Adaptive slow feature analysis - sparse autoencoder based fault detection for time-varying processes

•Slow feature analysis is used to extract the dynamic characteristics of the process and establish a "model update index" to realize the failure judgment of time-varying process models.•Sparse autoencoder is used to extract process features and establish "process monitoring index"...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Journal of the Taiwan Institute of Chemical Engineers Ročník 142; s. 104599
Hlavní autori: Tan, Shuai, Zhou, Xinjin, Shi, Hongbo, Song, Bing
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier B.V 01.01.2023
Predmet:
Adaptive process monitoring
Slow feature analysis
Sparse autoencoder
Time-varying process
Time-varying process
Sparse autoencoder
Slow feature analysis
Adaptive process monitoring
ISSN:1876-1070, 1876-1089
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•Slow feature analysis is used to extract the dynamic characteristics of the process and establish a "model update index" to realize the failure judgment of time-varying process models.•Sparse autoencoder is used to extract process features and establish "process monitoring index" to realize process monitoring.•Updated dataset is built using the normal neighbors of online samples. The incremental update strategy is proposed to establish adaptive model in order to describe the dynamic characteristics of time-varying process. Fault detection and diagnosis technology is of great significance for practical industrial processes. Industrial process characteristics change with time due to various reasons such as changing working conditions. This will cause false alarm or missing alarm of process monitoring. In this paper, an adaptive slow feature analysis (SFA) - sparse autoencoder (SAE) algorithm is proposed to establish an adaptive model for time-varying process monitoring. Model update index is built based on time-varying characteristics extracted using SFA model. Process monitoring index is built based on sparse characteristics extracted using SAE model. Through online adaptive update strategy, updated monitoring model is realized to adapt to the time-varying characteristics of the process. The proposed algorithm has good performance on penicillin fermentation process data set and can realize the task of adaptive process monitoring. [Display omitted]
ISSN:1876-1070
1876-1089
DOI:10.1016/j.jtice.2022.104599