Scholarlink

<p>Efficient waste management, especially in cities, is becoming more challenging due to the growing complexity of waste systems. The siting of hazardous healthcare waste treatment facilities is a complex decision-making problem that involves environmental, social, and economic factors under uncertainty. This study presents a novel integrated framework for locating hazardous healthcare waste treatment facilities that combines Multi-Criteria Decision Making (MCDM), Machine Learning (ML), and robust optimization. The framework consists of three phases: (i) sustainability-based site evaluation using BWM and MARCOS, (ii) an enhanced constrained K-means clustering method for creating geographically balanced and capacity-coherent clusters, and (iii) a bi-objective robust optimization model that maximizes spatial dispersion and sustainability. A case study in Tehran shows that, compared with the baseline (original clustering) under identical model settings, the enhanced clustering achieves a 32% relative improvement in the distance-based spatial dispersion objective (km) and a 3% relative improvement in the unitless MARCOS-derived sustainability objective. These gains remain consistent across the tested sensitivity and robustness settings.This approach provides a more effective solution for siting facilities that balances operational and environmental concerns under uncertainty, offering a practical decision support tool for urban waste management.</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

An Enhanced K-Means Clustering-Based Decision Framework for Multi-Objective Robust Optimization of Obnoxious Facility Location

  • 期刊论文

  • 发表日期：

    2026年3月7日

查看全文
  • Abbas Foroozanfar,
  • Amirhossein Amou Jafari,
  • Mohammad Ali Hassanabadi,
  • Masoud Rabbani

摘要

Efficient waste management, especially in cities, is becoming more challenging due to the growing complexity of waste systems. The siting of hazardous healthcare waste treatment facilities is a complex decision-making problem that involves environmental, social, and economic factors under uncertainty. This study presents a novel integrated framework for locating hazardous healthcare waste treatment facilities that combines Multi-Criteria Decision Making (MCDM), Machine Learning (ML), and robust optimization. The framework consists of three phases: (i) sustainability-based site evaluation using BWM and MARCOS, (ii) an enhanced constrained K-means clustering method for creating geographically balanced and capacity-coherent clusters, and (iii) a bi-objective robust optimization model that maximizes spatial dispersion and sustainability. A case study in Tehran shows that, compared with the baseline (original clustering) under identical model settings, the enhanced clustering achieves a 32% relative improvement in the distance-based spatial dispersion objective (km) and a 3% relative improvement in the unitless MARCOS-derived sustainability objective. These gains remain consistent across the tested sensitivity and robustness settings.This approach provides a more effective solution for siting facilities that balances operational and environmental concerns under uncertainty, offering a practical decision support tool for urban waste management.