Scholarlink

<p>In the initial phase of emergency response to geological disasters, decision-makers are frequently challenged by extreme environmental uncertainty and a critical shortage of quantitative monitoring data. To bridge this decision-making gap prior to the full establishment of monitoring networks, this study systematically selects emergency monitoring indicators by first characterizing geological disasters and identifying monitoring requirements through a literature review. An evaluation model is subsequently developed, comprising four primary factors— monitorability, timeliness, sensitivity, and feasibility— and nine secondary factors related to accuracy, stability, monitoring frequency, and sensitivity to catastrophic geological changes. The triangular fuzzy analytic hierarchy process (TriFAHP) is employed to address the inherent fuzziness in expert judgment, while a deep belief network (DBN) extracts expert decision features and generates an individual correction coefficient (β) to optimize weight allocation. The model yields a consistency ratio (CR) of 0.0329, confirming high reliability of the derived weights. Factor weights are ranked as follows: monitorability (0.3808) &gt; timeliness (0.3353) &gt; sensitivity (0.1874) &gt; feasibility (0.0966). Secondary factors, including accuracy (0.2083), monitoring frequency (0.2682), and indicator sensitivity to Disaster abrupt changes (0.0948), significantly influence the model’s early warning efficacy. The model is applied to an emergency monitoring scenario involving slope collapse, evaluating 16 commonly used indicators. Displacement, velocity, acceleration, and rainfall are identified as key monitoring indicators. These indicators are subsequently applied to the emergency monitoring of a slope collapse in Inner Mongolia, where they demonstrate effectiveness in supporting early warning decisions, thereby validating the model’s practicality and reliability. Further analysis reveals a decision-making tendency among experts to prioritize monitorability, while placing relatively less intrinsic value on emergency response speed. This study advances the theoretical framework of geological disaster management by shifting the focus from post-deployment data analysis to pre-deployment strategic configuration, offering a systematic and quantitative indexing tool to solve the initial monitoring configuration problem under data-scarce and highly uncertain emergency conditions.</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

Emergency monitoring index evaluation for collapse scenarios: A TriFAHP-DBN model addressing data-scarce conditions

  • 期刊论文

  • 发表日期：

    2026年7月29日

查看全文
  • Jikun Liu,
  • Chenghu Wang,
  • Guiyun Gao,
  • Haitao Ma,
  • Xiaolin Yang

摘要

In the initial phase of emergency response to geological disasters, decision-makers are frequently challenged by extreme environmental uncertainty and a critical shortage of quantitative monitoring data. To bridge this decision-making gap prior to the full establishment of monitoring networks, this study systematically selects emergency monitoring indicators by first characterizing geological disasters and identifying monitoring requirements through a literature review. An evaluation model is subsequently developed, comprising four primary factors— monitorability, timeliness, sensitivity, and feasibility— and nine secondary factors related to accuracy, stability, monitoring frequency, and sensitivity to catastrophic geological changes. The triangular fuzzy analytic hierarchy process (TriFAHP) is employed to address the inherent fuzziness in expert judgment, while a deep belief network (DBN) extracts expert decision features and generates an individual correction coefficient (β) to optimize weight allocation. The model yields a consistency ratio (CR) of 0.0329, confirming high reliability of the derived weights. Factor weights are ranked as follows: monitorability (0.3808) > timeliness (0.3353) > sensitivity (0.1874) > feasibility (0.0966). Secondary factors, including accuracy (0.2083), monitoring frequency (0.2682), and indicator sensitivity to Disaster abrupt changes (0.0948), significantly influence the model’s early warning efficacy. The model is applied to an emergency monitoring scenario involving slope collapse, evaluating 16 commonly used indicators. Displacement, velocity, acceleration, and rainfall are identified as key monitoring indicators. These indicators are subsequently applied to the emergency monitoring of a slope collapse in Inner Mongolia, where they demonstrate effectiveness in supporting early warning decisions, thereby validating the model’s practicality and reliability. Further analysis reveals a decision-making tendency among experts to prioritize monitorability, while placing relatively less intrinsic value on emergency response speed. This study advances the theoretical framework of geological disaster management by shifting the focus from post-deployment data analysis to pre-deployment strategic configuration, offering a systematic and quantitative indexing tool to solve the initial monitoring configuration problem under data-scarce and highly uncertain emergency conditions.