Adaptive process monitoring using principal component analysis and Gaussian Mixture Models
dc.contributor.advisor | Auret, Lidia | en_ZA |
dc.contributor.advisor | Kroon, R. S. (Steve) | en_ZA |
dc.contributor.author | Addo, Prince | en_ZA |
dc.contributor.other | Stellenbosch University. Faculty of Engineering. Dept. of Process Engineering. | en_ZA |
dc.date.accessioned | 2019-02-18T13:45:55Z | |
dc.date.accessioned | 2019-04-17T08:33:44Z | |
dc.date.available | 2019-02-18T13:45:55Z | |
dc.date.available | 2019-04-17T08:33:44Z | |
dc.date.issued | 2019-04 | |
dc.description | Thesis (MEng)--Stellenbosch University, 2019. | en_ZA |
dc.description.abstract | ENGLISH ABSTRACT: Principal component analysis (PCA) is a well-known technique used in combination with monitoring statistics for fault detection. Moving window PCA and recursive PCA are adaptive extensions of PCA that operate by periodically updating the monitoring model to incorporate new observations. This allows the monitoring model to cope with process behaviours that change slowly over time such as equipment aging, catalyst deactivation, and reaction kinetics drift and thereby improving monitoring performance. Recent demands and advancements in process industries, however, may result in multimodal operations, where distinct clusters are present in measurement data. The performance of the aforementioned PCA-based monitoring techniques is hindered due to the violation of the implicit assumption that all the observed process data belong to the same Gaussian distribution. To improve monitoring performance, multimodal techniques are required. The Gaussian mixture model (GMM) is a probabilistic model that can account for the observed modes in the process data and therefore be used in the monitoring of multimode processes. However, multimodal processes also exhibit behaviours that change slowly over time, which is challenging. This work develops a monitoring approach that extends adaptive PCA techniques to GMM, which effectively addresses the aforementioned challenge. This is done by continuously refreshing the model parameters and monitoring statistics for the PCA and GMM. Other key areas that the work focuses on are in improving the specifications for adaptive PCA protocol (taking into consideration the various model update methods) and Gaussian mixture model methods (taking into consideration the monitoring model types and data types). Also, the performance of unimodal and multimodal process monitoring approaches was assessed. The performance of the developed approach and the improved implementations of the pre-existing methods were assessed using various case studies including unimodal and multimodal processes both with and without drift as well as various fault types. The Tennessee Eastman process and the non-isothermal continuously stirred tank reactor process are the two main simulators considered. Results for the considered cases show improved performance for the developed approach (adaptive PCA-based GMM) as compared to PCA, adaptive PCA, and traditional GMM, in fault detection. The GMM, as expected, performed better for multimodal cases than the PCA approaches. Also, the adaptive PCA approach performed better than PCA when there is process drift. | en_ZA |
dc.description.abstract | AFRIKAANSE OPSOMMING: Hoofkomponentanalise (PCA) is ʼn welbekende tegniek wat gebruik word saam met moniteringstatistiek vir foutdeteksie. Bewegende venster PCA en rekursiewe PCA is aanpassende uitbreidings van PCA wat bedryf word deur die moniteringsmodel periodies op te dateer om nuwe waarnemings te inkorporeer. Dit laat die moniteringsmodel toe om by te hou met die proses gedrag wat geleidelik verander, soos toerusting wat verouder, katalis deaktivering, en reaksie kinetika dwaal, en sodoende die monitering werkverrigting verbeter. Onlangse eise en vooruitgang in proses industrieë mag egter multimodale bedrywe tot gevolg hê, waar duidelike groepe in metingsdata teenwoordig is. Die werkverrigting van voorafgenoemde PCA-gebaseerde moniteringstegnieke word belemmer as gevolg van die oortreding van die implisiete aanname dat al die prosesdata waargeneem, aan dieselfde Gauss-verdeling behoort. Om monitering werkverrigting te verbeter, word multimodale tegnieke benodig. Die Gauss-mengselmodel (GMM) is ʼn waarkynlikheidsmodel wat rekening kan hou met die waargenome modusse in die prosesdata en kan daarom gebruik word in die monitering van multimodale prosesse. Multimodale prosesse vertoon egter gedrag wat stadig met tyd verander, wat uitdagend is. Hierdie werk ontwikkel ʼn monitering benadering wat aanpassende PCA-tegnieke na GMM uitbrei, wat die voorafgenoemde uitdaging doeltreffend aanspreek. Dit word gedoen deur deurlopend die model parameters en moniteringstatistiek vir die PCA en GMM te verfris. Ander sleutelareas waarop die werk fokus is om die spesifikasies vir aanpassende PCA protokol te verbeter (die verskillende model opdatering metodes word in ag geneem) en Gauss-mengselmodelmetodes (die monitering model tipes en data tipes word in ag geneem). Die werkverrigting van unimodale en mulitmodale proses monitering benaderings is ook geassesseer. Die werksverrigting van die ontwikkelde benadering en die verbeterde implementasies van die voorafbestaande metodes is geassesseer deur verskillende gevallestudies te gebruik, insluitend unimodale en multimodale prosesse beide met en sonder dwaal sowel as verskillende fout tipes. Die Tennessee Eastman-proses en die nie-isotermiese kontinu geroerde tenk reaktor proses is die twee hoof simulators wat oorweeg is. | af_ZA |
dc.format.extent | 226 pages | en_ZA |
dc.identifier.uri | http://hdl.handle.net/10019.1/106193 | |
dc.language.iso | en_ZA | en_ZA |
dc.publisher | Stellenbosch : Stellenbosch University | en_ZA |
dc.rights.holder | Stellenbosch University | en_ZA |
dc.subject | Gaussian processes | en_ZA |
dc.subject | UCTD | |
dc.subject | Process engineering -- Monitoring | en_ZA |
dc.subject | Anomaly detection (Computer security) | en_ZA |
dc.subject | Distribution (Probability theory) | en_ZA |
dc.title | Adaptive process monitoring using principal component analysis and Gaussian Mixture Models | en_ZA |
dc.type | Thesis | en_ZA |