Solving the time window vehicle routing problem with improved artificial bee colony and firefly algorithms

Today, companies face difficulties sustainable and increasing profitability in a competitive environment. In order for companies to gain competitive advantage and improve their profitability in such an environment, company management needs to seek and implement innovative strategies in the field of logistics. It is very important and vital for the prestige of the companies to closely follow the developments in logistics management in order to deliver the products and services they need to the customers on time and at the desired location. In addition, while meeting the needs of customers, providing better working platforms for their own staff with minimum cost and maximum profit is the main purpose of companies. For this reason, companies focus on studies related to vehicle routing problems, which have an important place in the field of logistics. Vehicle routing problems, which have a wide-ranging importance in reducing transportation costs and increasing service quality and efficiency for logistics companies, are classified as NP-hard problems. Therefore, the solution of these problems with precise optimization methods in a reasonable time becomes more difficult as the problem size increases. Generally, the most suitable solution is tried to be found by using heuristic and metaheuristic methods. In this study, the time window vehicle routing problem, which is an extension of the vehicle routing problem, is discussed. Artificial Bee Colony and Firefly Algorithms, which are metaheuristic methods, have been proposed to find the appropriate solution to the problem. In the Artificial Bee Colony Algorithm, food sources (solutions) in the initial and scout bee stages, with the Nearest Neighbor heuristic; neighboring food sources in the worker bee and onlooker bee stages were created with the operators of insertion, replacement and random insertion of the subset. In the Firefly Algorithm, the initial fireflies were determined by the Nearest Neighbor algorithm, and the new fireflies were determined by the insertion, replacement and 〖2-opt〗^*operators. In order to investigate the effectiveness of the proposed methods, experiments were carried out with 58 known test problems and the results were compared with previous studies. As a result, Artificial Bee Colony Algorithm method provided the best results in terms of solution quality and number of vehicles. The best result in terms of solution time was obtained with the Firefly Algorithm.