A novel calibration approach of soft sensor based on multirate data fusion technology

Wu, Yao; Luo, Xionglin

doi:10.1016/j.jprocont.2010.09.003

Cited by 50 publications

(17 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Geralmente, quanto mais fácil de medir uma variável, menor será o tempo de amostragem. Como geralmente o sensor virtual é implementado para prever o valor de variáveis difíceis de medir, muitas vezes é preciso fazer um sincronismo entre estes tempos de amostragem, reduzindo o tempo de amostragem da variável de entrada do sistema [5]. No caso de dados perdidos, é comum observar que durante a amostragem de alguma variável, determinadas amostras podem vir sem valor nenhum.…”

Section: Coleta E Filtragem De Dadosunclassified

Desenvolvimento De Um Sensor Virtual Para Monitoramento De Índice De Wobbe Na Planta Do LTQ Da Arcelormittal Tubarão

Venturim¹,

Ferreira²,

Pereira³

et al. 2018

ABM Proceedings

View full text Add to dashboard Cite

Resumo Este trabalho estuda o desenvolvimento de um sensor virtual para monitoramento de índice de wobbe do gás misto (MG) utilizados em fornos de reaquecimento em indústrias siderúrgicas. Para isso foram testados quatro modelos de regressão. Dois deles sendo regressões por mínimos quadrados parciais (PartialLeastSquares, PLS), um utilizando o algoritmo NIPALS (NonlinearIterativePartialLeastSquares) e outro utilizando o algoritmo SIMPLS (simple partial least squares). As demais regressões foram por componentes principais (PCR) e regressão por vetor de suporte (SVR). As variáveis do processo foram selecionadas inicialmente com a ajuda de um especialista do processo e posteriormente pelo método de árvore de decisão. Os valores estimados são comparados com os valores reais obtidos na planta real. Os resultados mostram que foi possível prever o índice de wobbe com erro médio quadrático com valor de 14,43%.

show abstract

Section: Coleta E Filtragem De Dadosunclassified

Desenvolvimento De Um Sensor Virtual Para Monitoramento De Índice De Wobbe Na Planta Do LTQ Da Arcelormittal Tubarão

Venturim¹,

Ferreira²,

Pereira³

et al. 2018

ABM Proceedings

View full text Add to dashboard Cite

show abstract

“…The dataset consists of 149 samples for training and 62 samples for test. Four dynamic modeling methods are introduced for comparison purposes: 1) method proposed in this brief; 2) dynamic ANN method proposed by Wu and Luo [11], simply referred to as DANN where the model is optimized by the enhanced particle swam optimization (EPSO) method; 3) IR template (IRT) model with MDE optimization proposed in [9], which is simply referred to as IRT; and 4) static modeling using kernel-SVM proposed in [5].…”

Section: Simulation Case Studymentioning

confidence: 99%

“…It is, however, essentially a linear technique and fails to give desirable predictions for nonlinear processes. In alternative, a dynamic ANN is proposed to handle both dynamics and nonlinearity [10], [11]; however, its generalization ability cannot be guaranteed and overfitting is unavoidable. As a powerful technique, Bayesian inference has found wide applications in the process industry, including soft-sensor development [5], [6], [12], process monitoring [13], and fault diagnosis [14], [15].…”

Section: Introductionmentioning

confidence: 99%

Novel Bayesian Framework for Dynamic Soft Sensor Based on Support Vector Machine With Finite Impulse Response

Chen

Gao

Yang

et al. 2014

IEEE Trans. Contr. Syst. Technol.

View full text Add to dashboard Cite

Conventional data-driven soft sensors commonly rely on the assumption that processes are operating at steady states. As chemical processes involve evident dynamics, conventional soft sensors may suffer from transient inaccuracy and poor robustness. In addition, the control performance is unsatisfactory when the outputs of soft sensors serve as the feedback signals for quality control. This brief develops a dynamic soft-sensing model combining finite impulse response and support vector machine to describe dynamic and nonlinear static relationships. The model parameters are then estimated within a Bayesian framework. The results from both the simulated and the industrial case show its superiority to conventional static models in terms of dynamic accuracy and practical applicability.Index Terms-Bayesian inference, data-driven technique, dynamic soft sensor, process control, support vector machine (SVM). 1063-6536 © 2013 IEEE

show abstract

“…The proposed model and the concept of symmetry have relativity and complementarity, and the research direction is highly consistent with Symmetry, which is convenient for scholars in related fields as a reference. In chemical production, major process variables such as product quality are characterized by a slow sampling rate and time delay [1]. To ensure the stability of variable data in the main process, it is necessary to estimate the main process variables through some easily acquired process variables.…”

Section: Introductionmentioning

confidence: 99%

Dynamic Soft Sensor Development for Time-Varying and Multirate Data Processes Based on Discount and Weighted ARMA Models

Dai

2019

Symmetry

View full text Add to dashboard Cite

To solve the soft sensor modeling (SSMI) problem in a nonlinear chemical process with dynamic time variation and multi-rate data, this paper proposes a dynamic SSMI method based on an autoregressive moving average (ARMA) model of weighted process data with discount (DSSMI-AMWPDD) and optimization methods. For the sustained influence of auxiliary variable data on the dominant variables, the ARMA model structure is adopted. To reduce the complexity of the model, the dynamic weighting model is combined with the ARMA model. To address the weights of auxiliary variable data with different sampling frequencies, a calculation method for AMWPDD is proposed using assumptions that are suitable for most sequential chemical processes. The proposed method can obtain a discount factor value (DFV) of auxiliary variable data, realizing the dynamic fusion of chemical process data. Particle swarm optimization (PSO) is employed to optimize the soft sensor model parameters. To address the poor convergence problem of PSO, ω-dynamic PSO (ωDPSO) is used to improve the PSO convergence via the dynamic fluctuation of the inertia weight. A continuous stirred tank reactor (CSTR) simulation experiment was performed. The results show that the proposed DSSMI-AMWPDD method can effectively improve the SSM prediction accuracy for a nonlinear time-varying chemical process. The AMWPDD proposed in this paper can reflect the dynamic change of chemical process and improve the accuracy of SSM data prediction. The ω dynamic PSO method proposed in this paper has faster convergence speed and higher convergence accuracy, thus, these models correlate with the concept of symmetry.

show abstract

A novel calibration approach of soft sensor based on multirate data fusion technology

Cited by 50 publications

References 16 publications

Desenvolvimento De Um Sensor Virtual Para Monitoramento De Índice De Wobbe Na Planta Do LTQ Da Arcelormittal Tubarão

Desenvolvimento De Um Sensor Virtual Para Monitoramento De Índice De Wobbe Na Planta Do LTQ Da Arcelormittal Tubarão

Novel Bayesian Framework for Dynamic Soft Sensor Based on Support Vector Machine With Finite Impulse Response

Dynamic Soft Sensor Development for Time-Varying and Multirate Data Processes Based on Discount and Weighted ARMA Models

Contact Info

Product

Resources

About