Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning
Thesis (MEng)--Stellenbosch University, 2024.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | en_ZA en_ZA |
| Published: |
Stellenbosch : Stellenbosch University
2024
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613992748318720 |
|---|---|
| access_status_str | Open Access |
| author | Olifant, Fortunate Kenneth |
| author2 | Schutte, Cornelius Stephanus Lodewyk |
| author_browse | Olifant, Fortunate Kenneth Schutte, Cornelius Stephanus Lodewyk |
| author_facet | Schutte, Cornelius Stephanus Lodewyk Olifant, Fortunate Kenneth |
| author_sort | Olifant, Fortunate Kenneth |
| collection | Thesis |
| description | Thesis (MEng)--Stellenbosch University, 2024. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/130654 |
| institution | Stellenbosch University (South Africa) |
| language | en_ZA en_ZA |
| last_indexed | 2026-06-10T12:44:57.544Z |
| license_str | Not specified — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2024 |
| publishDateRange | 2024 |
| publishDateSort | 2024 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/130654 Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning Olifant, Fortunate Kenneth Schutte, Cornelius Stephanus Lodewyk Du Plessis, Johan Nicolaas Stellenbosch University. Faculty of Engineering. Dept. of Industrial Engineering. Deep learning (Machine learning) Water reticulation Deep-level mines UCTD Thesis (MEng)--Stellenbosch University, 2024. ENGLISH ABSTRACT: The aim of this study was to investigate whether emergency vehicle (EV) average response times can be improved by attempting to proactively avoid pre-emption conflicts through integrating emergency vehicle signal pre-emption (EVSP) and route selection. A literature review was conducted to gain valuable insight into, and knowledge of, a variety of topics related to the research statement. This included determining the causes of congestion, how it impacts the response times of EVs, as well as how EVSP and route selection can be used to reduce the delaying effect that congestion has on EV response times. Existing EVSP and route selection strategies were explored. It was found that the reviewed strategies are not typically able to consider and address cases in which multiple EVs need to traverse the same signalled intersection from conflicting directions, causing pre-emption conflicts in which one or more EVs are delayed due to the pre-emption of another. The strategies that do take these conflicts into consideration, were found to do so reactively. The literature review also explored simulation modelling, specifically traffic simulation modelling. Traffic simulation software packages were reviewed and PTV VISSIM was selected for use in this study, along with MATLAB. An EVSP and route selection strategy capable of proactively avoiding pre-emption conflicts was proposed and designed. A reference strategy was also designed, which was based on the proposed strategy, only it lacked its novel functionality. The two strategies were compared in a series of carefully constructed experiment runs using simulation modelling. To test the robustness of the proposed strategy, each experiment run was con- figured according to a unique combination of variables, such as the size of the network used, the number of EVs that required service, and the level of traffic congestion present in the network. It was empirically observed that, while not robust against the previously mentioned configurations, the proposed strategy was able to either perform equally well or better than the baseline strategy (when considering average EV response times) in the majority of cases. This indicated that an EVSP and route selection strategy that is capable of proactively anticipating and attempting to avoid pre-emption conflicts is a promising approach to reducing the response times of EVs. AFRIKAANSE OPSOMMING: Die doel van hierdie studie was om te ondersoek of die gemiddelde responstyd van noodvoertuie verbeter kan word deur proaktief seinvoorrangkonflikte (traffic signal pre- emption conflicts) te voorkom deur noodvoertuig-seinvoorrang (emergency vehicle signal pre-emption) en roeteseleksie te integreer. ’n Literatuurstudie is uitgevoer om waardevolle insig in en kennis oor ’n verskeiden- heid onderwerpe, wat by die navorsingsverklaring aansluit, te verkry. Daar is onder- soek ingestel na wat die oorsake van verkeersopeenhopings is, hoe verkeersopeenho- pings die responstyd van noodvoertuie be¨ınvloed, asook hoe noodvoertuig-seinvoorrang en roeteseleksie gebruik kan word om die vertragende effek, wat verkeersopeenhopings op noodvoertuig responstye het, te verminder. Bestaande noodvoertuig-seinvoorrang en roeteseleksie-strategie¨e is bestudeer. Daar is bevind dat die strategie¨e tipies nie in staat is om gevalle, waarin twee of meer noodvoertuie dieselfde interseksie vanuit verskeie rigtings moet deurkruis, doeltreffend te kan hanteer nie. Dit veroorsaak seinvoorrangkonflikte waarin een of meer noodvoertuie vertraag word deur die seinvoorrang wat aan ’n ander toegewys word. Die strategie¨e wat wel sulke konflikte in ag neem, doen so op ’n reaktiewe wyse. Die literatuurstudie het ook simulasiemodellering bestudeer, met ’n spesifieke fokus op verkeersimulasie-modellering. Sagtewarepakette van verkeersimulasie-modellering is ondersoek en PTV VISSIM, tesame met MATLAB, is geselekteer vir gebruik in hierdie studie. ’n Noodvoertuig-seinvoorrang en roeteseleksie-strategie, wat proaktief seinvoorrangkon- flikte kan voorkom, is voorgestel en ontwerp. ’n Verwysingstrategie gebaseer op die voorgestelde strategie is ook ontwerp, sonder dat dit die nuwe funksionaliteit van die voorgestelde strategie besit. Die twee strategie¨e is met mekaar vergelyk in ’n reeks van sorgvuldig opgestelde eksperimentele lopies deur middel van simulasiemodellering. Om die robuustheid van die voorgestelde strategie te toets, is elke eksperimentele lopie vol- gens ’n unieke kombinasie van veranderlikes, soos die grootte van die padnetwerke wat gebruik is, die hoeveelheid noodvoertuie wat in ag geneem moes word, asook die vlak van verkeer wat in ’n netwerk teenwoordig is, gekonfigureer. Dit is empiries waargeneem dat, terwyl die voorgestelde strategie nie robuust in verhouding tot hierdie veranderlikes is nie, die strategie in staat was om ´of dieselfde ´of beter as die verwysingstrategie te vaar in die meerderheid van toetsgevalle. Dit dui daarop dat ’n noodvoertuig-seinvoorrang en roeteseleksie-strategie, wat seinvoorrang konflikte proaktief kan antisipeer en probeer vermy, ’n belowende benadering is om die responstyd van noodvoertuie te probeer verkort. Masters 2024-02-13T08:27:47Z 2024-04-27T01:28:49Z 2024-02-13T08:27:47Z 2024-04-27T01:28:49Z 2024-02 Thesis https://scholar.sun.ac.za/handle/10019.1/130654 en_ZA en_ZA xiii, 185 pages : illustrations. application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Deep learning (Machine learning) Water reticulation Deep-level mines UCTD Olifant, Fortunate Kenneth Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning |
| title | Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning |
| title_full | Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning |
| title_fullStr | Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning |
| title_full_unstemmed | Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning |
| title_short | Improving energy cost savings in a complex dewatering system of a deep-level mine using machine learning |
| title_sort | improving energy cost savings in a complex dewatering system of a deep level mine using machine learning |
| topic | Deep learning (Machine learning) Water reticulation Deep-level mines UCTD |
| url | https://scholar.sun.ac.za/handle/10019.1/130654 |
| work_keys_str_mv | AT olifantfortunatekenneth improvingenergycostsavingsinacomplexdewateringsystemofadeeplevelmineusingmachinelearning |