Improving the interpretability of causality maps for fault identification

Date
2020-12
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: Worldwide competition forces modern mineral processing plants to operate at high productivity. This high productivity is achieved by implementing process monitoring to maintain the desired operating conditions. However, a fault originating in one section of a plant can propagate throughout the plant and so obscure its root cause. Causality analysis is a method that identifies the cause-effect relationships between process variables and presents these in a causality map which can be used to track the propagation path of a fault back to its root cause. A major obstacle to the wide acceptance of causality analysis as a tool for fault diagnosis in industry is the poor interpretability of causality maps. This study identified, proposed and assessed ways to improve the interpretability of causality maps for fault identification. All approaches were tested on a simulated case study and the resulting maps compared to a standard causality map or its transitive reduction. The ideal causality map was defined and all comparisons were performed based on its characteristics. Causality maps were produced using conditional Granger causality (GC), with a novel heuristic approach for selecting sampling period and time window. Conditional GC was found to be ill-suited to plant-wide causality analysis, due to large data requirements, poor model order selection using AIC, and inaccuracy in the presence of multiple different residence times and time delays. Methods to incorporate process knowledge to constrain connections and potential root causes were investigated and found to remove all spurious connections and decrease the pool of potential root cause variables respectively. Tools such as visually displaying node rankings on the causality map and incorporating sliders to manipulate connections and variables were also investigated. Furthermore, a novel hierarchical approach for plant-wide causality analysis was proposed, where causality maps were constructed in two subsequent stages. In the first stage, a less-detailed plant-wide map was constructed using representatives for groups of variables, and used to localise the fault to one of those groups of variables. Variables were grouped according to plant sections or modules identified in the data, and the first principal component (PC1) was used to represent each group (PS-PC1 and Mod-PC1 respectively). PS-PC1 was found to be the most promising approach, as its plant-wide map clearly identified the true root cause location, and the stage-wise application of conditional GC significantly reduced the required number of samples from 13 562 to 602. Lastly, a usability study in the form of a survey was performed to investigate the potential for industrial application of the tools and approaches presented in this study. Twenty responses were obtained, with participants consisting of Stellenbosch University final-year/postgraduate students, employees of an industrial IoT firm, and Anglo American Platinum employees. Main findings include that process knowledge is vital; grouping variables improves interpretability by decreasing the number of nodes; accuracy must be maintained during causality map simplification; and sliders add confusion by causing significant changes in the causality map. In addition, survey results found PS-PC1 to be the most user-friendly approach, further emphasizing its potential for application in industry.
AFRIKAANSE OPSOMMING: Wêreldwye kompetisie forseer moderne mineraalprosesseringaanlegte om by hoë produktiwiteit bedryf te word. Hierdie hoë produktiwiteit word bereik deur prosesmonitering te implementeer om die gewenste bedryfskondisies te handhaaf. ’n Fout wat in een deel van ’n aanleg ontstaan kan egter regdeur die aanleg voortplant en so die grondoorsaak verberg. Oorsaaklikheidanalise is ’n metode wat die oorsaak-en-gevolg-verhouding tussen prosesveranderlikes identifiseer en hierdie in ’n oorsaaklikheidskaart toon wat gebruik kan word om die voortplantings roete van ’n fout terug na sy grondoorsaak te volg. ’n Groot hindernis vir die wye aanvaarding van oorsaaklikheidanalise as instrument vir foutdiagnose in industrie, is die swak interpreteerbaarheid van oorsaaklikheidskaarte. Hierdie studie het maniere om die interpreteerbaarheid van oorsaaklikheidskaarte vir foutidentifikasie te verbeter, geïdentifiseer, voorgestel en geassesseer. Alle benaderings is getoets op ’n gesimuleerde gevallestudie en die resulterende kaarte is vergelyk met ’n standaard oorsaaklikheidskaart of sy transitiewe inkrimping. Die ideale oorsaaklikheidskaart is gedefinieer en alle vergelykings is uitgevoer gebaseer op sy karakteristieke. Oorsaaklikheidskaarte is geproduseer deur kondisionele Granger-oorsaaklikheid (GC) te gebruik, met ’n nuwe heuristiese benadering om steekproefperiode en tydgleuf te selekteer. Kondisionele GC is gevind om nie gepas te wees vir aanlegwye oorsaaklikheidanalise nie, as gevolg van groot datavereistes, swak seleksie van modelorde as AIC gebruik word, en onakkuraatheid in die teenwoordigheid van veelvoudige, verskillende verblyftye en tydvertraging. Metodes om proseskennis te inkorporeer om konneksies en potensiële grondoorsake te bedwing, is ondersoek en gevind om alle konneksies wat vals is te verwyder en die groep van potensiële grondoorsaakveranderlikes te verminder, onderskeidelik. Instrumente soos om node-ordes op die oorsaaklikheidskaart visueel te vertoon en skuiwers te inkorporeer om konneksies en veranderlikes te manipuleer is ook ondersoek. Verder is ’n nuwe hiërargiese benadering vir aanlegwye oorsaaklikheidanalise voorgestel, waar oorsaaklikheidskaarte in twee opeenvolgende fases gebou is. In die eerste fase is ’n minder gedetaileerde aanlegwye kaart gebou deur verteenwoordigers vir groepe veranderlikes te gebruik, en is gebruik om die fout na een van daardie groepe van veranderlikes te lokaliseer. Veranderlikes is gegroepeer volgens aanlegdele of modules geïdentifiseer in die data, en die eerste hoof komponent (PC1) is gebruik om elke groep te verteenwoordig (PS-PC1 en Mod-PC1 onderskeidelik). PS-PC1 is gevind om die mees belowende benadering te wees, want sy aanlegwye kaart het duidelik die ware grondoorsaakligging geïdentifiseer, en die stap-gewyse toepassing van kondisionele GC het die vereisde aantal steekproewe beduidend verminder van 13 562 tot 602. Laastens, ’n bruikbaarheidstudie in die vorm van ’n opname is uitgevoer om die potensiaal vir industriële toepassing van die instrumente en benaderinge voorgestel in hierdie studie, te ondersoek. Twintig antwoorde is verkry, met deelnemers wat bestaan het uit Universiteit van Stellenbosch se finale jaar/nagraadse studente, werknemers van ’n industriële IoT-firma, en Anglo American Platinum werknemers. Hoofbevindinge het ingehou dat proseskennis noodsaaklik is; om veranderlikes te groepeer verbeter interpreteerbaarheid deur die aantal nodes te verminder; akkuraatheid moet gehandhaaf word gedurende vereenvoudiging van oorsaaklikheidskaarte; en skuiwers dra by tot verwarring deur beduidende veranderinge in die oorsaaklikheidskaart te maak. Daarmee saam het die opname se resultate gevind dat PS-PC1 die meer gebruiksvriendelike benadering was, wat sy potensiaal vir toepassing verder beklemtoon.
Description
Thesis (MEng)--Stellenbosch University, 2020.
Keywords
Process monitoring, UCTD, Fault location (Engineering), Mineral processing
Citation