Mixture robust semi-supervised probabilistic principal component regression with missing input data

Memarian, Alireza; Varanasi, Santhosh Kumar; Huang, Biao

doi:10.1016/j.chemolab.2021.104315

Cited by 16 publications

(4 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In process industries, virtually one-third of the control loops experience oscillations, hence delivering unsatisfactory performance. Oscillations originated in one control loop can propagate to other control loops through interactions between the control loops. – An oscillating control loop cannot maintain process variables of interest at their desired values, leading to excessive consumption of energy and raw materials, curtailment in product quality, increased production cost, etc. – The oscillations are often the outcomes of poorly tuned controllers, control valve faults, sensor faults, or equipment faults. – The identification and localization of the source of oscillations are the first and foremost step in oscillation removal. This task can be performed by the manual inspection of each control loop.…”

Section: Introductionmentioning

confidence: 99%

Shape-Based Pattern Recognition Approaches toward Oscillation Detection

Memarian,

Damarla,

Huang

et al. 2024

Ind. Eng. Chem. Res.

Self Cite

View full text Add to dashboard Cite

Oscillation in control loops is a frequent problem in the process industries. These oscillations directly impact product quality, leading to a decreased plant profit. Additionally, oscillations increase energy consumption and waste raw materials and pose a significant restriction on the performance of the operational unit. Therefore, it is essential to isolate the loops that exhibit oscillations. In this work, two practical and effective methods are proposed to detect oscillations in process control loops. Both methods are aimed at detecting the presence of a triangle-like shape in the "D vs process variable (PV)" [or "D vs controller output (OP)"] plot to identify oscillations in the control loops. Here, "D" represents the Euclidean distance, which involves the deviations from the mean values of the variables OP and PV. Method 1 accomplishes this objective in an unsupervised manner by fitting a nonlinear algebraic function to the data: "D" and "PV". Method 2 uses a deep convolutional neural network for detecting the triangle-like shape. The performance of both the methods was evaluated by applying them to benchmark control loops sourced from various industries, including chemical, paper, mining, and metal industries, along with control loops in a local refinery unit. While both methods have their own advantages and application scenarios, the results demonstrated that both the proposed methods identified oscillatory control loops for the majority of the cases studied.

show abstract

Section: Introductionmentioning

confidence: 99%

Shape-Based Pattern Recognition Approaches toward Oscillation Detection

Memarian,

Damarla,

Huang

et al. 2024

Ind. Eng. Chem. Res.

Self Cite

View full text Add to dashboard Cite

show abstract

“… 11 With the continuous development of distributed control systems, the influx of large amounts of real-time data has driven the development of data-driven soft sensors for various processes. 12 The initial generation of data-driven soft sensors includes principal component analysis (PCA), 13 principal component regression (PCR), 14 partial least-squares (PLS), 15 support vector machine (SVM), 16 artificial neural network (ANN), 17 etc. Researchers have proposed new modeling strategies based on PCA and PLS to cope with different processes.…”

Section: Introductionmentioning

confidence: 99%

Supervised Attention-Based Bidirectional Long Short-Term Memory Network for Nonlinear Dynamic Soft Sensor Application

et al. 2023

View full text Add to dashboard Cite

Soft sensors are mathematical methods that describe the dependence of primary variables on secondary variables. A nonlinear characteristic commonly appears in modern industrial process data with increasing complexity and dynamics, which has brought challenges to soft sensor modeling. To solve these issues, a novel supervised attention-based bidirectional long short-term memory (SA-BiLSTM) is first proposed in this paper to handle the nonlinear industrial process modeling with dynamic features. In this SA-BiLSTM model, an attention mechanism is introduced to calculate the correlation between hidden features in each time step, thus avoiding the loss of important information. Furthermore, this approach combines historical quality information and a moving window through a supervised strategy of quality variables. Such manipulation not only extracts and exploits nonlinear dynamic latent information from the process and quality variables but also enhances the model's learning efficiency and overall prediction performance. Finally, two real industrial examples demonstrate the superiority of the proposed method compared to conventional methods.

show abstract

“…As a result, building soft sensors becomes a perfect alternative solution. There are two general classes of soft sensors, one being physics (e.g., first-principle model)-driven and the other being data-driven . The physics-driven family has predominantly been applied to the design and planning of processing plants focusing on ideal steady-state operation.…”

Section: Introductionmentioning

confidence: 99%

“…There are two general classes of soft sensors, one being physics (e.g., first-principle model)-driven and the other being data-driven. 4 The physics-driven family has predominantly been applied to the design and planning of processing plants focusing on ideal steady-state operation. The data-driven soft-sensor overcomes this drawback as they are built using the data obtained during plant operation, which gives a better representation of the true process conditions, allowing them to be described in a more meaningful manner.…”

Section: Introductionmentioning

confidence: 99%

Integrating Autoencoder and Heteroscedastic Noise Neural Networks for the Batch Process Soft-Sensor Design

Kay

Mowbray

et al. 2022

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

Viscosity represents a key product quality indicator but has been difficult to measure in-process in real-time. This is particularly true if the process involves complex mixing phenomena operated at dynamic conditions. To address this challenge, in this study, we developed an innovative soft sensor by integrating advanced artificial neural networks. The soft sensor first employs a deep learning autoencoder to extract information-rich process features by compressing high-dimensional industrial data and then adopts a heteroscedastic noise neural network to simultaneously predict product viscosity and associated uncertainty. To evaluate its performance, predictions of product viscosity were made for a number of industrial batches operated over different seasons. Furthermore, probabilistic machine learning techniques, including the Gaussian process and the Bayesian neural network, were selected to benchmark against the heteroscedastic noise neural network. Through comparison, it is found that the proposed soft-sensor has both high accuracy and high reliability, indicating its potential for process monitoring and quality control.

show abstract

Mixture robust semi-supervised probabilistic principal component regression with missing input data

Cited by 16 publications

References 29 publications

Shape-Based Pattern Recognition Approaches toward Oscillation Detection

Shape-Based Pattern Recognition Approaches toward Oscillation Detection

Supervised Attention-Based Bidirectional Long Short-Term Memory Network for Nonlinear Dynamic Soft Sensor Application

Integrating Autoencoder and Heteroscedastic Noise Neural Networks for the Batch Process Soft-Sensor Design

Contact Info

Product

Resources

About