A Subsidized Public Medical Service Facility Network Optimization Under Two Decision-makers

The network layout of public medical service facilities directly affects the level of public health service,and determines the fairness and harmony of urban and rural public medical service system construction.The network layout of public medical service facilities is of great significance to improve the health level of the whole people and promote social harmony.With the increasing gap between urban and rural development in terms of government public input and residents' income,public medical service resources show significant differences between urban and rural areas and groups.As a result,there is a lack of good matching degree between the total supply of public medical service and the increasing demand of public medical service,which seriously affects the improvement of public health service level.Therefore,it is urgent to optimize the public medical service facility network.Since the renovation project of public medical service facility network relies on a large number of human and material resources,it requires the coordination of the government department and the public health related functional department to make joint decisions and formulate reasonable and effective network optimization schemes.This also makes the optimization problem of the public medical service facility network more complex.The challenge of the network optimization of the public medical service facility is mainly reflected in the description and modeling of the game relationship between the two decision-makers and the design of the effective solution method of large-scale discrete bilevel programming with black box model.On the one hand,in the optimization problem of public medical service facility network,the government department provides financial subsidy for the renovation project of the public medical service facility network,and devotes herself to improving the utility of the public medical service facility network and improving public health service level.The public health related functional department carries out the specific facility location and the service area division with the support of the government financial subsidy.The functional department is more inclined to pursue the maximization of operating benefits.Therefore,the decision-making goals of the government department and the functional department are in conflict.The decision-making process between them belongs to Stackelberg's game.In this thesis,combining the theory of bilevel programming and the characteristics of the optimization problem of the public medical service facility network,we use discrete bilevel programming to model the optimization problem of the subsidized public medical service facility network under two decision-makers.On the other hand,it is extremely challenging to solve the subsidized public medical service facility network optimization problem under two decision-makers.For example,the measurement of the utility of the public medical service facility network is difficult to directly analyze,the black box model of the utility function constructed based on historical data is difficult to express analytically,the feasible region of discrete bilevel programming is very complex and difficult to describe,and the solution of large-scale discrete bilevel programming instances in practical problems is extremely challenging.How to design effective data-driven methods and efficient exact algorithms to solve large-scale discrete bilevel programming with black box model is the key to the optimization problem of subsidized public medical service facility network under two decision-makers.Based on the characteristics of the public medical service facility network optimization problem,this thesis tailores a strengthened branching rule for the discrete bilevel programming model,designs an exact branch-and bound algorithm,and combines with the practical examples and data to design the simulation method,so as to provide the optimal solution for the large-scale public medical service facility network optimization problem.The main research contents are as follows:(a)Problem analysis and modeling of subsidized public medical service facility network optimization under two decision-makers.The thesis defines the optimization problem of subsidized public medical service facility network under two decision-makers.The thesis describes the decision-making objectives of the government department and the functional department,and analyzes the decision-making process of Stackelberg game between two decision-makers.Combined with the government financial subsidy scheme,macro guidelines and physical conditions of the public medical service facility site selection,a universal discrete bilevel programming model is established.A simulation scheme is also designed based on the analytic representation of the network utility function by stratified sampling and symbolic regression.(b)The solution method of public medical service facility network optimization problem with two decision-makers under total subsidy situation.In view of the structure characteristics of the public medical service facility network optimization model under total subsidy situation,the upper level contains one-dimensional discrete decision variable and the lower level contains multidimensional discrete decision variables,an enhanced branching idea is proposed,which can effectively improve the efficiency of general branching rules.Through numerical experiments,it is found that the enhanced branching idea is more effective in accelerating the solution of large-sized instances and can effectively shorten the calculation time of large-sized instances.(c)The solution method of public medical service facility network optimization problem with two decision-makers under the targeted-subsidy list situation.Aiming at the optimization problem of public medical service facility network under targeted-subsidy list situation,an efficient and accurate algorithm in the branch-and-bound framework is designed to solve the general discrete bilevel programming.In the numerical experiments,a set of experimental instances is established,including all the general bilevel integer programming examples that have been published in the field of discrete bilevel programming algorithm research,as well as the large-sized randomly generated instances in this thesis.Compared with the best exact algorithms in the literature,the computational results show that the algorithm proposed in this thesis has obvious advantages in solving the largest and most complex instances.(d)Application research.Combined with a certain area of emergency medical history data and traumatic events,this thesis applies the theory study of the subsidized public medical service facility network optimization under two decision-makers to the trauma center network optimization problem decided by the government department and the hospital system,to provide optimal solutions for the trauma center network redesign.With computational experiments,the effectiveness of the optimized trauma center network can be significantly improved.The status of frequent mistriage errors in grading diagnosis and the treatment due to uneven distribution of trauma centers in the medical emergency system in this region can also be improved.It helps provide better emergency medical services to citizens in the region,and then verify the validity and practicability of the proposed methodology in this thesis.This thesis,in the theoretical level,can enrich the applications of Stackelberg game theory,the research of data driven theory and method,and the research of discrete bilevel programming.It can also enrich the research of the network optimization problem based on data-driven method and fill in the field of the exact algorithm design of the discrete bilevel programming problem for large-scale instances.This thesis,in the application level,can significantly improve the effectiveness of the public medical service facility network and the service level,maintain the operation and management of public medical service facilities,and effectively improve the defect of the uneven distribution of the public medical service facility network.It can also improve the public medical service supply and compatibility between the increasing demand of public medical service,and thus promote the level of public health service.