Model-based Unknown Input Estimation via Partially Observable Markov Decision Processes

Wei Liu, Zhilu Lai, Charikleia D. Stoura, Kiran Bacsa, Eleni Chatzi

Keywords: unknown input estimation, partially observable markov decision process, model-based reinforcement learning, model predictive control, cross-entropy method, dynamics modeling

Abstract: In the context of condition monitoring for structures and industrial assets, the estimation of unknown inputs, usually referring to acting loads, is of salient importance for guaranteeing safe and performant engineered systems. In this work, we propose a novel method for estimating unknown inputs from measured outputs, particularly for the case of dynamical systems with known or learned dynamics. The objective is to search for those system inputs that will reproduce the actual measured outputs, which can be reformulated as a Partially Observable Markov Decision Process (POMDP) problem and solved with well-established planning algorithms for POMDPs. The cross-entropy method is adopted in this paper for solving the POMDP due to its efficiency and robustness. The proposed method is demonstrated using simulated dynamical systems for structures with known dynamics, as well as a real wind turbine with learned dynamics, which is inferred via use of a Replay Overshooting (RO) scheme, a deep learning-based dynamics method for learning stochastic dynamics.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Machine Learning for Sciences (eg biology, physics, health sciences, social sciences, climate/sustainability )