Incipient fault diagnosis of chemical processes via artificial neural networks

Watanabe, Kajiro; Matsuura, I.; Abe, Masahiro; Kubota, Mamoru; Himmelblau, D. M.

doi:10.1002/aic.690351106

Cited by 278 publications

(64 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In chemical engineering, Watanabe et al (1989); Venkatasubramanian and Chan (1989); Ungar et al (1990) were among the first to demonstrate the usefulness of neural networks for the problem of fault diagnosis. Later, Venkatasubramanian et al (1990) presented a more detailed and thorough analysis of the learning, recall and generalization characteristics of neural networks for detecting and diagnosing process failures in steady-state processes.…”

Section: Supervised Methodsmentioning

confidence: 99%

Sequential anomalies: a study in the Railway Industry

2016

View full text Add to dashboard Cite

Concerned with predicting equipment failures, predictive maintenance has a high impact both at a technical and at a financial level. Most modern equipments have logging systems that allow us to collect a diversity of data regarding their operation and health. Using data mining models for anomaly and novelty detection enables us to explore those datasets, building predictive systems that can detect and issue an alert when a failure starts evolving, avoiding the unknown development up to breakdown. In the present case, we use a failure detection system to predict train door breakdowns before they happen using data from their logging system. We use sensor data from pneumatic valves that control the open and close cycles of a door. Still, the failure of a cycle does not necessarily indicates a breakdown. A cycle might fail due to user interaction. The goal of this study is to detect structural failures in the automatic train door system, not when there is a cycle failure, but when there are sequences of cycle failures. We study three methods for such structural failure detection: outlier detection, anomaly detection and novelty detection, using different windowing strategies. We propose a two-stage approach, where the output of a point-anomaly algorithm is post-processed by a low-pass filter to obtain a subsequence-anomaly detection. The main result of the two-level architecture is a strong impact in the false alarm rate.

show abstract

Section: Supervised Methodsmentioning

confidence: 99%

Sequential anomalies: a study in the Railway Industry

2016

View full text Add to dashboard Cite

show abstract

“…Various applications of ANN are an approach to fault diagnosis in chemical processes [10]. fault diagnosis in complex chemical plants [11], incipient fault diagnosis of chemical process [12], leak detection in liquefied gas pipeline (AIChE 44 (12), (1998)) [13], leak detection in liquefied gas pipeline (Chem. Engg.…”

Section: Introductionmentioning

confidence: 99%

Osmotic Drying Rate Estimation for Dehydration of Beetroot Slices using Artificial Neural Network

Pandharipande¹,

Bele²

2013

IJCA

View full text Add to dashboard Cite

Osmotic dehydration can be viewed as an alternative method for drying of food materials with advantages of retention of gloss, texture & colour of dried products. Artificial neural network is emerging as a modeling tool for complex operations involving non linear multivariable relationships. The present work is aimed at estimation of the osmotic drying rates & weight reduction of beetroot slices as a function of concentration of sodium chloride, time & temperature using artificial neural network. Based on the observations, results & discussion, it can be said that, beetroot slices can be partially dewatered by osmotic dehydration in salt solution and percent weight loss is from 10 to 29 % depending upon the operating parameters. It can be concluded that the present work has successfully demonstrated the potential of ANN in modeling of osmotic dehydration of beetroot slices with high accuracy.

show abstract

“…Various applications of ANN are, an approach to fault diagnosis in chemical processes (8) , fault diagnosis in complex chemical plants (9) , incipient fault diagnosis of chemical process (10) , leak detection in liquefied gas pipeline (11,12) , for estimation of mass transfer coefficient for fast fluidized bed solids (13) , modeling of distillation column (14) , detergent formulation (15) , modeling of unsteady heat conduction in semi infinite solid (16) , prediction of mass transfer coefficient in down flow jet loop reactor (17) and modeling of packed column (18) and similar other (19,20) were also reported.…”

Section: Introductionmentioning

confidence: 99%

Artificial Neural Network Modeling for Adsorption of Dyes from Aqueous Solution using Rice Husk Carbon

Khonde¹,

Pandharipande²

2012

IJCA

View full text Add to dashboard Cite

Adsorption is one of the important industrial processes used for removal of colour, odour, turbidity & reduction of COD. In adsorption, the solute present in dilute concentration in liquid or gas phase is removed by contacting with suitable solid adsorbent so that the transfer of the component first takes place on the surface of solid and then into the pore of the solid.The present work is aimed at exploring rice husk carbon as low cost adsorbent for removal of various dyes from synthetically prepared aqueous solutions and modeling the adsorption process using artificial neural network.Rice husk carbon, developed in the present work at the laboratory scale, is analyzed for its BET surface area & is observed to be very effective for removal of dyes, namely bromocresol red, alizarin red, malachite green and methylene blue from their aqueous solutions. Further, the RHC developed. The present work highlighted the efficacy of ANN as an effective tool in modeling adsorption.The ANN models developed are using elite-ANN © .The architecture of artificial neural network is initialized & training has been carried out using the experimental data. The trained neural network is used to predict output for the given set of input parameters.

show abstract

Incipient fault diagnosis of chemical processes via artificial neural networks

Cited by 278 publications

References 12 publications

Sequential anomalies: a study in the Railway Industry

Sequential anomalies: a study in the Railway Industry

Osmotic Drying Rate Estimation for Dehydration of Beetroot Slices using Artificial Neural Network

Artificial Neural Network Modeling for Adsorption of Dyes from Aqueous Solution using Rice Husk Carbon

Contact Info

Product

Resources

About