Diagnosis using a first-order stochastic language that learns

Chakrabarti, Chaitali; Rammohan, Roshan; Luger, George F.

doi:10.1016/j.eswa.2006.01.048

Cited by 8 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we illustrate how to use the model to reason about specific scenarios. The DBN is a knowledge-based system capable, using a form of EM 12,13 of inferring current states of reactor systems and plant parameters. In some cases, the input of just a few observations change the predicted plant state.…”

Section: Example Results From the Dbnmentioning

confidence: 99%

“…Early attempts to use AI technology in support of nuclear power generation focused on the components of the nuclear power plants. 23 Other researchers 13,24 have used BNs to diagnose device and equipment failures. Abu-Samah et al use BNs for identifying failure signatures for breakdown remediation.…”

Section: Related Workmentioning

confidence: 99%

“…Other researchers 13,24 have used BNs to diagnose device and equipment failures. Abu-Samah et al use BNs for identifying failure signatures for breakdown remediation.…”

Section: Related Workmentioning

confidence: 99%

“…This parameter fitting, with most likely values for missing information given the system’s current priors, is usually accomplished in stochastic models with some form of the expectation–maximization (EM) algorithm. 12,13 In the near term, parameter fitting can be used to help determine the most likely values for reactor parameters that are expected to be available in the control room. We demonstrate examples of parameter fitting in this article when we present our two accident scenarios: a particular parameter is adjusted manually, and given that new reactor state, the full system responds appropriately.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Building and using dynamic risk-informed diagnosis procedures for complex system accidents

Groth

Denman

Darling

et al. 2018

Proceedings of the Institution of Mechanical Engineers, Part O:

View full text Add to dashboard Cite

Accidents pose unique challenges for operating crews in complex systems such as nuclear power plants, presenting limitations in plant status information and lack of detailed monitoring, diagnosis, and response planning support. Advances in severe accident simulation and dynamic probabilistic risk assessment provide an opportunity to garner detailed insight into accident scenarios. In this article, we demonstrate how to build and use a framework which leverages dynamic probabilistic risk assessment, simulation, and dynamic Bayesian networks to provide real-time monitoring and diagnostic support for severe accidents in a nuclear power plant. We use general purpose modeling technology, the dynamic Bayesian network, and adapt it for risk management of complex engineering systems. This article presents a prototype model for monitoring and diagnosing system states associated with loss of flow and transient overpower accidents in a generic sodium fast reactor. We discuss using this framework to create a risk-informed accident management framework called Safely Managing Accidental Reactor Transients procedures. This represents a new application of risk assessment, expanding probabilistic risk assessment techniques beyond static decision support into dynamic, real-time models which support accident diagnosis and management.

show abstract

Section: Example Results From the Dbnmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

“…Other researchers 13,24 have used BNs to diagnose device and equipment failures. Abu-Samah et al use BNs for identifying failure signatures for breakdown remediation.…”

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations