Application of Uncertainty Analysis of Artificial Neural Networksfor Predicting Coagulant and Alkalizer Dosages in a Water Treatment Process

Menezes, Fábio Conceição de; Fontes, Raony M.; Oliveira-Esquerre, Karla Patrícia; Kalid, Ricardo

doi:10.1590/0104-6632.20180354s20170039

Cited by 11 publications

(9 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A preliminary sensitivity analysis of the model was performed to understand which input variables have a higher influence on the output prediction and to reduce the number of input variables, removing the variables with lower impact on the model output . Monte Carlo simulations were employed for the sensitivity analysis because they are frequently used to quantify model uncertainty from the model parameters, input data, or model structure . The model output uncertainty was expressed in terms of prediction intervals which are typically related to the quantiles of the model output distribution.…”

Section: Methodsmentioning

confidence: 99%

“…31 Monte Carlo simulations were employed for the sensitivity analysis because they are frequently used to quantify model uncertainty from the model parameters, input data, or model structure. 32 The model output uncertainty was expressed in terms of prediction intervals which are typically related to the quantiles of the model output distribution. Details on the Monte Carlo simulation methodology are reported in the Supporting Information, §4.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Soft Sensing for On-Line Fault Detection of Ammonium Sensors in Water Resource Recovery Facilities

Cecconi¹,

Rosso

2021

Environ. Sci. Technol.

View full text Add to dashboard Cite

The increasing demand for online sensors applied to advanced control strategies in water resource recovery facilities has resulted in the increasing investigation of fault-detection methods to improve the reliability of sensors installed in harsh environments. The study herein focuses on the fault detection of ammonium sensors, especially for effluent monitoring, given their potential in ammonium-based aeration control applications. An artificial neural network model was built to predict the ammonium content in the effluent by employing the information from five other sensors installed in the activated sludge tank: NH4 +, pH, ORP, DO, and TSS. The residual between the model prediction and the effluent ammonium sensor signal was utilized in a fault-detection mechanism based on principal component analysis and Shewhart monitoring charts. In contrast to previous studies, the present work utilizes typical faults collected from a 1 year historic dataset of an actual sensor setup. Treatment process anomalies, calibration bias faults, and fouling drifts were the most common issues identified from the historic dataset, and they were promptly identified by the proposed fault-detection methodology. Once a fault is detected, the model prediction can be actively used in place of the sensor for process control without affecting the treatment process by utilizing faulty datasets.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Soft Sensing for On-Line Fault Detection of Ammonium Sensors in Water Resource Recovery Facilities

Cecconi¹,

Rosso

2021

Environ. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…The surrogates with the greatest influence on the prediction outcome are the recommended surrogates for viral risk indication. Monte Carlo simulation can then be applied to quantify model uncertainty from model parameters, input data, or model structure [157].…”

Section: Modeling Of Infectious Viruses Using Artificial Neural Networkmentioning

confidence: 99%

Integrating Virus Monitoring Strategies for Safe Non-Potable Water Reuse

Jiang

Bischel

Goel

et al. 2022

Water

View full text Add to dashboard Cite

Wastewater reclamation and reuse have the potential to supplement water supplies, offering resiliency in times of drought and helping to meet increased water demands associated with population growth. Non-potable water reuse represents the largest potential reuse market. Yet, economic constraints for new water reuse infrastructure and safety concerns due to microbial water quality, especially viral pathogen exposure, limit the widespread implementation of water reuse. Cost-effective, real-time methods to measure or indicate the viral quality of recycled water would do much to instill greater confidence in the practice. This manuscript discusses advancements in monitoring and modeling viral health risks in the context of water reuse. First, we describe current wastewater reclamation processes and treatment technologies with an emphasis on virus removal. Second, we review technologies for the measurement of viruses, both culture- and molecular-based, along with their advantages and disadvantages. We outline promising viral surrogates and specific pathogenic viruses that can serve as indicators of viral risk for water reuse. We suggest metagenomic analyses for viral screening and flow cytometry for quantification of virus-like particles as new approaches to complement more traditional methods. Third, we describe modeling to assess health risks through quantitative microbial risk assessments (QMRAs), the most common strategy to couple data on virus concentrations with human exposure scenarios. We then explore the potential of artificial neural networks (ANNs) to incorporate suites of data from wastewater treatment processes, water quality parameters, and viral surrogates. We recommend ANNs as a means to utilize existing water quality data, alongside new complementary measures of viral quality, to achieve cost-effective strategies to assess risks associated with infectious human viruses in recycled water. Given the review, we conclude that technologies will be ready to identify and implement viral surrogates for health risk reduction in the next decade. Incorporating modeling with monitoring data would likely result in a more robust assessment of water reuse risk.

show abstract

“…On the other hand, the literature has presented several advances in evaluating uncertainty in AI models. Menezes et al [19] employ the uncertainty propagation law of distributions proposed by Bipm et al [5] (Supplement 1 to the Guide to the expression of uncertainty in measurement (GUM-S1)) to evaluate the uncertainty of static models of artificial neural networks (ANN). Abdar et al [1] , in its turn, present a review of uncertainty analysis methods for deep learning models.…”

Section: Introductionmentioning

confidence: 99%

An uncertainty approach for Electric Submersible Pump modeling through Deep Neural Network

Almeida Costa,

de Menezes Rebello,

Viena Santana

et al. 2024

Heliyon

View full text Add to dashboard Cite

Application of Uncertainty Analysis of Artificial Neural Networksfor Predicting Coagulant and Alkalizer Dosages in a Water Treatment Process

Abstract: Artificial neural networks (ANNs) were built to predict coagulant (Model I) and alkalizer (Model II) dosages given raw and treated water parameters from a water clarifying process. Different ANN architectures were tested and optimal results were obtained with [

Cited by 11 publications

References 29 publications

Soft Sensing for On-Line Fault Detection of Ammonium Sensors in Water Resource Recovery Facilities

Soft Sensing for On-Line Fault Detection of Ammonium Sensors in Water Resource Recovery Facilities

Integrating Virus Monitoring Strategies for Safe Non-Potable Water Reuse

An uncertainty approach for Electric Submersible Pump modeling through Deep Neural Network

Contact Info

Product

Resources

About