Optimizing EV truck fleet charging schedules on depot

As the commercial transport sector transitions toward electrification, the need for intelligent energy management tools becomes increasingly critical. This thesis presents a modular, datadriven optimization framework for scheduling electric truck fleet charging at a single depot. Developed alongside...

Descripción completa

Detalles Bibliográficos
Autor: Martinez Sanchez, Julian David
Tipo de recurso: tesis de maestría
Fecha de publicación:2025
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/444949
Acceso en línea:https://hdl.handle.net/2117/444949
Access Level:acceso embargado
Palabra clave:Electric vehicles
Mathematical optimization
Battery charging stations (Electric vehicles)
Vehicles elèctrics
Optimització matemàtica
Estacions de càrrega (Vehicles elèctrics)
Àrees temàtiques de la UPC::Enginyeria elèctrica
Descripción
Sumario:As the commercial transport sector transitions toward electrification, the need for intelligent energy management tools becomes increasingly critical. This thesis presents a modular, datadriven optimization framework for scheduling electric truck fleet charging at a single depot. Developed alongside a partner job with Scania, the tool generates cost-minimizing charging plans by aligning operational constraints, such as truck availability, charger power limits, and State-of-Charge (SoC) requirements, with dynamic electricity prices from the Nord Pool dayahead market. The optimizer was implemented in Python using linear programming techniques (Pyomo + GLPK) and deployed entirely through Scania’s cloud-based CI/CD infrastructure to ensure automation and scalability. A detailed case study using real operational data from a Swedish Scania customer demonstrated substantial benefits: over 90% cost savings compared to flat-rate electricity contracts, improved km/SEK efficiency as routes lengthen, and a clear sensitivity to infrastructure limitations and fleet scale. Sensitivity analyses further quantified how small reductions in charger count or fleet expansions could lead to infeasibility, emphasizing the importance of strategic infrastructure planning. The tool is modular, requiring only standardized input files and no local computation, making it adaptable across clients and markets within the Nord Pool region. While the current version focuses on static, day-ahead planning for a single depot, proposed extensions include multidepot scheduling, battery degradation modelling, real-time re-optimization, and integration of telematics-based consumption forecasts. This thesis contributes both a practical tool for fleet electrification and a research foundation for future development. By combining operational realism with robust optimization, the methodology helps bridge the gap between sustainable transport goals and real-world implementation challenges.