Design of a sensor based on neural networks to determine sludge level of industrial thickeners

Mauricio A Leiva, Claudio A Acuña and Claudio A Leiva

International Journal of Distributed Sensor Networks, 2020, vol. 16, issue 6, 1550147720933153

Abstract: In copper mining, there are two main separation processes: leaching for oxidized minerals and flotation for sulfide minerals. In Chile, the increase of sulfide minerals in the deposits and the decrease of oxidized minerals have led to an increase in the investigation of flotation processes and the optimization of their associated operations. One of the concentration processes is the use of thickeners, whose main objective is to treat the tailings that leave the plants in pulp form with approximately 30% solids and obtain a pulp with a concentration greater than 50% with clear water flow. The recovery of water is the main goal, so knowing the concentration profiles of solids and sedimentation is crucial. However, the characteristics of the pulp and the operation of the thickeners are complex because a great variety of forms can be found in the concentration profile of said pulp. This limits conventional measurement techniques and makes it difficult to apply deterministic models to the solids profile, given the high nonlinearities and variability of the system. In this article, a solution is proposed by developing a sensor that allows the online estimation of sludge level and concentration of solids, based on a model of neural networks (with the model of Maxwell for dispersions), allowing to measure the solids profile regardless of the operating conditions. The structure selected has nine inputs with a hidden layer, two neurons, and two outputs being trained with tailings from Chuquicamata, obtaining the information from a 50 L pilot with 10-electrode bar of 60 cm of length, resulting in an estimation error of 0.8 cm with a network of 26 parameters.

Keywords: Neural networks; knowledge-based system; thickeners; sludge level (search for similar items in EconPapers)
Date: 2020
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://journals.sagepub.com/doi/10.1177/1550147720933153 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:sae:intdis:v:16:y:2020:i:6:p:1550147720933153

DOI: 10.1177/1550147720933153

Access Statistics for this article

More articles in International Journal of Distributed Sensor Networks
Bibliographic data for series maintained by SAGE Publications ().