scholarly journals Temporal-Spatial Neighborhood Enhanced Sparse Autoencoder for Nonlinear Dynamic Process Monitoring

Processes ◽  
2020 ◽  
Vol 8 (9) ◽  
pp. 1079
Author(s):  
Nanxi Li ◽  
Hongbo Shi ◽  
Bing Song ◽  
Yang Tao

Data-based process monitoring methods have received tremendous attention in recent years, and modern industrial process data often exhibit dynamic and nonlinear characteristics. Traditional autoencoders, such as stacked denoising autoencoders (SDAEs), have excellent nonlinear feature extraction capabilities, but they ignore the dynamic correlation between sample data. Feature extraction based on manifold learning using spatial or temporal neighbors has been widely used in dynamic process monitoring in recent years, but most of them use linear features and do not take into account the complex nonlinearities of industrial processes. Therefore, a fault detection scheme based on temporal-spatial neighborhood enhanced sparse autoencoder is proposed in this paper. Firstly, it selects the temporal neighborhood and spatial neighborhood of the sample at the current time within the time window with a certain length, the spatial similarity and time serial correlation are used for weighted reconstruction, and the reconstruction combines the current sample as the input of the sparse stack autoencoder (SSAE) to extract the correlation features between the current sample and the neighborhood information. Two statistics are constructed for fault detection. Considering that both types of neighborhood information contain spatial-temporal structural features, Bayesian fusion strategy is used to integrate the two parts of the detection results. Finally, the superiority of the method in this paper is illustrated by a numerical example and the Tennessee Eastman process.

Author(s):  
Rui-Cheng Zhang ◽  
Yu-Ting Li ◽  
Wei-Zheng Liang ◽  
Wei Xiong

Aiming at the problems of inaccurate fault detection and error alarm in the process of hot strip mill process, a fault detection scheme of canonical independent component analysis is proposed. The new scheme first uses canonical variable analysis to calculate the canonical variable matrix of observation data, which effectively solves the problem of autocorrelation and cross-correlation. Then the canonical variable matrix is decomposed by independent component analysis to obtain independent elements. Finally, the data are monitored online through constructing statistics. It is proved that the accuracy of the scheme for identifying fault data is reached to 100%, and the misjudgment rate data are reduced to less than 0.6% through the simulation study of the hot strip mill process data.


TAPPI Journal ◽  
2014 ◽  
Vol 13 (1) ◽  
pp. 33-41
Author(s):  
YVON THARRAULT ◽  
MOULOUD AMAZOUZ

Recovery boilers play a key role in chemical pulp mills. Early detection of defects, such as water leaks, in a recovery boiler is critical to the prevention of explosions, which can occur when water reaches the molten smelt bed of the boiler. Early detection is difficult to achieve because of the complexity and the multitude of recovery boiler operating parameters. Multiple faults can occur in multiple components of the boiler simultaneously, and an efficient and robust fault isolation method is needed. In this paper, we present a new fault detection and isolation scheme for multiple faults. The proposed approach is based on principal component analysis (PCA), a popular fault detection technique. For fault detection, the Mahalanobis distance with an exponentially weighted moving average filter to reduce the false alarm rate is used. This filter is used to adapt the sensitivity of the fault detection scheme versus false alarm rate. For fault isolation, the reconstruction-based contribution is used. To avoid a combinatorial excess of faulty scenarios related to multiple faults, an iterative approach is used. This new method was validated using real data from a pulp and paper mill in Canada. The results demonstrate that the proposed method can effectively detect sensor faults and water leakage.


Author(s):  
Iyappan Murugesan ◽  
Karpagam Sathish

: This paper presents electrical power system comprises many complex and interrelating elements that are susceptible to the disturbance or electrical fault. The faults in electrical power system transmission line (TL) are detected and classified. But, the existing techniques like artificial neural network (ANN) failed to improve the Fault Detection (FD) performance during transmission and distribution. In order to reduce the power loss rate (PLR), Daubechies Wavelet Transform based Gradient Ascent Deep Neural Learning (DWT-GADNL) Technique is introduced for FDin electrical power sub-station. DWT-GADNL Technique comprises three step, normalization, feature extraction and FD through optimization. Initially sample power TL signal is taken. After that in first step, min-max normalization process is carried out to estimate the various rated values of transmission lines. Then in second step, Daubechies Wavelet Transform (DWT) is employed for decomposition of normalized TLsignal to different components for feature extraction with higher accuracy. Finally in third step, Gradient Ascent Deep Neural Learning is an optimization process for detecting the local maximum (i.e., fault) from the extracted values with help of error function and weight value. When maximum error with low weight value is identified, the fault is detected with lesser time consumption. DWT-GADNL Technique is measured with PLR, feature extraction accuracy (FEA), and fault detection time (FDT). The simulation result shows that DWT-GADNL Technique is able to improve the performance of FEA and reduces FDT and PLR during the transmission and distribution when compared to state-of-the-art works.


2020 ◽  
Vol 53 (2) ◽  
pp. 4202-4207
Author(s):  
Anass Taoufik ◽  
Michael Defoort ◽  
Mohamed Djemai ◽  
Krishna Busawon ◽  
Juan Diego Sánchez-Torres

Sign in / Sign up

Export Citation Format

Share Document