Optimal inspection and maintenance planning for deteriorating structural components through dynamic Bayesian networks and Markov decision processes

Morato, Pablo G.; Παπακωνσταντίνου, Κωνσταντίνος; Andriotis, C. P.; Nielsen, Jannie Sønderkær; Rigo, Philippe

doi:10.1016/j.strusafe.2021.102140

Cited by 45 publications

(24 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A key challenge of a preposterior decision analysis is the identification of the optimal set of actions conditional on data a opt|z . The optimization of inspection and maintenance plans forms a stochastic sequential decision problem [33,34], the solution of which requires large computational efforts [11,35].…”

Section: Solution Of the Sequential Decision Problem Via Adoption Of ...mentioning

confidence: 99%

“…The left panel of Figure 8 plots the filtered deterioration state estimate, and the right panel plots the filtered failure rate estimate, as obtained in view of intermittent visual inspection data. For w * 1 = [p I * th = 5 • 10 −4 , p R * th = 1 • 10 −3 , ∆t I = 5], a visual inspection is performed at t insp = [5,10,15,20,25,30,35,40,45] years, as dictated by ∆t I = 5. Since p I * th and p R * th are not exceeded, no additional inspection or repair takes place.…”

Section: Voshm Quantificationmentioning

confidence: 99%

“…We consider the same single underlying true realization of the deterioration process as in Figures 10,11 Time t (years) panel of Figure 12 to the top left panel of Figure 11, one observes that the estimation with SHM barely differs between the two case studies, due to the continuity of the SHM data. The actions corresponding to Figure 12 are inspections at t insp = [32,33,34,35,36], as informed by exceedance of p I th , and repair at t rep = 36. On the other hand, comparing Figure 13 to Figure 10, it is clear that the absence of visual inspection measurements right after the deterioration jump events leads to a severe underestimation of the underlying true deterioration state.…”

Section: Second Case Study: Gradual Deterioration and Shock Deteriora...mentioning

confidence: 99%

“…On the other hand, comparing Figure 13 to Figure 10, it is clear that the absence of visual inspection measurements right after the deterioration jump events leads to a severe underestimation of the underlying true deterioration state. For the same heuristic thresholds as the ones in Figure 10, the actions corresponding to Figure 13 (without SHM) are inspections at t insp = [5,10,15,20,25,30,35,37,38,44,49] and repair at t rep = 38, i.e., two years after the repair informed by SHM, with an increased risk of failure during this time period. For this single realization, in the second case study,…”

Section: Second Case Study: Gradual Deterioration and Shock Deteriora...mentioning

confidence: 99%

See 3 more Smart Citations

A framework for quantifying the value of vibration-based structural health monitoring

Kamariotis¹,

Chatzi²,

Straub³

2022

Preprint

View full text Add to dashboard Cite

The difficulty in quantifying the benefit of Structural Health Monitoring (SHM) for decision support is one of the bottlenecks to an extensive adoption of SHM on real-world structures. In this paper, we present a framework for such a quantification of the value of vibration-based SHM, which can be flexibly applied to different use cases. These cover SHM-based decisions at different time scales, from near-real time diagnostics to the prognosis of slowly evolving deterioration processes over the lifetime of a structure. The framework includes an advanced model of the SHM system. It employs a Bayesian filter for the tasks of sequential joint deterioration state-parameter estimation and structural reliability updating, using continuously identified modal and intermittent visual inspection data. It also includes a realistic model of the inspection and maintenance decisions throughout the structural life-cycle. On this basis, the Value of SHM is quantified by the difference in expected total life-cycle costs with and without the SHM. We investigate the framework through application on a numerical model of a two-span bridge system, subjected to gradual and shock deterioration, as well as to changing environmental conditions, over its lifetime. The results show that this framework can be used as an a-priori decision support tool to inform the decision on whether or not to install a vibration-based SHM system on a structure, for a wide range of SHM use cases.

show abstract

Section: Solution Of the Sequential Decision Problem Via Adoption Of ...mentioning

confidence: 99%

Section: Voshm Quantificationmentioning

confidence: 99%

Section: Second Case Study: Gradual Deterioration and Shock Deteriora...mentioning

confidence: 99%

Section: Second Case Study: Gradual Deterioration and Shock Deteriora...mentioning

confidence: 99%

See 2 more Smart Citations

A framework for quantifying the value of vibration-based structural health monitoring

Kamariotis¹,

Chatzi²,

Straub³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Further details on consistently coupling inference of dynamic Bayesian networks with POMDPs for deteriorating structures can be found in [72,73]. The final state vector for each component is s (i) =(x (i) ,τ (i) ,f (i) ,t), where x (i) is the damage state; τ (i) is the deterioration…”

Section: Environment Detailsmentioning

confidence: 99%

Deep reinforcement learning driven inspection and maintenance planning under incomplete information and constraints

Andriotis

Παπακωνσταντίνου

2021

Reliability Engineering & System Safety

Self Cite

View full text Add to dashboard Cite

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 decisionsteps; (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 wellestablished 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.

show abstract

Assessing the Optimality of Decentralized Inspection and Maintenance Policies for Stochastically Degrading Engineering Systems

Bhustali,

Andriotis

2024

Communications in Computer and Information Science

View full text Add to dashboard Cite

Optimal inspection and maintenance planning for deteriorating structural components through dynamic Bayesian networks and Markov decision processes

Cited by 45 publications

References 41 publications

A framework for quantifying the value of vibration-based structural health monitoring

A framework for quantifying the value of vibration-based structural health monitoring

Deep reinforcement learning driven inspection and maintenance planning under incomplete information and constraints

Assessing the Optimality of Decentralized Inspection and Maintenance Policies for Stochastically Degrading Engineering Systems

Contact Info

Product

Resources

About