The Application Of Progressively Interactive Evolutionary Multi-objective Optimization Algorithm In Waste Disposal Facilities’ Location Problem

Recent years, with the increasing of urban population and the improvement of people’s consumption capacity, municipal living waste production grows rapidly. Municipal living waste disposal facilities’location problem is an important research direction in reverse logistics research. It’s not only the basis for efficient garbage collection processing system, but also has a direct impact on people’s living environment. Facility location problem has been the concern of experts and scholars. Because of the municipal solid waste’distributive and hazardous characteristics, so when come to the location problem of the waste disposal site not only need to consider the cost factor but also should consider the social and environmental factors. In recent years, land prices continue to rise, the price of land falls far of the city center and the suburbs. Therefore, as the decision-making authorities-the government should take every factor into account.In this paper, we improved the traditional waste disposal station location model by incorporating the land price factor, and consider the general distribution of urban land prices. For solve the multi-objective municipal living waste disposal facilities’ location model aiming at minimal costs (including land costs, construction costs, and the total transportation costs) and the negative effects on residents, a progressively interactive evolutionary multi-objective optimization (PI-EMO) algorithm was proposed. We periodically added the decision maker’s preferences to progressively improve the domination principle in directing algorithm’s search to more preference solutions. It can avoid the search of the entire Pareto front. Through solving the model, the locations and numbers of these sites and how to assign the generation sites can be determined. Finally, the computational results showed that:1) the algorithm can quickly find the Pareto optimal solution in conformity with the preferences of decision makers;2) compared with the NSGA-Ⅱ, the algorithm can effectively reduce the computation time, and improve the efficiency to solve this problem.