【作者】 李彬；

【导师】 左兴权；

【作者基本信息】 北京邮电大学 ， 计算机技术， 2016， 硕士

【摘要】 随着人们对医疗资源的需求的不断增大,一些大型医院的接诊的病人数量不断增多。急诊科又是医院中重症病人最集中、病情最紧急、抢救和管理任务最重的科室。一个好的布局方案能够增强医院服务水平,减少病人行走距离,进而提高医院急诊科室吞吐率和提高病人满意度,因此医院急诊科室布局问题具有非常重要的研究意义。本文以一家医院的急诊科室作为研究对象。通过对医院建筑布局图的观察和实际调查,我们将该医院急诊科室的布局问题转化为了三行科室布局问题,并设置了一些约束条件。通过对医院提供的接诊数据的整理我们了解了科室之间的病人流量和不同病人在急诊科的就诊流程。为了求解医院急诊科室布局问题,我们对影响急诊科室布局的因素做了分析,决定将流量因素和非流量因素两个影响因素作为优化目标。为了能够解决这个问题,在算法选择上,先后使用了局部搜索算法、模拟退火算法和多目标模拟退火算法三种算法。为了表示一种布局方案,我们设计了一种解的编码方式来表示。为了减少算法的运行时间,使用了最小流量优先启发式算法来产生初始解。在搜索邻域解时,设计了三种适合医院急诊科室布局问题的操作算子。在多目标模拟退火算法中,对于新解的处理,根据支配关系使用了三种非支配解检测及相应的处理方法。三种算法的优化结果,对比原始解都有一定的优化效果,有效性上多目标模拟退火算法最好,模拟退火算法次之,局部搜索算法最差。为了验证算法的正确性,我们使用了一款专业的医院仿真软件MedModel来进行仿真。通过对原始解和从多目标模拟退火算法得出的解集中选出的两个解进行仿真,仿真软件给出的统计结果验证了多目标模拟退火算法的有效性。多目标模拟退火算法得出结果显示,流量成本平均减少了 11.57%,关系紧密度提升了 27.3%。多目标模拟退火算法能够有效的解决医院急诊科室布局问题。更多还原

【Abstract】 With the increasing demand for medical resources,the number of patients is growing rapidly.And the emergency department of hospital is the part in hospital where the patients is most concentrated,the condition is most emergency and the task of rescue is most urgent.So a good layout scheme can improve the service level of hospital by reduce patients’walking distance,and then the throughput of emergency department and patients’ satisfaction while also improved.Therefore the problem of the layout of hospital emergency department has an important research significance.In this paper,we take a real hospital emergency department as the research object.Through the observation of the hospital building layout and the actual investigation,we transformed the layout problem of the emergency department to three rows layout problem,and we put forward some constraints to the problem.By organizing hospital admissions data,we know the patient flow between departments and different treatment processes of different patient in the emergency department.To solve the problem of hospital emergency department layout,we analyzed the factors that affect the layout of the emergency department,and decided to take the two factors as the optimization objective.In order to solve this problem,we tried thee algorithms:local search algorithm,simulated annealing algorithm and simulated annealing algorithm for multi-objective.To represent the layout scheme,we designed a solution for encoding.In order to reduce the running time of the algorithm,we use the minimum flow first heuristic algorithm to generate the initial solution.Three operators are devised to produce a neighbor from the current slution.In the multi-objective simulated annealing algorithm,we used three non-dominated solutions testing and the proper disposal method based on relations of domination to deal with the new solution.Compared to the original solution,the three algorithms archive better optimization results,the effectiveness of the multi-objective simulated annealing algorithm is the best,the simulated annealing algorithm is the second,and the local search algorithm is the worst.In order to verify the correctness of the algorithm,we use a professional hospital simulation software MedModel to carry out simulation.Compare to the two solutions of the simulation of the original layout and the multi-objective simulated annealing algorithm,the simulation results verify the algorithm.The results of Multi-objective simulated annealing algorithm show that the average cost of the flow is reduced by 11.57%,the relationship between the degree of tightness increased by 27.3%,compared with the original layout.Therefore,we can get conclusion that the multi-objective simulated annealing algorithm can effectively solve the problem of hospital emergency department layout.更多还原