Loading...

A Traffic-Aware Approach to the Time-Dependent Generalized Pollution Routing Problem

Soroush Hassani | 2025

72 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 58649 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Varmazyar, Mohsen
  7. Abstract:
  8. This thesis models and solves a time-dependent generalized pollution-routing problem with explicit social considerations. The objective is to balance total operating cost and pollutant emissions while maintaining schedule regularity. We propose an integrated framework that jointly determines routing, departure times, and cruising speeds and evaluates solution quality through academically established Pareto-front indicators. For small instances, a mixed-integer linear programming formulation is developed and solved exactly with commercial optimizers. For medium and large instances, two metaheuristics are employed, namely a genetic algorithm based on NSGA-II and a multi-objective ant colony optimization scheme. The embedded speed–departure optimization subproblem is reformulated as a shortest-path problem on an event network, solved exactly, and used as the fitness evaluator within the metaheuristics. Benchmark instances from pollution-routing studies are adapted to the time-dependent setting; distance- and demand-based clustering is applied to improve scalability. Performance is assessed using generational distance, its improved variant, distance to the ideal point, spacing uniformity, and computing time. Computational results show that the exact approach yields high-quality Pareto fronts on small instances but its runtime grows rapidly with size. The proposed metaheuristics generate more stable and uniform fronts at substantially lower computational cost. Sensitivity analysis indicates that longer congestion intervals shift the Pareto front toward more expensive and tardier solutions, whereas higher traffic speeds enable simultaneous improvement in both objectives. The main contributions are a unified and implementable formulation for the time-dependent problem, an exact integration of the speed–departure module within the evaluation pipeline, and two scalable metaheuristics for large-scale instances
  9. Keywords:
  10. Non-Dominate Sorting Genetic Algorithm (NSGAII) Method ; Multi-Objective Max-Min Ant Colony Algorithm ; Fuel Consumption Optimization ; Speed Optimization ; Departure Optimization ; Generalized Pollution Routing Problem ; Time-Dependent Vehicle Routing

 Digital Object List

 Bookmark

No TOC