Multi-moving-window neural network for modeling of purified terephthalic acid solvent system

Author(s):  
Yuan Xu ◽  
Qunxiong Zhu
2011 ◽  
Vol 467-469 ◽  
pp. 389-394
Author(s):  
Yuan Xu ◽  
Qun Xiong Zhu ◽  
Yan Qing Wang

To enhance the work efficiency and systematization degree of fault diagnosis for process production, an extension CBR fault diagnosis (ECFD) method is proposed in which the extension theory applies the formal model to expand the possibility of matter and explore the innovative principle, and the CBR technology provides the ways for solution based on the historical experiences. Considering the characters of the two technologies, some researches are made on the extension basic-element case-description, distance-based extension case-retrieval, and character difference-based extension case-revision for fault diagnosis. Taking the purified terephthalic acid (PTA) solvent system as an application example for fault diagnosis, the ECFD method is proved to be feasible and effective, which provides a new way for diagnosing the production faults accurately.


2020 ◽  
Vol 77 (6) ◽  
pp. 1081-1088
Author(s):  
Tayyibe Alpay ◽  
Burcin Karabey ◽  
Nuri Azbar ◽  
Guven Ozdemir

Sign in / Sign up

Export Citation Format

Share Document