Skripsi
OPTIMASI TRANSPORTASI PINTAR MENGGUNAKAN HIBRID METAHEURISTIK DRAGONFLY ALGORITHM (DA)
Smart transportation optimization in modern logistics systems requires route efficiency to minimize operational costs and travel time. However, the Vehicle Routing Problem with Time Windows (VRPTW) is an NP-hard problem that is highly complex to solve. The standard Dragonfly Algorithm (DA) has strong global exploration capabilities but weak local exploitation, which often causes it to become trapped in suboptimal solutions. This study aims to develop a smart transportation optimization model through the hybridization of the Optimized Dragonfly Algorithm (ODA) and Variable Neighborhood Search (VNS) to minimize total travel distance and the number of vehicles used. The research methodology includes the development of ODA enhanced with adaptive weighted initialization strategies, constraint repair mechanisms, and heuristic route refinement. This model is integrated with VNS as a local search intensification phase using eight neighborhood movement techniques, such as relocate, exchange, and two-opt. The evaluation was conducted on 29 instances from the Solomon Benchmark Dataset classes C1, C2, and R1. The results show that the hybrid ODA-VNS model provides more stable and competitive performance compared to the pure ODA approach. For the C1 and C2 dataset classes, the algorithm successfully achieved the Best Known Solution (BKS) with a 0.00% gap in the majority of test instances. In the R1 dataset class, which features randomly distributed customers, ODA-VNS significantly reduced distance deviation and optimized the number of vehicles used. Benchmarking results against the HHHSA, HGA, and LNS-MRSO algorithms demonstrate the superiority of the proposed model in generating precise routing solutions. This study provides a significant contribution to the development of intelligent logistics systems that require high efficiency in vehicle route scheduling.
No other version available