Scholarlink

<p>Home healthcare services play a vital role in ensuring that elderly, disabled, and chronically ill patients receive personalized, high-quality care in the comfort of their own homes. In this context, efficient and adaptive planning tools are essential to support complex decision-making under uncertainty. This study proposes an intelligent Decision Support System (DSS) designed to assist healthcare planners in the scheduling and routing of home healthcare tasks. The proposed DSS is based on a novel hybrid three-stage approach that integrates fuzzy logic, grey relational analysis, and a multi-objective optimization model solved using the Non-dominated Sorting Genetic Algorithm II (NSGA-II). In the first stage, potential healthcare facility locations are assessed through a knowledge-based framework combining fuzzy logic and grey relational analysis to handle uncertainty and prioritize criteria. The second stage formulates a robust multi-period, multi-objective model with three key goals: minimizing total operational costs, avoiding the misallocation of overqualified staff, and maximizing qualitative service attributes. In the final stage, NSGA-II is employed to solve the optimization problem and generate high-quality, Pareto-optimal scheduling and routing solutions. Experimental results on benchmark datasets confirm that the hybrid NSGA-II approach significantly outperforms baseline strategies (e.g., Simple Random Assignment, Greedy Matching) across multiple performance metrics, including travel time reduction, workload distribution, and patient satisfaction. For a representative case involving 25 patients and 7 nurses, the proposed approach successfully served 100% of patients while reducing total travel distance to as low as 562 units, achieving the highest objective function value of 23.13 among competing solutions. Monte Carlo simulations further demonstrate the robustness of the generated schedules, with up to 93.33% of patients served within their time windows (V0) and over 98% service feasibility when allowing minimal time-window violations (V1–V3) in several benchmark instances. Compared to baseline strategies and a state-of-the-art robust optimization model, the proposed fuzzy-NSGA-II approach achieves comparable or superior solution quality while reducing average computational time by up to 22% on large-scale instances. This highlights the system’s practical relevance and computational efficiency for real-time deployment in healthcare settings. The implementation code is publicly available at: <a href="https://github.com/DrMaroueneCHAIEB/HCSP_Fuzzy">https://github.com/DrMaroueneCHAIEB/HCSP_Fuzzy</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-objective optimization for home health care scheduling under uncertainty using fuzzy logic and non-dominated sorting genetic algorithm II

  • 期刊论文

  • 发表日期：

    2026年3月14日

查看全文
  • Marouene Chaieb,
  • Mohamed Amine Bouhachem,
  • Ayoub Rouin,
  • Ines Hilali Jaghdam

摘要

Home healthcare services play a vital role in ensuring that elderly, disabled, and chronically ill patients receive personalized, high-quality care in the comfort of their own homes. In this context, efficient and adaptive planning tools are essential to support complex decision-making under uncertainty. This study proposes an intelligent Decision Support System (DSS) designed to assist healthcare planners in the scheduling and routing of home healthcare tasks. The proposed DSS is based on a novel hybrid three-stage approach that integrates fuzzy logic, grey relational analysis, and a multi-objective optimization model solved using the Non-dominated Sorting Genetic Algorithm II (NSGA-II). In the first stage, potential healthcare facility locations are assessed through a knowledge-based framework combining fuzzy logic and grey relational analysis to handle uncertainty and prioritize criteria. The second stage formulates a robust multi-period, multi-objective model with three key goals: minimizing total operational costs, avoiding the misallocation of overqualified staff, and maximizing qualitative service attributes. In the final stage, NSGA-II is employed to solve the optimization problem and generate high-quality, Pareto-optimal scheduling and routing solutions. Experimental results on benchmark datasets confirm that the hybrid NSGA-II approach significantly outperforms baseline strategies (e.g., Simple Random Assignment, Greedy Matching) across multiple performance metrics, including travel time reduction, workload distribution, and patient satisfaction. For a representative case involving 25 patients and 7 nurses, the proposed approach successfully served 100% of patients while reducing total travel distance to as low as 562 units, achieving the highest objective function value of 23.13 among competing solutions. Monte Carlo simulations further demonstrate the robustness of the generated schedules, with up to 93.33% of patients served within their time windows (V0) and over 98% service feasibility when allowing minimal time-window violations (V1–V3) in several benchmark instances. Compared to baseline strategies and a state-of-the-art robust optimization model, the proposed fuzzy-NSGA-II approach achieves comparable or superior solution quality while reducing average computational time by up to 22% on large-scale instances. This highlights the system’s practical relevance and computational efficiency for real-time deployment in healthcare settings. The implementation code is publicly available at: https://github.com/DrMaroueneCHAIEB/HCSP_Fuzzy.