An improved partial comparison optimization for utilizing landfill facilities in a waste collection vehicle routing problem

The Waste Collection Vehicle Routing Problem (WCVRP) involves optimizing vehicle routes to minimize travel distance, carbon dioxide (CO₂) emissions, and the number of vehicles used. However, existing models often neglect distant landfill sites, leading to imbalanced landfill utilization and shortened facility lifespans. Thus, this study introduces a new constraint to ensure the effective use of all landfill sites in WCVRP. To integrate this constraint, an enhanced Partial Comparison Optimization (PCO) algorithm is proposed. PCO is a single-solution-based metaheuristic previously shown to be effective in solving Vehicle Routing Problems (VRP). The improved PCO incorporates a Nearest Greedy (NG) algorithm for initial solution construction, dynamic parameter adjustment, and two additional neighborhood operators. The algorithm is tested on datasets containing up to 2000 customers across three landfill capacity scenarios. Scenario 1 assumes unlimited landfill capacity, Scenario 2 distributes waste equally among landfills, while Scenario 3 assigns higher capacities to certain landfills. The results indicate that Scenarios 2 and 3 increase travel distances and vehicle counts compared to Scenario 1. However, both scenarios achieve a more balanced distribution of landfill usage. On average, NG produced 1157.37 kg of CO₂ emissions, reduced by 3.49% using PCO and 10.07% with improved PCO. Similarly, NG’s travel distance of 1208.11 miles decreased by 15.77% with PCO and 29.57% with improved PCO. Additionally, NG required 106 vehicles, which is reduced by six through PCO and improved PCO. These findings demonstrate that the improved PCO significantly outperforms both the NG and the existing PCO. Consequently, the proposed approach offers a more efficient and environmentally sustainable solution to WCVRP. This study provides valuable insights for waste management authorities in selecting operational strategies that optimize both resource utilization and environmental impact.