Solving the Prize-collecting Location Routing Problem in Chinese Rural Logistics Network by a Bi-level Genetic Algorithm

The prize-collecting location and routing problem (PCLRP) is a new variant of the location-routing problem (LRP), which considers the profit that can be obtained by visiting each demand node. In this research, the PCLRP is applied to the planning of rural logistics network to help the enterprise make a trade-off between the revenue and cost. The model takes the maximization of enterprise profit as the optimization objectives, in which enterprise profit equals the difference between the enterprise distribution revenue and cost. Then, a bi-level genetic algorithm with a new crossover method are proposed to solve the PCLRP. The model and algorithm are validated by the benchmark instances, and the effects of different parameter values on the target value of the model are analyzed. The results show that as the price of unit demand rises, PCLRP makes the rural logistics enterprises profitable earlier than the LRP. Meanwhile, there is a strong relationship between the size of the enterprise profits and the number of demands being satisfied. The results can provide relevant decision-making basis for government and enterprises.