Scholarlink

This work focuses on developing an anomaly detection, prediction, and localization system using deep learning methods applied to the data collected from sensors built in IoT-enabled machines. The primary objective is to detect faults in real-time, estimate the machine’s behavior and health for a future time period, and precisely localize errors within the machine. In this paper we use Auto-encoder Neural Networks and LSTM Recurrent Networks and One Class SVM’s to actualise the previously mentioned objectives.

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

Autoencoder, LSTM, RNN, and SVM for Anomaly Detection, Prediction, and Localization in Industrial Systems

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Dalila Cherifi,
  • Mohammed Amine Bedri,
  • Said Boutaghane

摘要

This work focuses on developing an anomaly detection, prediction, and localization system using deep learning methods applied to the data collected from sensors built in IoT-enabled machines. The primary objective is to detect faults in real-time, estimate the machine’s behavior and health for a future time period, and precisely localize errors within the machine. In this paper we use Auto-encoder Neural Networks and LSTM Recurrent Networks and One Class SVM’s to actualise the previously mentioned objectives.