Document Type

Thesis

Degree Name

Master of Applied Computing

Department

Physics and Computer Science

Program Name/Specialization

Applied Computing

Faculty/School

Faculty of Science

First Advisor

Dr. Dariush Ebrahimi

Advisor Role

Supervisor

Second Advisor

Dr. Sukhjit Singh Sehra

Advisor Role

Co-supervisor

Abstract

Electric vehicles (EVs) are becoming central to sustainable and intelligent transportation systems, particularly in urban logistics, where reducing emissions, congestion, and operating inefficiencies is increasingly important. Despite their environmental advantages, the large-scale deployment of EVs in delivery operations remains constrained by limited driving range, time-consuming energy replenishment, and the difficulty of making routing decisions under dynamic traffic and load-dependent energy consumption. Addressing these challenges is essential for improving the practicality and reliability of EV-based freight transportation.

This thesis investigates optimization frameworks for EV routing in urban delivery systems by studying two complementary energy replenishment strategies: modular battery segment swapping and the joint use of charging stations (CS) and electric road systems. In the first part, the thesis introduces an Electric Vehicle Routing Problem with Battery Segment Swapping (EVRP-BSS), where a single electric delivery truck dynamically adapts its route using real-time traffic information while deciding when, where, and how many battery modules to swap. In the second part, the thesis studies EV routing in networks equipped with conventional charging stations and dynamic wireless charging (DWC) through electric road systems (ERS), where the vehicle must determine efficient delivery sequences, route choices, and charging modality decisions under battery and traffic constraints. In both problems, vehicle energy consumption is modeled as load-aware, reflecting the fact that battery depletion changes as payload decreases during deliveries.

To solve these two problems, this thesis formulates problem-specific Mixed-Integer Linear Programming (MILP) models and investigates multiple scalable solution approaches. The MILP models are used to obtain optimal solutions for small-scale instances and to provide rigorous benchmark baselines. To address larger instances, where exact optimization becomes computationally prohibitive, the thesis develops and evaluates several heuristic and metaheuristic methods. For the modular battery segment swapping problem, these include the proposed heuristic called Minimizing Travel Time and Energy Swapping Cost (MT2EWC), a Genetic Algorithm (GA), Clarke and Wright Savings (CW) Algorithm, and Ant Colony Optimization (ACO). For the electric-road-integrated routing problem, the thesis similarly examines GA, CW, and ACO as scalable alternatives to the optimization model.

Computational results demonstrate that the proposed methods effectively balance solution quality and scalability. The battery-swapping framework achieves strong real-time performance while maintaining near-optimal behavior on small instances and robust operation on large-scale and real-world road networks. The electric-road routing framework shows that strategically exploiting electrified road segments can reduce total delivery time compared with relying only on stationary charging infrastructure. Overall, this thesis provides a complementary contribution to intelligent EV logistics by advancing real-time, energy-aware routing models that improve operational efficiency, enhance vehicle utilization, and support the development of more sustainable urban transportation systems.

Convocation Year

2026

Convocation Season

Spring

Download

Available for download on Tuesday, October 27, 2026

Share

COinS