Distribution Center Location Problem With An Example Of Sugarhouse Purchasing Stations

Facility location is about how to determine the location of facilities scientificallywith the purpose of enhancingcorporate profits, reducinglogistics costs andimprovingthe efficiency of society. Decision-makers have to consider many factors,such as population, economy, traffic, environment, policy and enterprise developmentstrategy. Providing valuable reference for decision-makers quickly throughmathematical models and algorithms is important.Models and methods related tofacility locationhave been carried out research adequately, and have been widely used inreal life. Takingthe site selection of collectingstations fora sugarhouse as an example,this paper not only provides the enterprise with a solution of facility location, but alsoprovidesan example for similar problems.This paper first introduces the background and logistics situation of the sugarhouse,then introduces the history of research on facility location problems. In accordance withthe order of siting problemsstudies, this paper introduces the models and solvingmethods of facility location problems. After defining afunctionin whichthe degree ofcoverage is a decreasing function of the distance to the nearest facility between themaximum and minimum coverage, a bi-objective facility location model which containstwo goals of market share and total logistics cost is established. Meanwhile, this paperalso establishes a set covering model and a P-center model, both of which can providehelp in determining a recommended solution.Next, the paper uses the NSGA-II to solve the multi-objective facility locationmodel and gets an approximate Pareto optimal solutions set. The set covering model andP-center model are solved by ILOG Cplex. A preliminary recommended solution isfound from approximate Pareto optimal solutions set based on the goal of maximizingcompany’s profit. Then this paper solves the set covering model and P-center modelwith the same number of facilities as the preliminary recommended solution. The resultsshow that the exact solutions obtained outperform the preliminary recommendedsolution in terms ofcoverage percent and corporate profit, and P-center model is betterthan set covering model. Finally, artificial adjustment is applied on the exact solutionobtainedby P-center model and the establishment of a station in the southernmost pointis canceled. As a compensation, the enterprise should pay some freight subsidy. After a further comparison between these solutions, the adjustedsolution is identified as thefinal recommendation.In actual decision, decision-makers can choose solutions from non-dominatedsolutions set, set covering solutions set and P-center solutions set according to suchfactorsas layout opinion, budget or traffic situation. The final solution obtainedby thispaper is suggested as a recommended solution for decision-makers.