Model Predictive Control
Also known as MPC
Overview
Model Predictive Control (MPC) predicts how a system will evolve, optimizes a sequence of actions over a finite horizon, applies the first action, and replans as measurements arrive. Work spans linear and nonlinear optimization, sampling-based methods such as MPPI, and learned dynamics from Gaussian processes, Koopman models, and neural world models; these methods are used in robotics, autonomous vehicles, energy systems, and process control. Robust and stochastic MPC, control barrier functions, reachability, and distributed optimization address uncertainty, safety, and coordination, while GPU and embedded solvers aim to meet real-time limits. A central tension is that longer, richer predictions can improve planning but increase computation and sensitivity to model error, so practical controllers must balance performance, safety guarantees, and reliable execution.
Latest papers
337 papers30 papers in the last four weeks, down 32% on the four weeks before. 0.2% of all new papers.
A Unified Framework for Empowerment and Predictive Control
SD-DPC: Sparse Dictionary Differentiable Predictive Control
ReCo: Response-Consistent Locomotion with Policy-Aware MPC for Legged Manipulation
Closed-Loop Refinement and Execution for Learned Driving Planners
RL-Guided PAC-NMPC for Probabilistically-Safe Perception-Based Navigation in Unknown Environments
Beyond Policy Alignment: Closing the Planning-Learning Loop for Robot Control with Learned World Models
balance-hard, where Total Average Return (TAR) increases from to , and hurdle, from to ; performance across the broader benchmark remains task dependent, and PL-MPC remains competitive on DMControl. Controlled ablations show different component interactions across the two tasks. We further demonstrate zero-shot sim-to-real transfer on wrench-nut alignment with a 7-DoF KUKA IIWA14, obtaining higher observed success than TD-M(PC)^2 on the training object size and two unseen sizes. Code and data will be available at: https://pl-mpc-humanoid.github.io.