Neural network model of mill-fan system elements vibration for predictive maintenance

Balabanov, Todor; Hadjiski, Mincho; Koprinkova–Hristova, Petia; Beloreshki, Simeon; Doukovska, Lyubka

doi:10.1109/inista.2011.5946102

Cited by 8 publications

(3 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(1) Predictive maintenance for the mechanical component: This is the most common topic of predictive maintenance. Research has been conducted on bearings [13], engines [14], turbines [15], fans [16], pumps [17], gearboxes [18], milling machines [19], and centrifugal pumps [18]. Usually, the mechanical component fault includes vibration, sound, or abnormal patterns of sensor data.…”

Section: Related Workmentioning

confidence: 99%

Residual Life Prediction for Induction Furnace by Sequential Encoder with s-Convolutional LSTM

et al. 2021

View full text Add to dashboard Cite

Induction furnaces are widely used for melting scrapped steel in small foundries and their use has recently become more frequent. The maintenance of induction furnaces is usually based on empirical decisions of the operator and an explosion can occur through operator error. To prevent an explosion, previous studies have utilized statistical models but have been unable to generalize the problem and have achieved a low accuracy. Herein, we propose a data-driven method for induction furnaces by proposing a novel 2D matrix called a sequential feature matrix(s-encoder) and multi-channel convolutional long short-term memory (s-ConLSTM). First, the sensor data and operation data are converted into sequential feature matrices. Then, N-sequential feature matrices are imported into the convolutional LSTM model to predict the residual life of the induction furnace wall. Based on our experimental results, our method outperforms general neural network models and enhances the safe use of induction furnaces.

show abstract

Section: Related Workmentioning

confidence: 99%

Residual Life Prediction for Induction Furnace by Sequential Encoder with s-Convolutional LSTM

et al. 2021

View full text Add to dashboard Cite

show abstract

“…It is called intrinsic plasticity (IP) and is aimed at increasing the entropy of the reservoir neurons outputs thus stabilizing its dynamic behavior. Such combined IP-RLS training was already successfully applied in [8], [9] In our previous work, we [10] applied ESN as a model for prediction of changes in vibrations tendencies of mill fan system. Here our aim is to compare this newly developed kind of RNN with historical Elman RNN architecture.…”

Section: Introductionmentioning

confidence: 99%

Recurrent Neural Networks for Predictive Maintenance of Mill Fan Systems

Koprinkova–Hristova¹,

Hadjiski²,

Doukovska³

et al. 2011

International Journal of Electronics and Telecommunications

View full text Add to dashboard Cite

In the present paper we focus on online monitoring system for predictive maintenance based on sensor automated inputs. Our subject was a device from Maritsa East 2 power plant-a mill fan. The main sensor information we have access to is based on the vibration of the nearest to the mill rotor bearing block. Our aim was to create a (nonlinear) model able to predict on time possible changes in vibrations tendencies that can be early signal for system work deterioration. For that purpose, we compared two types of recurrent neural networks: historical Elman architecture and a recently developed kind of RNN named Echo stet networks (ESN). The preliminary investigations showed better approximation and faster training abilities of ESN in comparison to the Elman network. Direction of future work will be increasing of predications time horizon and inclusion of our predictor at lower level of a complex predictive maintenance system.

show abstract

“…The standard statistical and probabilistic (Bayesian) approaches for diagnostics are inapplicable to estimate mill fan vibration state due to non-stationarity, nonergodicity and the significant noise level of the monitored vibrations. Promising results are obtained only using computational intelligence methods (fuzzy logic, neural and neuro-fuzzy networks), [15][16][17][18][19] ].…”

mentioning

confidence: 99%

CBR Approach for Technical Diagnostics of Mill Fan System

Hadjiski¹,

Doukovska²

2013

Proceeding BAS

View full text Add to dashboard Cite

Fault diagnostics shares with other process operations the realization that with powerful knowledge representation schemes one can capture the expertise of operators and control engineers all that was gained over the years of experience with process plants. Process specific knowledge can be used to improve general purpose methodologies. There is a close coupling between diagnostics and process operations design of plants. The proper design of a plant can reduce the burden upon the task of diagnostics. Also, the information from diagnostics can be used to continuously improve the performance of process operations. The information from fault diagnostics can be incorporated into the traditional solution paradigms of other process operations. The aim of this paper is to provide a Case Based Reasoning (CBR) approach for technical diagnostics of mill fan system that would particularly share information with the fault diagnostics module and also outline the nature of interaction that one can expect.

show abstract

Neural network model of mill-fan system elements vibration for predictive maintenance

Cited by 8 publications

References 8 publications

Residual Life Prediction for Induction Furnace by Sequential Encoder with s-Convolutional LSTM

Residual Life Prediction for Induction Furnace by Sequential Encoder with s-Convolutional LSTM

Recurrent Neural Networks for Predictive Maintenance of Mill Fan Systems

CBR Approach for Technical Diagnostics of Mill Fan System

Contact Info

Product

Resources

About