System risk quantification and decision making support using functional modeling and dynamic Bayesian network

Kim, Junyung; Zhao, Xin; Shah, Asad Ullah Amin; Kang, Hyun Gook

doi:10.1016/j.ress.2021.107880

Cited by 22 publications

(6 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…DBN structures can be designed using various data-driven approaches [8,9]. In this study, we utilized the IARF [6], which is a physics-based approach that selects key process variables based on causal relations among subsystems and the laws of physics to design the MDP structure. The IARF approach involves transforming system schematics into an MDP.…”

Section: … …mentioning

confidence: 99%

“…To effectively map state transitions, it is essential to establish clear and comprehensible definitions of state cells and to control the number of system states. An integrated artificial reasoning framework (IARF) [6] transforms the physical representation of a system into a format that can be used for MDP. This study presents a decisiontheoretic approach to optimizing system control logic by connecting artificial reasoning and decisionmaking to the state transition and reward models of MDP.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Operation Optimization Using Reinforcement Learning with Integrated Artificial Reasoning Framework

Kim,

Mikkelson,

Wang

et al. 2023

13th Nuclear Plant Instrumentation, Control &Amp; Human-Machine Interface Technologies (NPIC&HMIT 2023)

View full text Add to dashboard Cite

In large and complex systems, operational decision-making requires a systematic analysis with a vast amount of data from both process parameters and component status monitoring. In this paper, we present an integrated artificial reasoning approach for system state transition models that can help operational decision-making with explainable and traceable reasoning. The integrated artificial reasoning framework is a physics-based approach of defining the system structure in a Bayesian network, so we leveraged it in a Markov decision process (MDP) for finding optimal operational solutions. In our proposed framework, the MDP is implemented on a dynamic Bayesian network (DBN), which represents causalities in a system. The multilevel flow modeling was utilized in order to extract these causalities in a more efficient and objective manner. Since multilevel flow modeling is based on the fundamental energy and mass conservation laws, the target system is decomposed into several mass, energy, and information structures, which serve as the basis for a DBN. The MDP consists of the processes of finding a solution for the Bellman equation, which can be derived from the conditional probability equations of the constructed DBN. System operators can capture stochastic system dynamics as multiple subsystem state transitions based on their physical relations and uncertainties coming from the component degradation process or random failures. We analyzed a simplified example system to illustrate finding an optimal operational policy with this approach.

show abstract

Section: … …mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Operation Optimization Using Reinforcement Learning with Integrated Artificial Reasoning Framework

Kim,

Mikkelson,

Wang

et al. 2023

13th Nuclear Plant Instrumentation, Control &Amp; Human-Machine Interface Technologies (NPIC&HMIT 2023)

View full text Add to dashboard Cite

show abstract

“…They can also design control methods to address these uncertainties, such as adaptive control [6,7], model predictive control [8,9], nonlinear control [10], filtering and estimation [11,12], and robust control [13,14], among others. With the increasing demand for uncertainty modeling and probabilistic inference, BNs have been widely used in various fields, including risk assessment [15], fault diagnosis [16], decision systems [17], gene sequence analysis [18], biomedical image processing [19], and other areas. BN learning consists of parameter learning and structure learning.…”

Section: Introductionmentioning

confidence: 99%

Bayesian network structure learning with a new ensemble weights and edge constraints setting mechanism

Liu,

Zhou,

Huang

2024

Complex Intell. Syst.

View full text Add to dashboard Cite

Bayesian networks (BNs) are highly effective in handling uncertain problems, which can assist in decision-making by reasoning with limited and incomplete information. Learning a faithful directed acyclic graph (DAG) from a large number of complex samples of a joint distribution is currently a challenging combinatorial problem. Due to the growing volume and complexity of data, some Bayesian structure learning algorithms are ineffective and lack the necessary precision to meet the required needs. In this paper, we propose a new PCCL-CC algorithm. To ensure the accuracy of the network structure, we introduce the new ensemble weights and edge constraints setting mechanism. In this mechanism, we employ a method that estimates the interaction between network nodes from multiple perspectives and divides the learning process into multiple stages. We utilize an asymmetric weighted ensemble method and adaptively adjust the network structure. Additionally, we propose a causal discovery method that effectively utilizes the causal relationships among data samples to correct the network structure and mitigate the influence of Markov equivalence classes (MEC). Experimental results on real datasets demonstrate that our approach outperforms state-of-the-art methods.

show abstract

“…W. Kaiming et al [29] used dynamic Bayesian network to analyze the reliability of traction substation of high-speed railway. J. Y. KIM et al [30] established a decision-making process combining dynamic probabilistic risk assessment, dynamic Bayesian network and functional modeling, which can accurately predict and analyze the risk situation.…”

Section: Introductionmentioning

confidence: 99%

Analysis of traffic safety risk state of highway reconstruction and expansion project based on dynamic Bayesian network

Sun

Tian

Zhong-guang

et al. 2023

3rd International Conference on Internet of Things and Smart City (IoTSC 2023）

View full text Add to dashboard Cite

In order to accurately analyze the causes of accidents in highway reconstruction and expansion project and grasp the law of traffic safety state change, a traffic safety situation analysis method based on dynamic Bayesian network model is proposed. Firstly, on the basis of summarizing the existing research results and expert experience, eight key influencing factors reflecting state evolution were selected, and the dynamic Bayesian network structure was established. Secondly, based on fuzzy evidence theory, an improved mixed interval evidence synthesis rule was designed to solve the negative impact of high conflict evidence on parameter learning. Finally, by using Markov theory and time factor management method, the state transition probability is determined, and the dynamic Bayesian network analysis model of traffic safety situation of highway reconstruction and expansion project is constructed. The empirical analysis is carried out by taking the traffic accident occurred during the construction of a highway reconstruction and expansion project as an example. The results show that the traffic safety situation analysis method based on dynamic Bayesian network model can accurately predict and describe the changes of safety situation with time before and after traffic safety accidents. The analysis of the sensitivity of each factor under different situation levels shows that the traffic composition (the proportion of large vehicles) is the most sensitive. Through THE DIAGNOSIS and analysis of the cause of the traffic accident, it is shown that the main reason of the accident is that the temporary traffic safety facilities are not installed in time, which leads to the great difference of vehicle speed, which is consistent with the actual situation at the scene when the accident occurs. The research results provide a scientific basis for taking the corresponding safety control measures.

show abstract

System risk quantification and decision making support using functional modeling and dynamic Bayesian network

Cited by 22 publications

References 37 publications

Operation Optimization Using Reinforcement Learning with Integrated Artificial Reasoning Framework

Operation Optimization Using Reinforcement Learning with Integrated Artificial Reasoning Framework

Bayesian network structure learning with a new ensemble weights and edge constraints setting mechanism

Analysis of traffic safety risk state of highway reconstruction and expansion project based on dynamic Bayesian network

Contact Info

Product

Resources

About