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

TL;DR

This paper presents SIT-LMPC, a novel control algorithm that combines information-theoretic learning and model predictive control to enhance safety and performance in iterative robotic tasks within uncertain environments.

Contribution

It introduces an adaptive penalty-based safety mechanism and a normalizing flow-based value function learning approach for improved uncertainty modeling.

Findings

SIT-LMPC outperforms existing methods in safety and efficiency.

The approach enables real-time control on GPU hardware.

Iterative learning improves system performance over time.

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.