Scholarlink

This paper applies deep learning techniques to the signals acquired during a long-term dynamic monitoring campaign conducted on the Matilde donjon, a fortified keep belonging to the Old Fortress in the Medicean Port of Livorno, Italy. The time series collected during the dynamic monitoring complemented with the environmental parameters (temperature, wind speed) were used to train a deep learning neural network and forecast the dynamical behaviour of the tower. Although the signals are sparse and noisy, the algorithm can learn the main features of the tower’s dynamic response and detect anomalies and events occurring in the surrounding environment.

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

Using Artificial Intelligence for the Dynamic Monitoring of an Old Tower in the Medicean Port of Livorno (Italy)

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Gianni Bartoli,
  • Michele Betti,
  • Maria Girardi,
  • Gianmarco Gurioli,
  • Nicola Messina,
  • Cristina Padovani,
  • Daniele Pellegrini,
  • Giacomo Zini

摘要

This paper applies deep learning techniques to the signals acquired during a long-term dynamic monitoring campaign conducted on the Matilde donjon, a fortified keep belonging to the Old Fortress in the Medicean Port of Livorno, Italy. The time series collected during the dynamic monitoring complemented with the environmental parameters (temperature, wind speed) were used to train a deep learning neural network and forecast the dynamical behaviour of the tower. Although the signals are sparse and noisy, the algorithm can learn the main features of the tower’s dynamic response and detect anomalies and events occurring in the surrounding environment.