A new adaptive PCA based thresholding scheme for fault detection in complex systems

Bakdi, Azzeddine; Kouadri, Abdelmalek

doi:10.1016/j.chemolab.2017.01.013

Cited by 64 publications

(50 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Each sample has 2 variables, which are independent and identically distributed in each mode. In mode 1, 2 variables, x 1 and x 2 , both follow the normal distribution with [0, 0.1], while x 1 and x 2 in mode 2 both follow the normal distribution with Bakdi and Kouadri and Perk and Çinar . The scatter plot of all of the training samples is shown in Figure .…”

Section: Multimodal Fault Detection Methods Of Knn Based On Probabilitmentioning

confidence: 99%

“…In mode 1, 2 variables, x 1 and x 2 , both follow the normal distribution with [0, 0.1], while x 1 and x 2 in mode 2 both follow the normal distribution with Bakdi and Kouadri and Perk and Çinar. 6,8 The scatter plot of all of the training samples is shown in Figure 1. The blue star indicates the training samples in mode 1, and the black circle indicates the training samples in mode 2.…”

Section: Knn Rules Based On Probability Densitymentioning

confidence: 99%

See 1 more Smart Citation

kNN based on probability density for fault detection in multimodal processes

Guo

Wang

2018

Journal of Chemometrics

View full text Add to dashboard Cite

Recently, k‐nearest neighbor rules (kNN) have drawn increasing attention for fault detection of multimodal industrial processes. However, the traditional kNN method performs poorly for weak faults in a dense mode when the dispersion degree of each mode is quite different. The reason is that the kNN statistics of weak faults are usually submerged by those of normal data in a mode with a high dispersion degree. To improve the fault detection performance of kNN in this case, this paper proposes a new multimodal fault detection method of kNN based on probability density. The proposed method does not need to consider the different degrees of dispersion between modes and avoids the problem of weak faults in a mode that has a low dispersion degree that is submerged by normal data in a mode with a high dispersion degree. A multimodal numerical example with different dispersion degrees of each mode and an industrial application in a semiconductor manufacturing process are used to verify the effectiveness of the proposed method. The simulation results demonstrate that the proposed method shows better fault detection performance than the kNN, local outlier factor, and weighted difference principal component analysis methods.

show abstract

Section: Multimodal Fault Detection Methods Of Knn Based On Probabilitmentioning

confidence: 99%

Section: Knn Rules Based On Probability Densitymentioning

confidence: 99%

kNN based on probability density for fault detection in multimodal processes

Guo

Wang

2018

Journal of Chemometrics

View full text Add to dashboard Cite

show abstract

“…The TE process contains 53 variables, including 22 continuous process variables, 12 manipulated variables (12), and 19 composition variables. [14,28] Only 33 variables are adopted in this paper for process monitoring. The 19 composition variables are not used in this paper because they are difficult to obtain in real time.…”

Section: Process Descriptionmentioning

confidence: 99%

“…[13] Considering that PCA has a fixed threshold value for fault detection, adaptive threshold PCA is proposed. [14] Dynamic principal component analysis (DPCA) is proposed to take into account the autocorrelation as well as correlation of variables. [15,16] This method takes dynamic characteristics into consideration by introducing historical data.…”

Section: Introductionmentioning

confidence: 99%

Two‐step support vector data description for dynamic, non‐linear, and non‐Gaussian processes monitoring

Zhang

2020

Can J Chem Eng

View full text Add to dashboard Cite

In this study, a the two‐step support vector data description (TS‐SVDD) method is proposed to handle the problem of fault detection for dynamic, non‐linear, and non‐Gaussian processes. First, the dynamic structure of the data is identified and the data is divided into two components: innovation component and dynamic component. Then, the innovation component is used to make the SVDD model for fault detection. Moreover, in order to overcome the issue with two‐step principal component analysis (TS‐PCA) that the choice of parameters q and D affects the fault detection effect of the methods, a genetic algorithm (GA) is used to optimize the parameters. The proposed method combines the advantages of TS‐PCA in processing dynamic process data and SVDD in dealing with non‐linear and non‐Gaussian process data. In order to evaluate the effectiveness and superiority of the proposed method, TS‐SVDD is applied to the Tennessee Eastman (TE) process and the intelligent industrial processes control test facility (I2PC‐TF), and the fault detection performance is compared with TS‐PCA and SVDD in terms of dault detection rate (FDR) and false alarm rate (FAR). The results show that TS‐SVDD has a better monitoring performance.

show abstract

“…9 Currently, the selection method of user behavior characteristics is mainly adopted for the detection of click fraud users, and characteristic dimensionality reduction methods that are frequently used are as follows: primary component analysis, multi-dimensional zoom, linear discriminant analysis, etc. [10][11][12][13][14] However, there are still some problems such as huge analysis and detection data, difficulty in determining the identity of click users, and low analysis efficiency. Zoran et al 15 investigated e-commerce users' behavior in order to detect the identity theft.…”

Section: Introductionmentioning

confidence: 99%

User click fraud detection method based on Top-Rank-k frequent pattern mining

Pan

Wang

2019

Int. J. Mod. Phys. B

View full text Add to dashboard Cite

A user click fraud detection method based on Top-Rank-k frequent pattern mining algorithm is presented to solve the click fraud problem appearing in current online advertising. Firstly, this method combines the click frequency of event samples, calculates the real evaluation score of click stream, and the click stream density function and evaluation score expression under multi-dimensional variables, and further obtains the time complexity of the next user’s click fraud process. Secondly, according to the Top-Rank-k frequent pattern, the process of click fraud detection algorithm is designed, and the click fraud user is analyzed and obtained. The results show that this method has good efficiency and correctness, and is superior to other similar algorithms.

show abstract

A new adaptive PCA based thresholding scheme for fault detection in complex systems

Cited by 64 publications

References 41 publications

kNN based on probability density for fault detection in multimodal processes

kNN based on probability density for fault detection in multimodal processes

Two‐step support vector data description for dynamic, non‐linear, and non‐Gaussian processes monitoring

User click fraud detection method based on Top-Rank-k frequent pattern mining

Contact Info

Product

Resources

About