Scholarlink

<p>This study presented an integrated spatial optimisation framework for hazardous waste site selection in Markazi Province, Iran, utilising a hybrid approach that combines Geographic Information Systems (GIS), Analytic Network Process (ANP), and Genetic Algorithms (GA). Spatial and environmental datasets were analysed to evaluate critical criteria such as fault proximity, land use, geohydrology, and soil type. To ensure robustness and reliability, sensitivity analysis was conducted using Kappa statistics and Receiver Operating Characteristic (ROC) curves, which validated both model performance and the relative effect of selection criteria. In this framework, the ANP structure was optimised via GA to improve prioritisation accuracy. Comparative suitability maps were generated to provide a visual representation of multi-criteria assessments, facilitating simultaneous interpretation of environmental and spatial factors. The optimisation process demonstrated effective performance, as the best solution achieved a fitness value close to 1, indicating high-quality site selection. In addition, the average fitness value was greater than the final best fitness, which points again to low dispersion between solutions or potential convergence of the algorithm within a proximity of high-fitness solutions. Overall, the proposed approach demonstrates the complementary relationship between MCDM methods, evolutionary algorithms and statistical evaluation tools, and provides a robust, scalable, and sustainable method for managing hazardous waste strategically in environmentally sensitive areas.</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

Modeling and optimizing criteria in hazardous waste site selection with an integrated approach of GIS, ANP, and genetic algorithms

  • 期刊论文

  • 发表日期：

    2025年9月18日

查看全文
  • Razieh Nosrati Sakol,
  • Seyed Masoud Monavari,
  • Lobat Taghavi,
  • Shahram Bakpour,
  • Fariba Zamani

摘要

This study presented an integrated spatial optimisation framework for hazardous waste site selection in Markazi Province, Iran, utilising a hybrid approach that combines Geographic Information Systems (GIS), Analytic Network Process (ANP), and Genetic Algorithms (GA). Spatial and environmental datasets were analysed to evaluate critical criteria such as fault proximity, land use, geohydrology, and soil type. To ensure robustness and reliability, sensitivity analysis was conducted using Kappa statistics and Receiver Operating Characteristic (ROC) curves, which validated both model performance and the relative effect of selection criteria. In this framework, the ANP structure was optimised via GA to improve prioritisation accuracy. Comparative suitability maps were generated to provide a visual representation of multi-criteria assessments, facilitating simultaneous interpretation of environmental and spatial factors. The optimisation process demonstrated effective performance, as the best solution achieved a fitness value close to 1, indicating high-quality site selection. In addition, the average fitness value was greater than the final best fitness, which points again to low dispersion between solutions or potential convergence of the algorithm within a proximity of high-fitness solutions. Overall, the proposed approach demonstrates the complementary relationship between MCDM methods, evolutionary algorithms and statistical evaluation tools, and provides a robust, scalable, and sustainable method for managing hazardous waste strategically in environmentally sensitive areas.