SIT-LMPC: Safe Information-Theoretic Learning Model Predictive Control for Iterative Tasks

Zirui Zang; Ahmad Amine; Nick-Marios T. Kokolakis; Truong X. Nghiem; Ugo Rosolia; Rahul Mangharam

cs.ROFeb 18, 2026

SIT-LMPC: Safe Information-Theoretic Learning Model Predictive Control for Iterative Tasks

Authors: Zirui Zang, Ahmad Amine, Nick-Marios T. Kokolakis, Truong X. Nghiem, Ugo Rosolia, Rahul Mangharam

Organizations: University of Pennsylvania, Philadelphia, PA 19104, USA · Department of Electrical and Computer Engineering, University of Central Florida, Orlando, FL 32816, USA · Lyric, Italia 152 Catania Italy

Abstract

Robots executing iterative tasks in complex, uncertain environments require control strategies that balance robustness, safety, and high performance. This paper introduces a safe information-theoretic learning model predictive control (SIT-LMPC) algorithm for iterative tasks. Specifically, we design an iterative control framework based on an information-theoretic model predictive control algorithm to address a constrained infinite-horizon optimal control problem for discrete-time nonlinear stochastic systems. An adaptive penalty method is developed to ensure safety while balancing optimality. Trajectories from previous iterations are utilized to learn a value function using normalizing flows, which enables richer uncertainty modeling compared to Gaussian priors. SIT-LMPC is designed for highly parallel execution on graphics processing units, allowing efficient real-time optimization. Benchmark simulations and hardware experiments demonstrate that SIT-LMPC iteratively improves system performance while robustly satisfying system constraints.

Figures & tables

Fig. 1: SIT-LMPC architecture: starting from an initial trajectory, the algorithm iteratively updates the safe set and value function model (orange loop), while solving multiple MPPI problems in parallel (blue loop) to generate optimal trajectories.

Panel of Figure 2 — Fig. 2: Point mass experiment. (a) Convergence of lap time over iterations. (b) Layout of the experiment.

Panel of Figure 3 — Fig. 3: Top: Convergence of lap time for three simulated experiments (five independent rollouts in light color and averages in dark color; $\times$ denotes out-of-track). Bottom: Fastest lap trajectories with velocity colorbar.

Panel of Figure 4 — Fig. 4: Ablation study of key SIT-LMPC components for a CEM controller (left) and an MPPI controller (right), comparing NF and BNN value function models, with and without the AP method. Plotted are averages of five rollouts with $\times$ denoting out-of-track.

Panel of Figure 5 — Fig. 5: Experimental setup and results with a real vehicle. (a) Platform. (b) Track. (c) Laptime per iteration. (d) Boundary violations.

Iterative Learning Control

See all 11 papers

Towards Safe Learning-Based Non-Linear Model Predictive Control through Recurrent Neural Network Modeling
Mar 25, 2026Mihaela-Larisa Clement, Mónika Farsang, Agnes Poks +4Nonlinear MPC Model Predictive Control
Institute of Computer Engineering, TU Wien, Vienna, Austria · AIT Austrian Institute of Technology, Vienna, Austria · Institute of Mechanics and Mechatronics, TU Wien, Vienna, Austria
Quasi-Periodic Gaussian Process Predictive Iterative Learning Control
Feb 20, 2026Unnati Nigam, Radhendushka Srivastava, Faezeh Marzbanrad +1Iterative Learning Control Robotic Control
IITB-Monash Research Academy, IIT Bombay, Mumbai, India · Department of Mathematics, IIT Bombay, Mumbai, India · Department of Electrical and Computer Systems Engineering, Monash University, Clayton, Melbourne, Australia
Safety-Guaranteed Imitation Learning from Nonlinear Model Predictive Control for Spacecraft Close Proximity Operations
Mar 19, 2026Alexander Meinert, Niklas Baldauf, Peter Stadler +1Control Lyapunov Functions Imitation Learning
Space Applications Group, e:fs TechHub GmbH, Gaimersheim, Germany. · Faculty of Electrical Engineering and Information Technology, Ingolstadt University of Applied Sciences, Germany.

Mar 25, 2026J/K move · Enter open · S save

Towards Safe Learning-Based Non-Linear Model Predictive Control through Recurrent Neural Network Modeling

Mihaela-Larisa Clement, Mónika Farsang, Agnes Poks, Johannes Edelmann+3

Institute of Computer Engineering, TU Wien, Vienna, Austria · AIT Austrian Institute of Technology, Vienna, Austria · Institute of Mechanics and Mechatronics, TU Wien, Vienna, Austria

Nonlinear MPC Model Predictive Control Policy Learning Safety-Critical Control Recurrent Neural Networks

Open paper AskPDF arXiv

The practical deployment of nonlinear model predictive control (NMPC) is often limited by online computation: solving a nonlinear program at high control rates can be expensive on embedded hardware, especially when models are complex or horizons are long. Learning-based NMPC approximations shift this computation offline but typically demand large expert datasets and costly training. We propose Sequential-AMPC, a sequential neural policy that generates MPC candidate control sequences by sharing parameters across the prediction horizon. For deployment, we wrap the policy in a safety-augmented online evaluation and fallback mechanism, yielding Safe Sequential-AMPC. Compared to a naive feedforward policy baseline across several benchmarks, Sequential-AMPC requires substantially fewer expert MPC rollouts and yields candidate sequences with higher feasibility rates and improved closed-loop safety. On high-dimensional systems, it also exhibits better learning dynamics and performance in fewer epochs while maintaining stable validation improvement where the feedforward baseline can stagnate.

Abstract

Iterative Learning Control

Explore similar work

Towards Safe Learning-Based Non-Linear Model Predictive Control through Recurrent Neural Network Modeling

Quasi-Periodic Gaussian Process Predictive Iterative Learning Control

Safety-Guaranteed Imitation Learning from Nonlinear Model Predictive Control for Spacecraft Close Proximity Operations