| 郭凯* **,黎建辉* ***,师亮*.地震数据驱动的城市出行智能评估模型[J].高技术通讯(中文),2025,35(12):1300~1310 |
| 地震数据驱动的城市出行智能评估模型 |
| Intelligent urban travel evaluation model driven by seismic data |
| |
| DOI：10. 3772 / j. issn. 1002-0470. 2025. 12. 004 |
| 中文关键词: 城市出行强度； 梯度提升树； 地震数据； 长短期记忆 |
| 英文关键词: urban travel intensity， gradient boosting tree， seismic data， long short-term memory |
| 基金项目: |
| 作者 | 单位 | | 郭凯* ** | （*中国科学院计算机网络信息中心北京 100190）
（**中国科学院大学北京 100049）
（***南京大学空间地球科学研究院苏州 215163） | | 黎建辉* *** | | | 师亮* | |
|
| 摘要点击次数: 36 |
| 全文下载次数: 37 |
| 中文摘要: |
| 城市出行强度是衡量人类活动的重要指标，为预测趋势并理解人类活动对环境和社会的影响提供了关键参考。然而，受限于出行数据的全面性、准确性和隐私保护，量化城市出行强度面临巨大挑战。本研究提出了一种创新方法，利用连续地震数据来量化城市出行强度，从而有效解决了大规模数据收集和隐私问题。我们构建了结合梯度提升树和长短期记忆（long short-term memory,LSTM）网络的机器学习框架，以应对地震数据的复杂性和多样性，提取与出行强度相关的特征。研究结果表明，模型在湖北省、河北省和山西省的训练精度表现优异，并在其他8个城市中展现出良好的泛化性能。模型预测的出行强度变化准确反映了COVID-19疫情期间不同阶段的出行模式变化。本研究提供了一种客观、低成本且保护隐私的城市出行强度评估方法，为可持续的城市规划和政策制定提供了科学依据。 |
| 英文摘要: |
| Urban travel intensity serves as a critical indicator of human activity, offering valuable insights for forecasting trends and understanding the impact of anthropogenic activities on both the environment and society. However, challenges in collecting comprehensive, accurate travel data and ensuring privacy have made it difficult to accurately estimate urban travel intensity. In this study, an innovative approach by leveraging continuous seismic data is proposed to quantify urban travel intensity, thus effectively addressing issues of large-scale data collection and privacy concerns. We develop a machine learning framework that integrates gradient boosting trees and long short-term memory (LSTM) networks to extract features related to travel intensity, overcoming the complexity and variability of seismic data. The experimental results demonstrate high training accuracy in Hubei, Hebei, and Shanxi provinces, and strong generalization performance across eight cities in different regions. The model’s predictions successfully capture the significant changes in travel intensity patterns during various stages of the COVID-19 pandemic. This study offers an objective, low-cost, and privacy-preserving method for assessing urban travel intensity using seismic data, providing scientific support for sustainable urban planning and policy development. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
