Pseudo Time-Slice Construction Using a Variable Moving Window <i>k</i> Nearest Neighbor Rule for Sequential Uneven Phase Division and Batch Process Monitoring

Zhang, Shumei; Zhao, Chunhui; Wang, Shu; Wang, Fuli

doi:10.1021/acs.iecr.6b03743

Cited by 63 publications

(23 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Generally, the industrial processes with multi-state characteristics can be divided according to different states and then formulated using corresponding monitoring models. Furthermore, principal component analysis (PCA) has been widely applied for different industrial processes [28][29][30]. However, to the best of our knowledge, the partition of the floating state and the construction of the model have not been solved regarding an air cushion furnace.…”

Section: Introductionmentioning

confidence: 99%

Multi-Model- and Soft-Transition-Based Height Soft Sensor for an Air Cushion Furnace

Hou

Zhang

Dai

et al. 2020

Sensors

View full text Add to dashboard Cite

The floating height of the strip in an air cushion furnace is a key parameter for the quality and efficiency of production. However, the high temperature and high pressure of the working environment prevents the floating height from being directly measured. Furthermore, the strip has multiple floating states in the whole operation process. It is thus difficult to employ a single model to accurately describe the floating height in different states. This paper presents a multi-model soft sensor to estimate the height based on state identification and the soft transition. First, floating states were divided using a partition method that combined adaptive k-nearest neighbors and principal component analysis theories. Based on the identified results, a hybrid model for the stable state, involving a double-random forest model for the vibration state and a soft-transition model, was created to predict the strip floating height. In the hybrid model for the stable state, a mechanistic model combined thick jet theory and the equilibrium equation of force to cope with the lower floating height. In addition, a novel soft-transition model based on data gravitation that further reflects the intrinsic process characteristic was developed for the transition state. The effectiveness of the proposed approach was validated using a self-developed air cushion furnace experimental platform. This study has important value for the process prediction and control of air cushion furnaces.

show abstract

Section: Introductionmentioning

confidence: 99%

Multi-Model- and Soft-Transition-Based Height Soft Sensor for an Air Cushion Furnace

Hou

Zhang

Dai

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…The batch process is an important part of modern industry, and the safety monitoring of the batch process is very important and meaningful . Theoretical research based on data‐driven modeling methods, including soft sensor modeling and process monitoring, has made significant progress and has become increasingly intelligent with industrial processes .…”

Section: Introductionmentioning

confidence: 99%

“…The batch process is an important part of modern industry, and the safety monitoring of the batch process is very important and meaningful. [1][2][3] Theoretical research based on data-driven modeling methods, including soft sensor modeling and process monitoring, has made significant progress and has become increasingly intelligent with industrial processes. [4][5][6][7][8][9][10][11][12] For the batch process, multiway principal component analysis (MPCA) is one of the most basic and the most extensively used monitoring methods.…”

Section: Introductionmentioning

confidence: 99%

Just‐in‐time learning–multiple subspace support vector data description used for non‐Gaussian dynamic batch process monitoring

Yan

Jiang

et al. 2019

Journal of Chemometrics

View full text Add to dashboard Cite

Batch process data are time‐varying dynamic and non‐Gaussian distributed. In addition, for multivariate statistical process monitoring, their variability can be overwhelmed when considering local variability behavior. To address the abovementioned issues, an improved batch process monitoring approach is presented that integrates just‐in‐time learning and multiple subspace support vector data description (JITL‐MSSVDD). A new multiple subspace segmentation method is proposed that classifies a contribution array that is calculated on the mixing matrix of independent component analysis (ICA). Offline, the variable subspace segmentation rule can be obtained from the proposed method. The subspace monitoring models can reduce the risk of the variability being overwhelmed. Online, local modeling samples are collected through JITL, which can reduce the impact of the time‐varying dynamic on modeling accuracy. Then, in accordance with the variable subspace segmentation rule, which is obtained offline, MSSVDD models are constructed to solve non‐Gaussian problems. The advantage of the proposed JITL‐MSSVDD is demonstrated through the standard test model fed‐batch penicillin fermentation process.

show abstract

“…Some existing methods may be used to address this problem. One is the adaptive strategy [25][26][27] that consecutively updates the monitoring model to capture the nonstationary process behaviors which, however, may undesirably adapt to slow-varying faulty conditions. The second one is the data-differencing process 28 which, however, can cause the loss of dynamic information in the data and lead to poor modeling performance.…”

Section: Introductionmentioning

confidence: 99%

A full‐condition monitoring method for nonstationary dynamic chemical processes with cointegration and slow feature analysis

2017

Self Cite

View full text Add to dashboard Cite

Chemical processes are in general subject to time variant conditions because of load changes, product grade transitions, or other causes, resulting in typical nonstationary dynamic characteristic. It is of a considerable challenge for process monitoring to consider all possible operation conditions simultaneously including multifarious steady states and dynamic switchings. In this work, a novel full-condition monitoring strategy is proposed based on both cointegration analysis (CA) and slow feature analysis (SFA) with the following considerations: (1) Despite that the operation conditions may vary over time, they may follow certain equilibrium relations that extend beyond the current time, and (2) there may exist certain dynamic relations that stay invariant under normal process operation despite process may operate at different operating conditions. To monitor both equilibrium and dynamic relations, in the proposed method, nonstationary variables are separated from stationary variables first. Then by CA and SFA, the long-term equilibrium relation is distinguished from the specific relation held by the current conditions from both static and dynamic aspects. Various monitoring statistics are designed with clear physical interpretation. It can distinguish between the changes of operation conditions and real faults by checking deviations from equilibrium relation and deviations from the specific relation. Case study on a chemical industrial scale multiphase flow experimental rig shows the validity of the proposed full-condition monitoring method.

show abstract

Pseudo Time-Slice Construction Using a Variable Moving Window k Nearest Neighbor Rule for Sequential Uneven Phase Division and Batch Process Monitoring

Cited by 63 publications

References 33 publications

Multi-Model- and Soft-Transition-Based Height Soft Sensor for an Air Cushion Furnace

Multi-Model- and Soft-Transition-Based Height Soft Sensor for an Air Cushion Furnace

Just‐in‐time learning–multiple subspace support vector data description used for non‐Gaussian dynamic batch process monitoring

A full‐condition monitoring method for nonstationary dynamic chemical processes with cointegration and slow feature analysis

Contact Info

Product

Resources

About