Evolutionary Multi-Objective Optimization of a Multi-Echelon Humanitarian VRP for Smart City Logistics Under Uncertainty
MATHEMATICS, cilt.14, sa.14, ss.1-23, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 14 Sayı: 14
- Basım Tarihi: 2026
- Doi Numarası: 10.3390/math14142500
- Dergi Adı: MATHEMATICS
- Derginin Tarandığı İndeksler: Academic Search Ultimate (EBSCO), Scopus, Materials Science & Engineering Collection (ProQuest), Technology Collection (ProQuest), Aerospace Database, Science Citation Index Expanded (SCI-EXPANDED), zbMATH, Directory of Open Access Journals
- Sayfa Sayıları: ss.1-23
- Galatasaray Üniversitesi Adresli: Evet
Özet
Efficient and resilient humanitarian logistics is critical for smart cities facing large-scale disasters, where infrastructure disruptions, uncertain demand, and time-critical deliveries complicate operational planning. This study proposes a multi-echelon vehicle-routing framework that integrates trucks, electric unmanned aerial vehicles (UAVs), and micromobility systems under uncertainty. Demand and travel-time variability are modeled through scenario-based representations, while delivery flexibility is captured through triangular fuzzy time windows defined by earliest-acceptable, preferred, and latest-tolerable delivery times. The problem is formulated as a four-objective optimization model that minimizes total cost, response time, CO2 emissions, and fuzzy lateness. To solve the resulting highly constrained multi-objective problem, NSGA-II and NSGA-III are adapted with problem-specific repair operators, including capacity-splitting, range-feasibility correction, and fuzzy time-shift adjustment mechanisms, and embedded in a simulation-based evaluation framework. The proposed approach is validated using a geo-referenced earthquake scenario in Istanbul, constructed from open-source GIS, traffic, and demographic data. The computational results show that evolutionary methods generate feasible solutions within minutes, whereas exact optimization approaches fail to converge for realistic instances. Compared with NSGA-II, NSGA-III achieves superior performance, including a 9% higher hypervolume and improved robustness under stress-test scenarios. Furthermore, the hybrid truck–UAV–micromobility strategy reduces average cost by up to 30%, delivery time by 43%, and CO2 emissions by 65% relative to a truck-only baseline, while eliminating fuzzy lateness. These findings demonstrate that evolutionary multi-objective optimization provides an effective and scalable decision-support framework for uncertainty-aware and sustainable humanitarian logistics in smart cities.