A Stochastic Programming Based Approach For Hospital Location Planning

Hospital location planning is crucial to modern hospital construction. In the health care resource allocation process the demand of health care resource is often stochastic. The way of dealing with this uncertainty has raised increasing focus in location planning researches. Current hospital location planning methods are mostly static deterministic models and therefore can not fully explore the random feature of the problem. This research paper discusses the stochastic characteristic of parameter (health care demand: monthly visits) and proposes a stochastic programming based method to deal with the problem of hospital site selection.In Chapter 3 an integer programming model with stochastic parameter is proposed. Various factors such as traffic condition, patients'preference, etc. are taken into consideration. Stochastic parameter, monthly number of hospital visits, is introduced, with the objective of minimizing total weighted distance between population and health care resources.In Chapter 4, Monte Carlo method is used to simulate stochastic parameter, and results based on numerical test prove good convergence of the Monte Carlo algorithm, as well as its advantage compared to deterministic model.In the following chapter scenario-analysis based method is proposed to simulate the stochastic parameter. Consideration is made on several discrete monthly visits scenarios with associated probabilities to determine the location plan for new general hospital under capacity constraints, convergence of the algorithm is proved by numerical experiments.The numerical experiments are realized using Visual C++ and CPLEX.