The Application of a Hybrid Genetic Algorithm in distribution of goods with simultaneous pickup and delivery

Nowadays, a large volume of daily traffic is caused by vehicles of distribution centers which are delivering the commodities to the customers. The important point is that some customers, in addition to the delivered goods, have an amount of commodities that has to be collected. However, many distribution centers use separate batches of vehicles to pickup and deliver goods. This leads to increasing transportation costs as well as daily traffic. Therefore, a model is needed to organize the distribution routes with the minimum total costs and number of vehicles. In this study, we’ve used VRPSPD model to minimize total costs of transportation and to obtain the optimal routes. We have presented a new hybrid meta heuristic algorithm to solve the model. The proposed method consists of genetic algorithm and Nearest Neighborhood Heuristic (NNH). The initial population is generated with the help of NNH method. Thereafter, the genetic algorithm is used to continue the solving procedure. Ultimately, several standard numerical test instances have been solved to evaluate the performance of the proposed method comparing to classical genetic algorithm and Simulated Annealing method. As a result, the proposed algorithm outperformed the other two methods.