The structural complexity of modern distribution networks necessitates the simultaneous optimization of strategic facility location and operational routing to mitigate suboptimalities inherent in decoupled planning. This research develops a robust two-stage metaheuristic framework to solve the Capacitated Location-Routing Problem (CLRP) by minimizing the sum of fixed depot expenditures and variable transportation costs. The first phase employs a Location-Allocation (LAP) mechanism based on Euclidean distance minimization to establish optimal depot-customer clusters. In the second phase, a high-performance Genetic Algorithm (GA) integrated with an Evolutionary Path Relinking strategy is implemented using Evolver software to refine vehicle routing plans. The model’s efficacy was rigorously validated using Prins’ and Prodhon’s benchmark instances, specifically analyzing 50-customer configurations across varying vehicle capacities. Computational experiments demonstrate that increasing vehicle capacity and integrating decision layers significantly enhance network efficiency, resulting in substantial reductions in total logistics costs. The findings provide a scalable, analytically rigorous decision-support tool for industrial engineers seeking to optimize complex supply chain architectures under strict capacity and service-level constraints.

Keywords

CLRP Metaheuristic Synthesis, Evolutionary Path-Relinking, Dual-Phase CLRP Optimization, Evolver-GA Heuristic, Benchmark-Driven Network Design.