具有恶化效应及维护需求的医疗检查调度问题研究

发布时间：2018-08-30 11:32

【摘要】：医疗检查作为现代医学中不可或缺的一个环节,已在医学研究、大数据挖掘、智能诊断等领域得到广泛重视。随着我国老龄化人口不断增加及环境问题的不断加剧,人民生活水平的进一步提高,医疗检查成了老百姓看病取药的首要环节。然而患者年龄结构差异化大,患者自身情况的不确定性,都使得医疗检查调度问题极为复杂化。同时,患者接受检查的及时与否直接影响患者后续治疗。在医疗行业,目前还未对医疗调度问题给予足够的重视,尤其是针对结合医疗器械预防性维护的调度。如何将患者的检查数据合理化,问题的模型实际化,算法的应用普遍化,是从科学层面解决我国医疗行业中"看病慢"问题的有效手段。因此,从医疗合理化和社会高效化的角度出发,研究医疗检查调度问题,尤其是维护调度问题,并建立一套科学合理且适用于我国医疗现状的调度模型及算法,具有重大的现实意义和实用价值。患者作为医疗工作的核心,存在检查时间的不固定性。本课题参考生产调度领域中的恶化工件的情况,分析患者检查时间延误存在病情加重的现象,提出利用分段函数描述医疗检查过程中的时间;同时由于医疗器械需要及时展开预防性维护,在考虑上述情况的同时分析了预防性维护对调度方案的影响。本课题基于患者和设备维护两个方面出发,开展医疗检查调度研究,更加贴近医疗实际,使研究更具指导意义。本文主要工作包括以下四个方面:(1)归纳总结出医疗检查调度问题的研究现状及研究重点,并简要介绍相应的调度算法以及各个方法具备的特点及主要应用范围。通过实际调查研究发现在医疗检查环节中容易出现问题的原因,将患者实际接受医疗的时间特点形象化,将患者检查时间用基本检查时间和惩罚时间两部分描述。将医疗检查器械的预防性维护采用数学模型进行描述,在构建模型过程中结合实际情况,予以适当约束。(2)分析了惩罚值为恶化阶梯函数的单机医疗检查调度问题,以极小化总惩罚值为目标,提出改进遗传算法对问题进行求解,并与其它算法进行对比,验证算法的有效性。(3)将患者的检查时间及维护时间固定化,研究目标函数为最小化滞后检查患者数量的调度问题。构建了改进的变邻域搜索算法,并通过算例说明该算法的实用性及优越性。(4)从集成优化的角度出发,建立了预防性维护下的含有累积恶化效应的医疗检查调度问题模型及改进模型,并且设计了基于邻域搜索的遗传算法对其进行求解。并将求解结果与其他算法进行对比,凸显算法的优越性。
[Abstract]:As an indispensable link in modern medicine, medical examination has been widely paid attention to in the fields of medical research, big data mining, intelligent diagnosis and so on. With the increasing of aging population and environmental problems, and the further improvement of people's living standard, medical examination has become the most important link for the common people to take medicine. However, the difference of age structure of patients and the uncertainty of patient's own situation make the problem of medical examination scheduling extremely complicated. At the same time, the timely examination of patients directly affects the follow-up treatment. At present, medical dispatch problem has not been paid enough attention to in medical industry, especially in combination with preventive maintenance of medical devices. How to rationalize the patient's examination data, the practical model of the problem and the generalization of the algorithm are the effective means to solve the problem of "slow seeing a doctor" in the medical profession of our country from the scientific level. Therefore, from the point of view of medical rationalization and social efficiency, this paper studies the medical inspection scheduling problem, especially the maintenance scheduling problem, and establishes a set of scientific and reasonable scheduling model and algorithm suitable for the current medical situation in China. It has great practical significance and practical value. Patients as the core of medical work, there is uncertainty of the examination time. Referring to the situation of deteriorating workpiece in the field of production scheduling, this paper analyzes the phenomenon that the patient's condition is aggravated in the delay of examination time, and proposes to use the piecewise function to describe the time in the course of medical examination. At the same time, due to the need of preventive maintenance, the influence of preventive maintenance on scheduling scheme is analyzed. Based on the two aspects of patient and equipment maintenance, the research of medical inspection scheduling is carried out, which is closer to the medical practice and makes the research more instructive. The main work of this paper includes the following four aspects: (1) summarize the current research situation and research focus of medical inspection scheduling problem, and briefly introduce the corresponding scheduling algorithm, the characteristics of each method and the main application scope. Through the actual investigation and research, the causes of the problems in the medical examination are found, the time characteristics of the patients' actual medical treatment are visualized, and the time of the patients' examination is described in two parts: the basic examination time and the punishment time. The preventive maintenance of medical examination equipment is described by mathematical model. In the process of constructing the model, appropriate constraints are given according to the actual situation. (2) the scheduling problem of single machine medical examination with the penalty value of worsening step function is analyzed. Aiming at minimizing the total penalty value, an improved genetic algorithm is proposed to solve the problem, which is compared with other algorithms to verify the effectiveness of the algorithm. (3) the patient's inspection time and maintenance time are immobilized. The objective function of this paper is to minimize the number of delay check patients. The improved variable neighborhood search algorithm is constructed, and the practicability and superiority of the algorithm are illustrated by an example. (4) from the angle of integration optimization, The model and improved model of medical inspection scheduling problem with cumulative deterioration effect under preventive maintenance are established, and a genetic algorithm based on neighborhood search is designed to solve the problem. The results are compared with other algorithms to highlight the superiority of the algorithm.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R197.3;TP18

【参考文献】