Fault diagnosis in chemical processes with application of hierarchical neural networks

Rusinov, Leon A.; Rudakova, I.V.; Remizova, Olga A.; Kurkina, V.V.

doi:10.1016/j.chemolab.2008.09.004

Cited by 20 publications

(18 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…That is why, it is important to ensure small time for network retraining in the case of possible change in the normal process state (for example, as a result of a drift of the normal state of the process). This can be achieved by introducing a hierarchy in the model structure . The latter entails a necessity for process decomposition and separation of a number of structural units (blocks), which should be as autonomous as possible.…”

Section: The Diagnostic Model Structurementioning

confidence: 99%

“…Moreover, if any unconsidered situation occurs, HL‐NN isolates a faulty part of the process. Although LL‐NN is probably unable to diagnose the actual cause of the fault, some useful information concerning the fault location could be obtained …”

Section: The Diagnostic Model Structurementioning

confidence: 99%

“…They are the feedforward networks (FFNs) and the radial basis function networks. Although radial basis function networks have shown the better results in fault detection as compared with FFN, the gain is not so significant, especially when an essential increase in complexity is taken into account . Because of this, it was decided to use in this study the standard 3‐layer FFN with sigmoid function of neurons activation in the hidden and output layers.…”

Section: The Diagnostic Model Structurementioning

confidence: 99%

“…That is why, when the network architecture is designed, it is important to use the results of HAZOP analysis, the expert information analysis, and the available theoretical models. This gives a firm basis for the correct process decomposition and increases the probability of identifying all possible abnormal situations to be arisen in the process …”

Section: The Diagnostic Model Structurementioning

confidence: 99%

“…At the same time, the neural network (NN) approach is free of these problems . Neural networks are inherently nonlinear.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Diagnostics of sintering processes on the basis of PCA and two‐level neural network model

Egorova

Rudakova

Rusinov

et al. 2017

Journal of Chemometrics

Self Cite

View full text Add to dashboard Cite

The application of chemometrics methods to continuous monitoring and diagnostics of sintering process faults for improving iron‐ore sinter quality is considered in the article. The sintering process is among complex multivariate processes. A number of agglomeration process faults have often similar symptoms, resulting in late fault detection by an operator and as a consequence, wrong process control decisions. To support the efficient operative decision making, it is proposed to use the process fault monitoring and diagnostics system. The proposed system uses a two‐level neural network (NN) diagnostic model. The high‐level neural network is used to localize the process faults whereas their reasons are determined by the low‐level neural networks. To reduce essentially the time of HL‐NN training and retraining, the task dimension is preliminarily reduced with the principal component analysis method so that the scores obtained from initial data are fed into high‐level neural network inputs. The use of principal component analysis allowed detection of sintering process faults with T2 and Q statistics. Only upon detecting the fault, a NN diagnostic model starts working to determine the fault reason. The system algorithm provides for special measures to prevent the NN from possible “loss” of the identified fault due to operator's inactivity. To increase the diagnosis depth for controlling fault symptoms that are evident on the sinter cake surface, optical digital cameras are installed and images from them are processed with proposed algorithms on the basis of fuzzy clusterization to take into account uncertainties in the initial information.

show abstract

Section: The Diagnostic Model Structurementioning

confidence: 99%

Section: The Diagnostic Model Structurementioning

confidence: 99%

Section: The Diagnostic Model Structurementioning

confidence: 99%

Section: The Diagnostic Model Structurementioning

confidence: 99%

“…At the same time, the neural network (NN) approach is free of these problems . Neural networks are inherently nonlinear.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Diagnostics of sintering processes on the basis of PCA and two‐level neural network model

Egorova

Rudakova

Rusinov

et al. 2017

Journal of Chemometrics

Self Cite

View full text Add to dashboard Cite

show abstract

An Extension Sample Classification‐Based Extreme Learning Machine Ensemble Method for Process Fault Diagnosis

Chen

Zhu

2014

Chem Eng & Technol

View full text Add to dashboard Cite

In order to achieve higher accuracy and faster response in complex process fault diagnosis, an extension sample classification‐based extreme learning machine ensemble (ESC‐ELME) method is proposed. In the realization process, the extension sample classification is used to divide the fault types. For each fault type, a specific extreme learning machine (ELM) is established and trained independently. Then, all specific ELMs are integrated to determine which fault is happened by the majority voting method. The proposed ESC‐ELME method is compared with the traditional ELM and a duty‐oriented hierarchical artificial neural network in fault diagnosis of the Tennessee Eastman process. The results demonstrate that the proposed method provides higher diagnosis accuracy and faster response.

show abstract

Stationäre Siebklassierung

Hennig

Teipel

2016

Chemie Ingenieur Technik

View full text Add to dashboard Cite

Eines der wichtigsten mechanischen Trennverfahren ist die Siebklassierung. Zur Untersuchung der Trenncharakteristik soll diese modelliert werden. Der Trenngrad wird aus den gemessenen Partikelverteilungen und den Gesamtmassen der einzelnen Fraktionen bestimmt. Die Modelle sind von jeweils drei bzw. vier Funktionsparametern abhängig. Dazu gehören der Bypass‐Parameter a, der Parameter xcut, der die Median‐Trenngrenze definiert, sowie der Parameter α als die Trennschärfe der Klassierung. Zusätzlich wird ein dritter Schärfeparameter β eingeführt. Als Parameterstudie stehen durchgeführte Siebversuchsreihen an zwei Taumelsiebmaschinen zur Verfügung.

show abstract

Fault diagnosis in chemical processes with application of hierarchical neural networks

Cited by 20 publications

References 17 publications

Diagnostics of sintering processes on the basis of PCA and two‐level neural network model

Diagnostics of sintering processes on the basis of PCA and two‐level neural network model

An Extension Sample Classification‐Based Extreme Learning Machine Ensemble Method for Process Fault Diagnosis

Stationäre Siebklassierung

Contact Info

Product

Resources

About