A novel algorithm for dead time estimation between soft sensor inputs and output

Mejic, Luka; Stanisic, Darko; Ilić, Vojin; Jakovljević, Boris B.; Jorgovanović, Nikola

doi:10.1016/j.compchemeng.2019.02.012

Cited by 18 publications

(5 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The algorithm presented in this paper is focused on the model development of a soft sensor for which inputs are already selected, so the input selection can be done in any way that is suitable for the observed process. However, we find that the approach for input selection, as well as data preprocessing, presented in [ 30 ] is highly efficient for continuous processes, since it provides a way to choose signals and their time delays that contain the most information about the desired soft sensor output. It describes in detail how data for this type of industrial process should be preprocessed, and, after that, presents the automated algorithm that will select the inputs and their time delays, which should be used for a soft sensor.…”

Section: Casy Study: Cement Fineness Estimationmentioning

confidence: 99%

“…It describes in detail how data for this type of industrial process should be preprocessed, and, after that, presents the automated algorithm that will select the inputs and their time delays, which should be used for a soft sensor. In this case, study inputs and their time delays were selected by using the approach from [ 30 ] which resulted in the set of inputs presented in Table 1 .…”

Section: Casy Study: Cement Fineness Estimationmentioning

confidence: 99%

See 1 more Smart Citation

An Algorithm for Soft Sensor Development for a Class of Processes with Distinct Operating Conditions

Stanišić,

Mejić,

Jorgovanović

et al. 2024

Sensors

Self Cite

View full text Add to dashboard Cite

Soft sensors are increasingly being used to provide important information about production processes that is otherwise only available through off-line laboratory analysis. However, usually, they are developed for a specific application, for which thorough process analysis is performed to provide information for the appropriate selection of model type and model structure. Wide industrial application of soft sensors, however, requires a method for soft sensor development that has a high level of automatism and is applicable to a significant number of industrial processes. A class of processes that is very common in the industry are processes with distinct operating conditions. In this paper, an algorithm that is suitable for the development of soft sensors for this class of processes is presented. The algorithm possesses a high level of automatism, as it requires minimal user engagement regarding the structure of the model, which makes it suitable for implementation as a customary industrial solution. The algorithm is based on a radial basis function artificial neural network, and it enables the automatic selection of the model structure and the determination of model parameters, only based on the training data set. The testing of the presented algorithm is done on the cement production process, since it represents a process with distinct operating conditions. The results of the test show that, besides providing a high level of automatism in model development, the presented algorithm generates a soft sensor with high estimation performance.

show abstract

Section: Casy Study: Cement Fineness Estimationmentioning

confidence: 99%

Section: Casy Study: Cement Fineness Estimationmentioning

confidence: 99%

An Algorithm for Soft Sensor Development for a Class of Processes with Distinct Operating Conditions

Stanišić,

Mejić,

Jorgovanović

et al. 2024

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…O tempo necessário para que o material venha da entrada do processo até a saída é significativo, portanto é preciso considerar um atraso para as variáveis relacionadas à entrada. Um algoritmo desenvolvido por (MEJI 2019) foi utilizado para extrair os atrasos correspondentes a cada variável de entrada correlacionada à finura do cimento. Por outro lado, de acordo com (STANIšI 2015) a introdução do atraso de tempo em variável deve basear-se na experiência e conhecimento sobre o processo, em colaboração com especialistas da planta.…”

Section: Coleta E Pré-processamento Dos Dadosunclassified

Previsão on-Line Da Finura Do Cimento Em Moinho De Bolas Usando Sensor Virtual

Andreatta,

Apóstolo,

Nunes

2023

Tecnologias E Inovações Na Construção Civil: Os Avanços Em Técnicas, Materiais E Processos De Construção

View full text Add to dashboard Cite

Este capítulo apresenta três modelos de regressão supervisionada para previsão da finura do cimento: rede neural perceptron multicamadas (RNPMC), máquina de vetores de suporte (SVM- Support Vector Machine) e rede neural de função de base radial (RNFBR). Para o treinamento dos modelos, as variáveis de entrada foram selecionadas através do estudo do processo de moagem do cimento e, posteriormente, foram escolhidos dois subconjuntos a partir dessas informações adquiridas por meio de métodos estatísticos. A variável de saída foi o resultado da análise laboratorial da finura do cimento. A etapa de pré-processamento que abrange a extração, análise, tratamento e limpeza dos dados recebidos da fábrica foi realizada de forma intensiva, o que ajudou a obter excelentes resultados na estimativa da finura do cimento. Este trabalho descreve o projeto e implementação de um sensor virtual baseado no modelo RNPMC treinado com o algoritmo de retropropagação de Adam e também no SVM para prever a finura do cimento online em um moinho de bolas. Os sistemas desenvolvidos foram implementados como sensores virtuais e foram efetivamente testados em um processo de moagem de cimento, denominado processo 1. Eles demonstraram a capacidade de fornecer informações sobre as variáveis anteriormente obtidas apenas por meio de testes laboratoriais off-line, com destaque para o sensor virtual baseado no modelo RNPMC. Os sensores virtuais desenvolvidos podem oferecer novas oportunidades complementares na redução do consumo de energia e também na melhoria da qualidade do produto final, entre outras vantagens para o processo de moagem.

show abstract

“…The ability of soft sensors to address the high cost and time lag associated with the direct measurement of these variables renders them attractive alternatives. In addition, the rapid response of soft sensors provides valuable information for process control and monitoring (Mejić et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

A data-driven soft-sensing approach using probabilistic latent variable model with functional data framework

Tan

Guo

Liu

et al. 2023

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

Functional principal component analysis (FPCA) and functional partial least squares (FPLS) are two mainstream functional data analysis (FDA) methods, which have been commonly used to extract deep information hidden in the original data space. However, the process data always contain random noise, which affects the performance of FDA models. To overcome this issue, two functional probabilistic latent variable models (FPLVMs), including functional probabilistic principal component analysis (FPPCA) and functional probabilistic partial least squares (FPPLS) are proposed in this work. First, the process data are converted into functional data using the FDA. Subsequently, a log-likelihood function considering the noise factor and functional latent variables is designed. Finally, the regression model parameters are estimated using an expectation–maximisation algorithm. In contrast to FPPCA, FPPLS decomposes the process data and the key variable with constrained latent variables, which is similar to the partial least squares (PLS) and the principal component analysis (PCA). Moreover, the degeneration mechanism from FPLVMs into probabilistic latent variable models and latent variable models is discussed. An adaptive strategy with functional covariance is used to satisfy the online predictive capabilities of the model. Finally, the proposed approach is validated using a numerical case, the Tennessee Eastman process and an industrial o-xylene distillation column for evaluation.

show abstract

A novel algorithm for dead time estimation between soft sensor inputs and output

Cited by 18 publications

References 25 publications

An Algorithm for Soft Sensor Development for a Class of Processes with Distinct Operating Conditions

An Algorithm for Soft Sensor Development for a Class of Processes with Distinct Operating Conditions

Previsão on-Line Da Finura Do Cimento Em Moinho De Bolas Usando Sensor Virtual

A data-driven soft-sensing approach using probabilistic latent variable model with functional data framework

Contact Info

Product

Resources

About