.0% |
Model Predictive Control
Model Predictive Control optimizes future actions with a system model, repeatedly applying the first action as new information arrives.
See all 337 papers