A solution approach for the green vehicle routing problem by Q-learning based ant colony algorithm

Today, the rapid increase in consumption has also caused an increase in environmental problems. Enterprises also have major responsibilities in combating environmental problems. In this context, enterprises now need to focus not only on costs but also on issues such as CO2 emissions, recycling and sustainability. In this study, the gren vehicle routing problem is addressed. Since the amount of CO2 emissions decreases with the distance, it is focused on improving green logistics problems by minimizing the distance. First, the Ant ColonyAlgorithm (ACA) was used as a solution approach, and then the Q-Learning Algorithm-based Ant ColonyAlgorithm (Q-ACA) was proposed. The proposed integrated approach was coded with the Python programming language, solved for small, medium and large datasets in the literature and compared with the best solution values. As a result of the experiments, it is seen that beter results are achieved with Q-KKA by taking into account the past experiences of ants. In addition, KKA and Q-KKA were compared with the Genetic Algorithm (GA) approach, which has two different initial solutions used for the same datasets in the literature, and it was seen that Q-KKA achieves beter results than GA in clustered data sets.