Improved P-hub Network Model and GA Solution Based on Rough Set Theory

To solve NP-hard 0-1 discrete optimization problems of the uncapacitated multiple allocation p-hub median network (UMpHMP), this paper proposed an improved p-hub network model based on - approximate rough set data mining technology to reduce the range of hub choice from n alternative hubs to the limited airports, which greatly reduces total number of the variables and constraints for the models. As one of the classical shortest path problems, the genetic algorithm was developed to solve the improved model. In order to illustrate the effective of new model, an experimental example of domestic 15 cities route network designing for airlines was given, the simulation results show the established index number can be abbreviated to 4 main attributes by - reduction, and the efficiency of solution was improved for the modified network.