Pregled bibliografske jedinice broj: 528873
Fault tolerant system in a process measurement system based on the PCA method
Fault tolerant system in a process measurement system based on the PCA method // MIPRO 2011, Student Papers
Opatija, Hrvatska, 2011. (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 528873 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Fault tolerant system in a process measurement system based on the PCA method
Autori
Rosković, Andrijana ; Grbić, Ratko ; Slišković, Dražen
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
MIPRO 2011, Student Papers
/ - , 2011
Skup
MIPRO 2011 34. International Convention on Information and Communication Technology, Electronics and Microelectronics
Mjesto i datum
Opatija, Hrvatska, 23.05.2011. - 27.05.2011
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
process monitoring; fault detection; fault identification; PCA; contribution plot
Sažetak
Better process control is an important step towards increasing the efficiency of production facility. More complex control systems are introduced which require a more complex process measuring systems. Efficient process control is based upon quality and reliable process variable measurement. Process equipment failure can significantly deteriorate the product quality and even cause production outage, resulting in high additional costs. This paper analyzes automatic fault detection and identification of process measurement equipment or sensors. Different statistical methods can be used for this purpose. PCA based statistical process monitoring algorithms are applied on selected examples. For the purpose of fault detection and identification, the PCA method is used to model the correlation among process variables in the input space. Hotelling's (T2) and Q (SPE) statistics are used for fault detection because they provide an indication of unusual variability within and outside normal workspace. Contribution plots are used for fault identification. This paper also presents the estimation (reconstruction) of the value of faulty sensor process variable, which allows the continuation of the process, although the fault might have occurred. Results of all considered examples are compared and discussed.
Izvorni jezik
Engleski
Znanstvena područja
Elektrotehnika, Računarstvo
POVEZANOST RADA
Projekti:
165-0361621-2000 - Distribuirano računalno upravljanje u transportu i industrijskim pogonima (Hocenski, Željko, MZO ) ( CroRIS)
Ustanove:
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek