Combining model-based diagnosis and data-driven anomaly classifiers for fault isolation

Jung, Daniel; Ng, Kok Yew; Frisk, Erik; Krysander, Mattias

doi:10.1016/j.conengprac.2018.08.013

Cited by 58 publications

(38 citation statements)

References 30 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This relieves the diagnostician from the burden of creating a model of the system/process to be diagnosed, sometimes even from the burden of identifying proper features since deep learning techniques offer potential tools for automatic feature extraction from (physical) signals . However, in many industrial applications, faults are rare events and available training data from faulty conditions is usually limited . Hence, collecting a sufficient amount of data from relevant faulty situations for data‐driven diagnosis is a time‐consuming and expensive process.…”

Section: Diagnosis and The Myth Of Total Knowledge Compilationmentioning

confidence: 99%

See 1 more Smart Citation

Twin‐engined diagnosis of discrete‐event systems

Bertoglio

Lamperti

Zanella

et al. 2019

Engineering Reports

View full text Add to dashboard Cite

Diagnosis of discrete‐event systems (DESs) is computationally complex. This is why a variety of knowledge compilation techniques have been proposed, the most notable of them rely on a diagnoser. However, the construction of a diagnoser requires the generation of the whole system space, thereby making the approach impractical even for DESs of moderate size. To avoid total knowledge compilation while preserving efficiency, a twin‐engined diagnosis technique is proposed in this paper, which is inspired by the two operational modes of the human mind. If the symptom of the DES is part of the knowledge or experience of the diagnosis engine, then Engine 1 allows for efficient diagnosis. If, instead, the symptom is unknown, then Engine 2 comes into play, which is far less efficient than Engine 1. Still, the experience acquired by Engine 2 is then integrated into the symptom dictionary of the DES. This way, if the same diagnosis problem arises anew, then it will be solved by Engine 1 in linear time. The symptom dictionary can also be extended by specialized knowledge coming from scenarios, which are the most critical/probable behavioral patterns of the DES, which need to be diagnosed quickly.

show abstract

Section: Diagnosis and The Myth Of Total Knowledge Compilationmentioning

confidence: 99%

“…6 However, in many industrial applications, faults are rare events and available training data from faulty conditions is usually limited. 7 Hence, collecting a sufficient amount of data from relevant faulty situations for data-driven diagnosis is a time-consuming and expensive process.…”

Section: Diagnosis and The Myth Of Total Knowledge Compilationmentioning

confidence: 99%

Twin‐engined diagnosis of discrete‐event systems

Bertoglio

Lamperti

Zanella

et al. 2019

Engineering Reports

View full text Add to dashboard Cite

show abstract

“…For example, very recently, a fault diagnosis method for rolling bearing based on semi-supervised clustering and SVDD is presented in [21]. And in [22], A hybrid method combing model-based diagnosis and SVDD-based anomaly detector is proposed to identify unknown faults and also classify multiple-faults in an internal combustion engine using only single-fault training data. In [23], a method for power electronic circuits fault classification is proposed based on the SVDD classifier.…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Fault Localization in Distribution Systems with Distributed Energy Resources

et al. 2020

View full text Add to dashboard Cite

The integration of Distributed Energy Resources (DERs) introduces a non-conventional two-way power flow which cannot be captured well by traditional model-based techniques. This brings an unprecedented challenge in terms of the accurate localization of faults and proper actions of the protection system. In this paper, we propose a data-driven fault localization strategy based on multi-level system regionalization and the quantification of fault detection results in all subsystems/subregions. This strategy relies on the tree segmentation criterion to divide the entire system under study into several subregions, and then combines Support Vector Data Description (SVDD) and Kernel Density Estimation (KDE) to find the confidence level of fault detection in each subregion in terms of their corresponding p-values. By comparing the p-values, one can accurately localize the faults. Experiments demonstrate that the proposed data-driven fault localization can greatly improve the accuracy of fault localization for distribution systems with high DER penetration.

show abstract

“…Even though there are tools to systematically identify all these fault classes early in the system development phase, see for example [3], it is still a difficult task, especially for large-scale or complex systems. Therefore, there can be unknown faults that are not taken into consideration when training the diagnosis system [4].…”

Section: Introductionmentioning

confidence: 99%

“…Different faults can have similar effects on system dynamics resulting in fault classification ambiguities. Therefore, it is not desirable that a data-driven classifier only selects one fault class, since the true fault could be missed, but should instead identify and rank all plausible fault classes [4]. This type of information is useful, for example, at the workshop to support a technician during troubleshooting [7].…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Open-Set Fault Classification of Residual Data Using Bayesian Filtering

Jung

2020

IEEE Trans. Contr. Syst. Technol.

Self Cite

View full text Add to dashboard Cite

Data-driven fault classification in industrial applications is complicated by unknown fault classes and limited training data. In addition, different faults can have similar effects on sensor outputs resulting in fault classification ambiguities, i.e. multiple fault hypotheses can explain the data. One solution is to identify and rank all plausible fault classes which give useful information, for example at a workshop when performing troubleshooting. A probabilistic fault classification algorithm is proposed for residual data classification combining Weibullcalibrated one-class support vector machines for fault class modeling and Bayesian filtering for time-series analysis. The fault classifier ranks different fault classes and can identify sequences from unknown fault realizations, i.e. faults not represented in training data. Real residual data computed from sensor data and model analysis of an internal combustion engine are used as a case study illustrating the usefulness of the proposed method.

show abstract

Combining model-based diagnosis and data-driven anomaly classifiers for fault isolation

Cited by 58 publications

References 30 publications

Twin‐engined diagnosis of discrete‐event systems

Twin‐engined diagnosis of discrete‐event systems

Data-Driven Fault Localization in Distribution Systems with Distributed Energy Resources

Data-Driven Open-Set Fault Classification of Residual Data Using Bayesian Filtering

Contact Info

Product

Resources

About