2023
DOI: 10.1016/j.ces.2022.118338
|View full text |Cite
|
Sign up to set email alerts
|

Causal network inference and functional decomposition for decentralized statistical process monitoring: Detection and diagnosis

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2

Citation Types

0
3
0

Year Published

2023
2023
2025
2025

Publication Types

Select...
6

Relationship

0
6

Authors

Journals

citations
Cited by 7 publications
(3 citation statements)
references
References 82 publications
0
3
0
Order By: Relevance
“…In the causal inference stage, a causal network among the process variables with the direction of information propagation is determined by using the PC-NET model and the PCGC method. Subsequently, in the stage of fault detection, the occurrence of faults is detected by using a causality-based multivariate sensitivity enhancement transformation (MSET) method proposed by Reis et al 20 While in the fault isolation and identification stage, according to the inputs with a causal network, the causality-attributing RBC method is employed to locate the root cause of the fault occurrence. Finally, in the stage of fault classification, the isolated faults are classified into different grades and the corresponding measures are suggested.…”
Section: Introductionmentioning
confidence: 99%
“…In the causal inference stage, a causal network among the process variables with the direction of information propagation is determined by using the PC-NET model and the PCGC method. Subsequently, in the stage of fault detection, the occurrence of faults is detected by using a causality-based multivariate sensitivity enhancement transformation (MSET) method proposed by Reis et al 20 While in the fault isolation and identification stage, according to the inputs with a causal network, the causality-attributing RBC method is employed to locate the root cause of the fault occurrence. Finally, in the stage of fault classification, the isolated faults are classified into different grades and the corresponding measures are suggested.…”
Section: Introductionmentioning
confidence: 99%
“…For example, the incorporation with the idea of locality preserving or neighborhood embedding has motivated different analytical algorithms for process monitoring 14,16 . Moreover, the advances achieved in machine learning continuously provide creative solutions for data‐driven process monitoring 17–19 …”
Section: Introductionmentioning
confidence: 99%
“…14,16 Moreover, the advances achieved in machine learning continuously provide creative solutions for data-driven process monitoring. [17][18][19] Given that the operating stability of an industrial process could always generate time-serial correlated samples in a successive manner, the time-serial relationship within consecutive samples is required to be given full consideration. Ku et al 20 proposed to stack successive samples together so that the time-serial correlation could be further modeled.…”
mentioning
confidence: 99%