A Two‐Dimension Dynamic Bayesian Network for Large‐Scale Degradation Modeling with an Application to a Bridges Network

Abstract

Modelling the stochastic evolution of a largescale fleet or network generally proves to be challenging due to the large number of variables and their interactions. This difficulty may be compounded through complex relationships between various assets in the network. Although a great number of probabilistic graph-based models (e.g., Bayesian networks) have been developed recently to describe the behaviour of single assets, one can find significantly fewer approaches addressing a fully integrated network. An extension to the standard dynamic Bayesian network is proposed by introducing an additional dimension for multiple elements. These elements are then linked through a set of covariates which translate the probabilistic dependencies. A Markov chain is utilized to model the elements and develop a distribution-free mathematical framework to parametrize the transition probabilities without previous data. This is achieved by borrowing from Cooke's method for structured expert judgement and also applied to the quantification of the covariate relationships. Some metrics are also presented for evaluating the sensitivity of information inserted into the covariate-DBN where the focus is given on two specific type of configurations. The model is applied to a real-world example of steel bridge network in the Netherlands which are related through traffic load patterns. Numerical examples highlight the inference mechanism and show the sensitivity of information inserted in various ways. It is shown that information is most valuable very early and decreases substantially over time. Resulting observations entail the reduction of inference combinations and by extension a computational gain to select the most sensitive pieces of information.

Applying Non-Parametric Bayesian Network to estimate monthly maximum river discharge: potential and challenges

“…Different application of NPBNs based on the normal copula assumption can be found in the scientific literature, (e.g. Morales-Nápoles et al, 2014a;Jesionek and Cooke, 2007;Hanea and Ale, 2009;Kosgodagan-Dalla Torre et al, 2017). In reliability studies, Morales-Nápoles and Steenbergen (2014) implemented NPBNs for modelling complex traffic systems and showed that they can be used for computing design values for individual axles, vehicle weight, and maximum bending moments of bridges within certain time intervals.…”

Co-Active Prioritization by Means of Contingency Tables for Analyzing Element-level Bridge Inspection Results and Optimizing Returns

“…A few studies that account for the relationships between elements are available in the literature [33,34]. The fault tree and impact tree methods have been used to analyze how one element affects the other elements [35][36][37].…”

Roadway Asset Inspection Sampling Using High‐Dimensional Clustering and Locality‐Sensitivity Hashing

“…LMDP aims to either minimize total managing costs or maximize asset performance over finite/infinite planning horizon. A lot of efforts have been made to further refine LMDP method since it was first proposed (Durango‐Cohen and Madanat, ; Kosgodagan‐Dalla Torre et al., ; Mishalani and Gong, ; Smilowitz and Madanat, ). Smilowitz and Madanat () extended LMDP model to a network‐level problem by including network‐level constraints, for example, minimum/maximum percentage of assets allowed in certain condition, minimum/maximum yearly budget.…”