Research On Flexible Medical Scheduling Problem Based On Imperialist Competitive Algorithm

The outbreak of Novel Coronavirus Pneumonia has aroused people’s great attention to medical and health issues.Effective medical scheduling is of great significance to the development of medical cause.Based on the two realistic constraints,i.e.,the preparation time of the physical examiner and the switching time between the physical examination equipment,in this paper,the medical scheduling is modeled as the flexible job shop scheduling problem(FJSP).Aiming at the physical examination scheduling problem,outpatient scheduling problem,surgery scheduling problem in the medical scheduling,the mixed integer linear programming(MILP)models are established respectively.The structural characteristics of the medical scheduling problems are analyzed.The imperialist competitive algorithm(ICA)is utilized to solve the medical scheduling problems.In addition,several efficient algorithm strategies are included into the algorithm framework.The main contributions and achievements are as follows:1.For the physical examination scheduling problem in medical scheduling,the objective is to minimize the maximum completion time of the physical examiner.The sequence-based MILP model is established.The ICA with global search strategy is designed.Then,A two-vector coding scheme is adopted to generate a feasible coding solution.A local search strategy similar to simulated annealing(SA)is designed,and a global search strategy is proposed to improve the comprehensive performance of the algorithm.2.For the outpatient scheduling problem in medical scheduling,a hybrid optimization algorithm framework combining ICA,SA and estimation of distribution algorithm(EDA)is designed.The position-based MILP model is established.The mutation strategy with mutation probability is constructed,which avoids premature convergence and maintains population diversity.The diversified assimilation strategy is adopted to improve the global search ability of the algorithm.3.For the surgery scheduling problem in medical scheduling,an improved multi-objective ICA is proposed to minimize the maximum completion time of surgery and medical cost.The social hierarchy strategy is developed to initialize the empire,which improves the search efficiency of the algorithm.The concept of attraction and repulsion(AR)is introduced into the assimilation strategy to improve the global search ability of the algorithm.To increase the diversity of solutions,the revolution strategy is used.The variable neighborhood search(VNS)strategy is embedded to improve the local search ability of the algorithm.4.For the surgery rescheduling problem with fuzzy surgery time in medical scheduling,the ICA based on reinforcement learning(RL)is proposed,where a diversified revolution strategy including six mutation operations is designed to enhance the local search ability of the algorithm.A new competition strategy for imperialist is designed to improve the global search ability of the algorithm.In addition,the RL algorithm is adopted to adjust the parameters adaptively and improve the robustness of the algorithm.