联系我们

论文查询结果	返回搜索

论文编号：	16148
作者编号：	2120243791
上传时间：	2026/6/9 18:19:15
中文题目：	双阶段多目标家庭护理服务调度优化研究
英文题目：	Study on the optimization of two-stage and multi-objective home care service scheduling
指导老师：	梁峰
中文关键字：	家庭护理服务；前景理论；车辆路径规划问题；多目标优化；干扰管理
英文关键字：	Home health care; Prospect Theory; VRP; Multi-objective optimization; Disruption management
中文摘要：	随着全球人口老龄化程度持续加深，我国医疗护理资源短缺与老龄群体养老需求增长之间的矛盾日益突出。家庭护理服务作为居家养老的重要支撑，能够将专业医疗护理延伸至家庭，有效缓解医疗资源供需矛盾。但现有家庭护理调度研究多聚焦于系统效率目标，对客户主观偏好与行为心理的建模不足，较少考虑护理人员工作负荷，且缺乏对突发干扰事件的动态响应机制。 因此，本文以老龄化背景下家庭护理服务预约调度问题为研究对象，从客户、护理人员与护理中心三方主体视角出发，构建多目标优化模型并设计求解算法，研究涵盖预约调度与动态重调度两个阶段。在预约调度阶段，引入前景理论对客户感知满意度进行量化建模，通过构建客户等待时间满意度函数与对护理人员偏好满意度函数，刻画客户损失厌恶、参照依赖等非线性心理感知特征。在此基础上，以客户综合感知满意度最大化为主目标、护理中心运营成本最小化为次目标，建立多目标调度优化模型，采用粒子群算法与遗传算法双层嵌套结构进行求解。并进一步将护理人员工作负荷纳入调度目标，引入技能匹配约束与加班时长限制，构建以最大化客户满意度、最小化护理人员加班时长、最小化运营成本为目标的多目标优化模型，采用多目标遗传算法求解帕累托最优解集，为护理中心提供多样化的决策方案。在动态重调度阶段，针对护理服务过程中客户临时取消订单、变更时间窗、紧急插单等突发干扰事件，引入前景理论分别量化客户、护理人员和护理中心三方主体对于重调度方案的心理感知价值，构建以系统扰动程度最小化为目标的干扰管理模型。通过设计双阶段遗传算法，将预约调度阶段的最优方案作为参考基准，在设定的干扰事件处理时间点后触发重调度循环，动态调整订单分配与路径规划。通过相关算例仿真运行，验证了所建立的家庭护理服务预约调度模型的有效性，并结合相关灵敏性分析得到管理启示。
英文摘要：	As global population aging continues to intensify, the contradiction between the shortage of medical and nursing resources in China and the growing eldercare demands of the aging population has become increasingly prominent. Home care services, as a vital pillar of home-based eldercare, can extend professional medical and nursing care into the household, effectively alleviating the supply-demand imbalance in healthcare resources. However, existing research on home care scheduling largely focuses on system efficiency objectives, with insufficient modeling of customers' subjective preferences and behavioral psychology, limited consideration of caregivers' workload, and a lack of dynamic response mechanisms for unexpected disruptive events. Therefore, this study addresses the appointment scheduling problem in home care services within the context of population aging. Adopting the perspectives of three stakeholders—customers, caregivers, and care centers—this research develops multi-objective optimization models and designs solution algorithms, encompassing two phases: appointment scheduling and dynamic rescheduling. In the appointment scheduling phase, prospect theory is introduced to quantitatively model customers' perceived satisfaction. By constructing satisfaction functions for customer waiting time and for caregiver preference, the study captures nonlinear psychological perception characteristics such as loss aversion and reference dependence. On this basis, a multi-objective scheduling optimization model is established with the maximization of customers' overall perceived satisfaction as the primary objective and the minimization of care center operational costs as the secondary objective, solved using a bilevel nested structure combining particle swarm optimization and genetic algorithms. The model is further extended to incorporate caregiver workload into the scheduling objectives by introducing skill-matching constraints and overtime limits, resulting in a multi-objective optimization model that maximizes customer satisfaction, minimizes caregiver overtime, and minimizes operational costs. A multi-objective genetic algorithm is employed to obtain the Pareto-optimal solution set, providing care centers with diversified decision-making options. In the dynamic rescheduling phase, to address unexpected disruptive events during the care service process—such as order cancellations, time window changes, and urgent order insertions—prospect theory is applied to quantify the psychological perceived value of the rescheduling plan for all three stakeholders: customers, caregivers, and care centers. A disruption management model is then constructed with the objective of minimizing the degree of system perturbation. Through the design of a two-phase genetic algorithm, the optimal solution from the appointment scheduling phase serves as a reference baseline, triggering a rescheduling loop after predefined disruption-event processing time points to dynamically adjust order assignments and route planning. Simulation experiments on relevant numerical instances validate the effectiveness of the proposed home care service appointment scheduling models, and managerial insights are derived through corresponding sensitivity analyses.
查看全文：	预览  下载(下载需要进行登录)