On neural network structure selection to solve problem of iron ore preparation process identification

The possibility of creation the neural network model for predicting the iron content in the output product of grinding process of the ore-dressing industry is considered in this paper. The use of this range of tools is determined by the multifactoriness and non-linearity of the process, and by the n...

Full description

Saved in:
Bibliographic Details
Published in:Journal of physics. Conference series 2019-04, Vol.1202 (1), p.12005
Main Authors: Eremenko, Yu I, Poleshchenko, D A, Tsygankov, Yu A
Format: Article
Language:English
Subjects:
Beneficiation
Grinding
Iron ores
Neural networks
Neurons
Prediction models
Predictive control
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The possibility of creation the neural network model for predicting the iron content in the output product of grinding process of the ore-dressing industry is considered in this paper. The use of this range of tools is determined by the multifactoriness and non-linearity of the process, and by the need of the model to adapt to changing parameters. The study consisted in modeling of various structures to achieve the desired quality of the model's output signal. The network with a structure of 4 neurons in the input layer, 60 neurons in a hidden layer and 1 neuron in the output layer had the best result in the generalization and accuracy of test set reproduction. The result allows to assume the possibility of using a neural network tool for the development of aggregates of technological grinding process, and the creation of a single control system with predictive models.
ISSN:1742-6588
1742-6596
DOI:10.1088/1742-6596/1202/1/012005