Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles
Thesis (PhD)--Stellenbosch University, 2025.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English |
| Published: |
Stellenbosch : Stellenbosch University
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613970477613056 |
|---|---|
| access_status_str | Open Access |
| author | Lacock, Stephan |
| author2 | Booysen, M. J. (Thinus) |
| author_browse | Booysen, M. J. (Thinus) Lacock, Stephan |
| author_facet | Booysen, M. J. (Thinus) Lacock, Stephan |
| author_sort | Lacock, Stephan |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description | Thesis (PhD)--Stellenbosch University, 2025. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/134667 |
| institution | Stellenbosch University (South Africa) |
| language | English |
| last_indexed | 2026-06-10T12:44:36.436Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/134667 Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles Lacock, Stephan Booysen, M. J. (Thinus) Du Plessis, A. A. (Armand) Stellenbosch University. Faculty of Engineering. Dept. of Electrical & Electronic Engineering. Internal combustion engines -- Africa, Sub-Saharan Paratransit services -- Africa, Sub-Saharan Transportation -- Planning -- Africa, Sub-Saharan Electric vehicles -- Economic aspects -- Africa, Sub-Saharan Sustainable transportation -- Africa, Sub-Saharan Thesis (PhD)--Stellenbosch University, 2025. Lacock, S. 2025. Planning for Electric Paratransit: The Feasibility of Electrifying Internal Combustion Engine Vehicles. Unpublished doctoral dissertation. Stellenbosch: Stellenbosch University [online]. Available: https://scholar.sun.ac.za/items/98718595-25cc-4a20-9ee1-95b873fb0ab1 ENGLISH ABSTRACT: This research investigates the technical feasibility of electrifying paratransit vehicles, specifically minibus taxis, within the context of sub-Saharan Africa, where existing electric models are scarce. The study addresses this by employing a systematic and multi-faceted approach. Initially, a dynamic, force-based simulation model was developed, leveraging high-fidelity GPS tracking data from conventional internal combustion engine minibus taxis. This model facilitated the design and selection of a suitable electric powertrain, incorporating electric motor efficiency maps to dynamically assess powertrain performance relative to torque demands. Theoretical evaluations compared the energy consumption of configurations with and without a manual transmission, revealing a more efficient outcome with the inclusion of a manual transmission. The design requirements for the retrofitted vehicle’s propulsion unit were established based on driving-cycle data. Subsequently, a Toyota Hiace Ses’fikile minibus taxi was successfully retrofitted into a fully electric prototype, marking the first registered electric paratransit vehicle in the sub-Saharan region and enabling empirical validation. The average measured efficiency of the retrofitted taxi was recorded as 0.342 kWh/km. The study focused on validating and refining an energy-based simulation model known as EV Fleet Simulation using empirical data. Initially, the uncorrected model produced a vehicle efficiency of 0.53kWh/km. After adjusting the vehicle parameters based on data from the retrofitted taxi, the simulation revealed an improved efficiency of 0.372 kWh/km, with a mean absolute error (MAE) of 18.1% and a standard deviation of 18.3% for trips longer than 5 km. This was significantly enhanced through successive corrections. Ultimately, the final model achieved an MAE of 8.61% (standard deviation of 10.1%), closely aligning with the measured energy usage values: 0.331 kWh/km simulated compared to 0.327 kWh/km measured for trips exceeding 5 km. The research investigated the potential of machine learning by training models using GPS-derived driving-cycle data. A baseline Feed-Forward Neural Network (FFNN) achieved a per-sample Root Mean Square Error (RMSE) of 8.52 kW, outperforming both linear and polynomial regression models. When compared to the EV Fleet Simulation, the FFNN demonstrated a lower MAE in power prediction, with results of 5.62 kW compared to 7.59 kW. Additionally, the per-trip energy prediction MAE was 0.018 kWh/km, which is less than half that of the EV Fleet Simulation model, which had an MAE of 0.044 kWh/km. These findings emphasise the advantages of using machine learning for real-time energy predictions and feasibility assessments. The comparison between measured and simulated results highlights the effectiveness of the approach. Furthermore, machine learning models are more user-friendly, as they only require GPS coordinates of the driving cycle, while energy-based models necessitate various parameters related to the driving cycles and vehicles. The findings of this research conclusively demonstrate the technical feasibility of electrifying paratransit vehicles in sub-Saharan Africa. The empirical validation of simulation tools and the promising application of machine learning provide a robust foundation for future feasibility assessments and contribute valuable insights for optimising the design and deployment of electric paratransit vehicles in the region. AFRIKAANSE OPSOMMING: Hierdie navorsing ondersoek die tegniese uitvoerbaarheid van die elektrifisering van voertuie in die informele openbare vervoerstelsel, spesifiek minibustaxi’s, binne die konteks van Afrika suid van die Sahara, waar daar ’n tekort aan bestaande elektriese modelle is. Die studie spreek hierdie uitdaging aan deur ’n sistematiese en veelvlakkige benadering te volg. Eerstens is ’n dinamiese, krag-gebaseerde simulasie-model ontwikkel wat hoë-resolusie GPS-volgdata van bestaande binnebrandenjin-minibustaxi’s benut. Hierdie model het die ontwerp en keuse van ’n geskikte elektriese aandryfstelsel gefasiliteer, insluitend die gebruik van elektriese motordoeltreffendheidskaarte om die werkverrigting van die aandryfstelsel dinamies te evalueer in verhouding tot wringkragvereistes. Teoretiese evaluerings het die energieverbruik van konfigurasies met en sonder ’n handratkas vergelyk, Dit het getoon dat die insluiting van ’n handratkas tot ’n meer doeltreffende stelsel lei. Die ontwerpsvereistes vir die omskakelde elektriese voertuig se aandrywingseenheid is gebaseer op die gegewens van die rysiklus. Daarna is ’n Toyota Hiace Ses’fikile minibus-taxi suksesvol omgeskakel in ’n volledig elektriese prototipe, wat die eerste geregistreerde elektriese informele openbare voertuig in die sub-Sahara streek verteenwoordig en empiriese validering moontlik gemaak het. Die gemiddelde gemeetde doeltreffendheid van die aangepaste taxi is aangeteken as 0.342 kWh/km. Die studie het gefokus op die validering en verfyning van ’n energiegebaseerde simulasiemodel genaamd EV Fleet Simulation deur gebruik te maak van empiriese data. Aanvanklik het die ongekorrigeerde model ’n voertuigdoeltreffendheid van 0.53 kWh/km opgelewer. Nadat die voertuigparameters aangepas is op grond van data van die elektriefiseerde taxi, het die simulasie ’n verbeterde doeltreffendheid van 0.372 kWh/km getoon, met ’n gemiddelde absolute fout (MAE) van 18.1% en ’n standaardafwyking van 18.3% vir rite langer as 5 km. Hierdie resultate is aansienlik verbeter deur opeenvolgende korrekties. Uiteindelik het die finale model ’n MAE van 8.61% (standaardafwyking van 10.1%) behaal, wat nou ooreenstem met die gemete energieverbruikwaardes: 0.331 kWh/km gesimuleer teenoor 0.327 kWh/km gemeet vir ritte langer as 5 km. Die navorsing het die potensiaal van masjienleer ondersoek deur modelle op te lei met behulp van GPS-afgeleide ry-siklusdata. ’n Basiese Feed-forawrd Neural Network (FFNN) het ’n wortelgemiddelde-kwadraatfout (RMSE) van 8.52 kW per monster behaal, wat beter presteer het as beide lineêre en polinoomregressiemodelle. In vergelyking met EV Fleet Simulation, het die FFNN ’n laer MAE in kragvoorspelling getoon, met resultate van 5.62 kW teenoor 7.59 kW. Daarbenewens was die MAE vir per-rit energievoorspelling 0.018 kWh/km, minder as die helfte van die simulasie-model se MAE van 0.044 kWh/km. Hierdie bevindings beklemtoon die voordele van die gebruik van masjienleer vir intydse energievoorspellings en haalbaarheidsevaluasies. Die vergelyking tussen gemete en gesimuleerde resultate toon die doeltreffendheid van die benadering. Verder is masjienleermodelle meer gebruikersvriendelik, aangesien hulle slegs GPS-koördinate van die ry-siklus benodig, terwyl energiegebaseerde modelle verskeie parameters rakende die ry-siklusse en voertuie vereis. Die bevindings van hierdie navorsing toon onomwonde die tegniese haalbaarheid van die elektrifisering van paratransportvoertuie in sub-Sahara Afrika. Die empiriese validering van simulasie-instrumente en die belowende toepassing van masjienleer bied ’n stewige grondslag vir toekomstige haalbaarheidsevaluasies en dra waardevolle insigte by tot die optimalisering van die ontwerp en implementering van elektriese paratransportvoertuie in die streek. Doctoral 2025-12-22T13:32:02Z 2025-12-22T13:32:02Z 2025-12 Thesis https://scholar.sun.ac.za/handle/10019.1/134667 en Stellenbosch University xxvi, 219 pages : illustrations application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Internal combustion engines -- Africa, Sub-Saharan Paratransit services -- Africa, Sub-Saharan Transportation -- Planning -- Africa, Sub-Saharan Electric vehicles -- Economic aspects -- Africa, Sub-Saharan Sustainable transportation -- Africa, Sub-Saharan Lacock, Stephan Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles |
| title | Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles |
| title_full | Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles |
| title_fullStr | Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles |
| title_full_unstemmed | Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles |
| title_short | Planning for electric paratransit: the feasibility of electrifying internal combustion engine vehicles |
| title_sort | planning for electric paratransit the feasibility of electrifying internal combustion engine vehicles |
| topic | Internal combustion engines -- Africa, Sub-Saharan Paratransit services -- Africa, Sub-Saharan Transportation -- Planning -- Africa, Sub-Saharan Electric vehicles -- Economic aspects -- Africa, Sub-Saharan Sustainable transportation -- Africa, Sub-Saharan |
| url | https://scholar.sun.ac.za/handle/10019.1/134667 |
| work_keys_str_mv | AT lacockstephan planningforelectricparatransitthefeasibilityofelectrifyinginternalcombustionenginevehicles |