Model Identification Adaptive Control with $\rho$-POMDP Planning

TL;DR

This paper introduces a novel belief-space planning approach using $ ho$-POMDPs and BiLQR for adaptive control and system identification under partial observability, improving accuracy and safety.

Contribution

It formulates system identification adaptive control as a belief space planning problem with $ ho$-POMDPs and solves it using an adapted BiLQR, demonstrating superior performance.

Findings

Outperforms baseline methods in system identification accuracy.

Effective under partial observability and disturbances.

Applicable to cart-pole and aircraft flight domains.

Abstract

Accurate system modeling is crucial for safe, effective control, as misidentification can lead to accumulated errors, especially under partial observability. We address this problem by formulating informative input design and model identification adaptive control (MIAC) as belief space planning problems, modeled as partially observable Markov decision processes with belief-dependent rewards ($\rho$-POMDPs). We treat system parameters as hidden state variables that must be localized while simultaneously controlling the system. We solve this problem with an adapted belief-space iterative Linear Quadratic Regulator (BiLQR). We demonstrate it on fully and partially observable tasks for cart-pole and steady aircraft flight domains. Our method outperforms baselines such as regression, filtering, and local optimal control methods, even under instantaneous disturbances to system parameters.