The use of Kohonen self-organizing maps in process monitoring

Vermasvuori, Mikko; Enden, P.; Haavisto, S.; Jämsä‐Jounela, Sirkka‐Liisa

doi:10.1109/is.2002.1042576

Cited by 18 publications

(7 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Initialisation helps the algorithm to converge sooner to a good result in which weight vectors are given values either randomly or linearly. In this process, each neuron is assigned random weight vectors generally between zero and one (Vermasvuori et al., 2002). The central aim of training is to establish the Best Matching Unit (BMU) or winning node from the map units for each input prototype.…”

Section: Methodsmentioning

confidence: 99%

Anaerobic digestion process modeling using Kohonen self-organising maps

Ramachandran

Rustum

Adeloye

2019

Heliyon

View full text Add to dashboard Cite

Anaerobic digestion is a versatile method for wastewater treatment as it not only reduces the waste but also leads to production of renewable energy. Modeling of the anaerobic process requires knowledge of biological and physico-chemical conditions, bacterial growth kinetics, substrate utilization, and product synthesis. However, the complexity of the process calls for highly sophisticated models requiring very high level of expertise and knowledge in the subject. This paper presents an approach for modeling of anaerobic digestion process through which the correlation between various process parameters can be studied, knowledge can be extracted, and system behaviour can be predicted. The datasets have been generated using a synthetic Matlab-Simulink-Excel model and process modelling is done using Kohonen Self organizing maps (KSOM). The resulting KSOM provided a visual interpretation of the inter-relationships between parameters (OLR, Sac, pH, Shco3, Q, Sglu_in, Qgas_out, Sglu_out, and Sch4_gas_out) which would help semi-skilled operators for operation and control of such plants. The model accurately predicts the variations in methane and total gas output with respect to changes in input parameters as the correlation is more than 90% for most of the parameters. This methodology offers a platform for scientists and researchers in comprehending the system behaviour under various operating conditions, even with missing data.

show abstract

Section: Methodsmentioning

confidence: 99%

Anaerobic digestion process modeling using Kohonen self-organising maps

Ramachandran

Rustum

Adeloye

2019

Heliyon

View full text Add to dashboard Cite

show abstract

“…In these approaches, the most prominent technique is that of artificial neural networks. On the basis of measured historical process variables (training data), neural networks can use a learning algorithm to acquire the static andlor dynamic transmission behaviour of the process [4,7]. The advantage of this approach is that no analytically formulated process model is needed a priori, and thus the developer is not hampered by "unsafe" assumptions about the physical interactions of the process that are present when modelling.…”

Section: Diagnosis and Monitoring Of Process Behaviourmentioning

confidence: 99%

Monitoring of complex industrial processes based on self-organizing maps and watershed transformations

Frey

2012

2012 IEEE International Conference on Industrial Technology

View full text Add to dashboard Cite

An efficient operation of complex industrial processes requires the continuous diagnosis of the asset functionality. The early detection of potential failures and malfunctions, the identification and localization of present or impending component failures and, in particular, the monitoring of the underlying physical process behaviour is of crucial importance for a cost-effective operation of complex industry assets. With respect to these suppositions a monitoring concept based on machine learning methods has been developed, which allows an integrated and continuous diagnosis of the physical process behavior and phases. The present paper outlines briefly the architecture of the developed distributed diagnostic concept and presents in detail the developed approach for the identification of intrinsic process-phases and the monitoring functionality of the unknown process behaviour based on self-organizing-maps and watershed transformations

show abstract

“…In these approaches, the most prominent technique is that of artificial neural networks. On the basis of measured historical process variables (training data), neural networks can use a learning algorithm to acquire the static and/or dynamic transmission behaviour of the process [4,7]. The advantage of this approach is that no analytically formulated process model is needed a priori, and thus the developer is not hampered by "unsafe" assumptions about the physical interactions of the process that are present when modelling.…”

Section: Diagnosis and Monitoring Of Process Behaviourmentioning

confidence: 99%

Process Diagnosis and Monitoring of Field Bus based Automation Systems using Self-Organizing Maps and Watershed Transformations

Frey¹

2009

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

A cost-effective operation of complex automation systems requires the continuous diagnosis of the asset functionality. The early detection of potential failures and malfunctions, the identification and localization of present or impending component failures and, in particular, the monitoring of the underlying physical process are of crucial importance for the efficient operation of complex process industry assets. With respect to these suppositions a software agent based diagnosis and monitoring concept has been developed, which allows an integrated and continuous diagnosis of the communication network and the underlying physical process behavior. The present paper outlines the architecture of the developed distributed diagnostic concept based on software agents and presents the functionality for the diagnosis of the unknown process behaviour of the underlying automation system based on machine learning methods. I. MOTIVATIONECHNOLOGICAL progress in automation system engineering presents great challenges for system operators and especially for implementation and maintenance personnel. Functions which have hitherto been performed by mechanical or electromechanical devices are being increasingly replaced by software-based mechatronic systems. Functions that were previously visible and easily understood are now implemented using software and digital communications systems. Information that was previously transmitted, for example, as analogue electrical signals and was thereby simple to verify, is now transmitted as digital messages in fast data networks. Due to the decentralization of intelligence and the transition to distributed automation architectures, the engineering and implementation of systems is becoming ever more complex. The associated costs meanwhile exceed the basic hardware costs of the components by a large factor. It is here that effective diagnostic concepts offer great potential for cost savings. Verification that a system is functioning correctly, fault diagnosis, early detection of impending component failures and, in particular, monitoring of the process are of crucial importance for the cost-effective operation of complex automated processes.The continuous diagnosis of the functionality, the early

show abstract

The use of Kohonen self-organizing maps in process monitoring

Cited by 18 publications

References 7 publications

Anaerobic digestion process modeling using Kohonen self-organising maps

Anaerobic digestion process modeling using Kohonen self-organising maps

Monitoring of complex industrial processes based on self-organizing maps and watershed transformations

Process Diagnosis and Monitoring of Field Bus based Automation Systems using Self-Organizing Maps and Watershed Transformations

Contact Info

Product

Resources

About