| 基于历史致灾流程挖掘的城市暴雨灾害链智能构建方法 |
| An intelligent construction method of urban rainstorm disaster chains based on historical disaster process mining |
| 摘要点击 37 全文点击 0 投稿时间：2025-10-10 修订日期：2026-06-05 |
| 查看/发表评论 下载PDF阅读器 |
| 中文关键词 城市暴雨；灾害链；流程挖掘；案例推理；致灾流程 |
| 英文关键词 Urban rainstorms; disaster chain; process mining; case-based reasoning; disaster process |
| 基金项目 国家自然科学基金项目（面上项目，重点项目，重大项目）, |
| 投稿方向 灾害风险；应急管理；系统工程 |
| 作者 | 单位 | 邮编 | | 刘昭阁* | 厦门大学 | 361005 | | 李向阳 | 哈尔滨工业大学 | | | 乔立民 | 北京北科互联城市治理技术研究院有限公司 | |
|
| 中文摘要 |
| 考虑灾害案例的知识学习，阐述了一类集成“流程-案例”的暴雨灾害链智能构建方法。该方法将案例推理融入传统流程挖掘，通过城市暴雨风险情景的相似度匹配筛选适配案例，再基于ProM-Boost算子实现案例提供关联规则的自适应嵌入。对10个城市的实验结果表明：所提方法能够利用多源开放历史数据，实现灾害链模型的可靠生成；同时，能够利用历史案例实现规则迁移，同时通过本体匹配消减语义歧义，生成的灾害链模型效果优于5类传统方法。所提方法有助于城市暴雨灾害链的精准可靠生成，更好地支撑灾害智慧管理。 |
| 英文摘要 |
| Considering the knowledge learning of disaster cases, a kind of intelligent construction method of rainstorm disaster chain integrating ‘process’ and ‘case’ is proposed. In this method, case-based reasoning is integrated into traditional process mining, adaptive cases are screened through similarity matching of urban rainstorm risk scenarios, and adaptive embedding of association rules provided by cases is realized based on ProM-Boost operator. The experimental results on 10 cities show that the proposed method can utilize multi-source open historical data to achieve reliable generation of disaster chain models; At the same time, the proposed method can use historical cases to achieve rule transfer, while reducing semantic ambiguity through ontology matching. The effect of the generated disaster chain model is better than that of the five traditional methods. The proposed method is conducive to the accurate and reliable generation of urban rainstorm disaster chain, and better supports the intelligent management of disasters. |
| 关闭 |
