|
This article is cited in 1 scientific paper (total in 1 paper)
Models of self-organizing artificial neural networks for determination of stationary permanent industrial sources of air pollution
S. P. Dudarov Mendeleev University of Chemical Technology of Russia, Moscow
Abstract:
It is considered a problem of determination of one particular or few possible pollution sources are responsible on air medium quality violation as result of the norm of maximum permissible emission excess. It is solved a model task for a group of spatially separated stationary permanent industrial sources in the work. It is presented an determination task statement and a method of its solution by two architectures of artificial neural networks: Kohonen’s networks for learning vector quantization with fixed and adaptive structures as well as adaptive resonance theory network for analog inputs (ART-2). The method consists of data clustering which is supplied by self-learning algorithms (learning without a teacher). It is given estimation equations, it is described operation algorithms of Kohonen's and adaptive resonance theory networks at different life cycle stages. It is carried on a comparative analysis of model task solution results received by each of networks.
Keywords:
artificial neural network, Kohonen's neural network, learning vector quantization, adaptive resonance theory network, self-learning, self-organizing, clustering, cluster analysis, determination of free air pollution sources.
Received: 16.07.2015 Revised: 12.01.2016
Citation:
S. P. Dudarov, “Models of self-organizing artificial neural networks for determination of stationary permanent industrial sources of air pollution”, Matem. Mod., 29:1 (2017), 33–44; Math. Models Comput. Simul., 9:4 (2017), 481–488
Linking options:
https://www.mathnet.ru/eng/mm3805 https://www.mathnet.ru/eng/mm/v29/i1/p33
|
Statistics & downloads: |
Abstract page: | 509 | Full-text PDF : | 344 | References: | 46 | First page: | 19 |
|