<b>Intelligent system for improving dosage control

Abstract

ABSTRACT. Coagulation is one of the most important processes in a drinking-water treatment plant, and it is applied to destabilize impurities in water for the subsequent flocculation stage. Several techniques are currently used in the water industry to determine the best dosage of the coagulant, such as the jar-test method, zeta potential measurements, artificial intelligence methods, comprising neural networks, fuzzy and expert systems, and the combination of the above-mentioned techniques to help operators and engineers in the water treatment process. Current paper presents an artificial neural network approach to evaluate optimum coagulant dosage for various scenarios in raw water quality, using parameters such as raw water color, raw water turbidity, clarified and filtered water turbidity and a calculated Dose Rate to provide the best performance in the filtration process. Another feature in current approach is the use of a backpropagation neural network method to estimate the best coagulant dosage simultaneously at two points of the water treatment plant. Simulation results were compared to the current dosage rate and showed that the proposed system may reduce costs of raw material in water treatment plant.Keywords: water treatment plant, process control, coagulant dosage, artificial neural networks, optimization.Inteligência artificial aplicada ao controle da dosagem de coagulante em plantas de tratamento de água RESUMO. A coagulação é um dos processos mais importantes em uma planta de tratamento de tratamento de água e é utilizado para desestabilizar as impurezas contidas na água para possibilitar a floculação. Atualmente, algumas técnicas têm sido utilizadas na área para determinar a melhor dosagem, como ensaios de jar-teste, medidas do potencial zeta e métodos de inteligência artificial, como Redes Neurais Artificiais, Fuzzy e Sistemas Especialistas, as quais permitem combinações de técnicas para auxiliar os operadores e engenheiros no processo de tratamento de água. Este artigo apresenta uma abordagem de rede neural artificial para a avaliação da dosagem de coagulante otimizada em vários cenários de qualidade da água bruta, utilizando alguns parâmetros, tais como a cor da água bruta, turbidez das águas bruta, decantada e filtrada, além de uma entrada calculada, denominada Dose Rate para monitorar o processo de filtração. Outra característica presente nesta abordagem é a utilização do método backpropagation nas redes neurais artificiais para estimar a melhor dosagem coagulante em dois pontos da instalação de tratamento de água simultaneamente. Os resultados da simulação foram comparados com a dosagem atual e apontaram que o sistema proposto pode trazer redução nos custos de matéria-prima na estação de tratamento de água.Palavras-chave: planta de tratamento de água, controle de processos, dosagem de coagulante, redes neurais artificiais, otimização.

Exploring potential dual-stage attention based recurrent neural network machine learning application for dosage prediction in intelligent municipal management

“…Its value is always non-negative, typically, a lower RMSE is better than a higher RMSE. MAE represents the average of absolute errors between predicted and observed values, as shown in formula (11), the error cancellations can be avoided and the actual prediction error can be accurately reflected. MAPE is used to measure the relative errors between the average test value and the real value on the test set, it is defined as formula (12).…”

“…8 Data-driven solutions for maintenance or efficiency improvement have been proven to have a positive impact on social and environment aspects in real-world projects, 9 with the advancement of data analytical tools and their interdisciplinary applications in environmental science and engineering, powerful computerbased ML and artificial neural networks (ANN) have shown great potential for prediction and optimization of process control for wastewater treatment. 3,10,11 With the rapid development of ML algorithms, a favourable tool for precise dosing control is provided. [12][13][14][15] Heddam 16 proposed an extremely randomized tree, random forest (RF) and multiple linear regression (MLR) for predicting coagulant dosage in the Boudouaou drinking water treatment plant (DWPT).…”

A flocculation tensor to monitor water quality using a deep learning model

“…Many water treatment plants still require the development of a dosing system to meet the effluent quality requirements, and excessive manual dosing is frequently used (Zhang 2005). This situation is not conducive to the development of modern production, water supply, and drinking water quality requirements, resulting in a range of issues such as high chemical consumption, poor economic benefits, unstable water quality, and high labor intensity for workers (dos Santos et al 2017;Imen et al 2016;Zaque et al 2018). In the water treatment industry, research on automatic control of coagulation dosing is both necessary and urgent.…”

Elaboración de un modelo neuronal artificial para la estimación de turbiedad y proposición de dosificaciones en el tratamiento de aguas residuales de la industria avícola

“…Al respecto, la literatura muestra un buen número de trabajos que abordan la problemática de la optimización del proceso de tratamiento de aguas usando redes neuronales artificiales (Fernández;Galvis, 2003;Acuña, 2008;Olanrewaju;Muyibi;Salawudeen;Aibinu, 2012;Salgado et al, 2013;Villarreal-Campos;Caicedo-Bravo, 2013;Haghiri et al, 2014;Peña-Rojas;Flóres del Pino, 2014;Prasannasangeentha, 2015;Bui;Giang-Duong;Nguyen, 2016;Barajas-Garzón;León-Luque, 2016;Peña-Rojas, 2016;Rodrigues-dos Santos;Henriques-Librantz;Días;Gozzo-Rodrigues, 2017;De Menezes;Fontes;Oliveira-Esquerre;Kalid, 2018, Haghiri;Moharramzadeh, 2018;Messaoud;Hellal;Imed;, siendo limitado en el tratamiento de aguas residuales de procesos agroindustriales y específicamente en la industria avícola.…”