Print Email Facebook Twitter Deep reinforcement learning driven inspection and maintenance planning under incomplete information and constraints Title Deep reinforcement learning driven inspection and maintenance planning under incomplete information and constraints Author Andriotis, C. (TU Delft Structural Design & Mechanics) Papakonstantinou, K. G. (The Pennsylvania State University) Date 2021 Abstract Determination of inspection and maintenance policies for minimizing long-term risks and costs in deteriorating engineering environments constitutes a complex optimization problem. Major computational challenges include the (i) curse of dimensionality, due to exponential scaling of state/action set cardinalities with the number of components; (ii) curse of history, related to exponentially growing decision-trees with the number of decision-steps; (iii) presence of state uncertainties, induced by inherent environment stochasticity and variability of inspection/monitoring measurements; (iv) presence of constraints, pertaining to stochastic long-term limitations, due to resource scarcity and other infeasible/undesirable system responses. In this work, these challenges are addressed within a joint framework of constrained Partially Observable Markov Decision Processes (POMDP) and multi-agent Deep Reinforcement Learning (DRL). POMDPs optimally tackle (ii)-(iii), combining stochastic dynamic programming with Bayesian inference principles. Multi-agent DRL addresses (i), through deep function parametrizations and decentralized control assumptions. Challenge (iv) is herein handled through proper state augmentation and Lagrangian relaxation, with emphasis on life-cycle risk-based constraints and budget limitations. The underlying algorithmic steps are provided, and the proposed framework is found to outperform well-established policy baselines and facilitate adept prescription of inspection and intervention actions, in cases where decisions must be made in the most resource- and risk-aware manner. Subject Constrained stochastic optimizationDecentralized multi-agent controlDeep reinforcement learningInspection and maintenance planningPartially observable Markov decision processesSystem risk and reliability To reference this document use: http://resolver.tudelft.nl/uuid:0713e10e-31f1-499d-89dc-940e50faef3a DOI https://doi.org/10.1016/j.ress.2021.107551 Embargo date 2021-09-11 ISSN 0951-8320 Source Reliability Engineering & System Safety, 212 Bibliographical note Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. Part of collection Institutional Repository Document type journal article Rights © 2021 C. Andriotis, K. G. Papakonstantinou Files PDF 1_s2.0_S095183202100106X_main.pdf 3.04 MB Close viewer /islandora/object/uuid:0713e10e-31f1-499d-89dc-940e50faef3a/datastream/OBJ/view