Penerapan Kombinasi Genetic Algorithm dan Iterated Local Search Pada Multi-Depot Capacitated Vehicle Routing Problem

Abstract

The distribution is a system for distributing goods or products from the company to customers. Multi-depot capacitated vehicle routing problem (MDCVRP) is a variation of the vehicle routing problem (VRP) which is based on distribution problems and MDCVRP is a topic of optimization problems in applied mathematics. In this study, the authors apply a combination of two metaheuristic algorithms, namely the Genetic Algorithm (GA) and Iterated Local Search (ILS), hereinafter referred to as the GA&ILS algorithm. This study aims to analyze the effectiveness and efficiency of the application of the GA&ILS algorithm to solve MDCVRP on 20 simulation data grouped into four sizes (25, 50, 75, and 100 customer points). Based on the results of the research, it was found that the GA&ILS algorithm is quite effective for small-scale data, but less effective for large-scale data. From the results of the analysis carried out, the GA&ILS algorithm still has drawbacks, it is less able to avoid being trapped by the local optimum. In addition, the GA&ILS algorithm requires a long computational time, making it less efficient.