Scholarlink

It’s paramount for the productivity of any company that industrial machinery is operating in an efficient manner and that the idle time is minimal. Anomaly detection plays a pivotal role in predictive maintenance – it focuses on abnormal behaviour patterns. The major challenge of RNNs is the computational complexity and the training of the models. On the other hand, Echo State Networks (ESNs), which belong to the class of Reservoir Computing, are generally more effective and efficient in modelling temporal relations. This paper investigates ESNs for the detection of industrial machinery anomalies using time-series sensor data. We describe the application of an ESN model on a standard industrial data set. Our results further indicate that ESN models are able to provide good results with less computation time and hence they can be used in real world industrial scenarios.

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

Anomaly Detection in Industrial Machines Using Echo State Networks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mann Patel,
  • Mithil Mistry,
  • Hasti Vakani,
  • Sachin Patel,
  • Tandel Vaibhav,
  • Rishi Patel,
  • Amit Nayak,
  • Rajesh Patel

摘要

It’s paramount for the productivity of any company that industrial machinery is operating in an efficient manner and that the idle time is minimal. Anomaly detection plays a pivotal role in predictive maintenance – it focuses on abnormal behaviour patterns. The major challenge of RNNs is the computational complexity and the training of the models. On the other hand, Echo State Networks (ESNs), which belong to the class of Reservoir Computing, are generally more effective and efficient in modelling temporal relations. This paper investigates ESNs for the detection of industrial machinery anomalies using time-series sensor data. We describe the application of an ESN model on a standard industrial data set. Our results further indicate that ESN models are able to provide good results with less computation time and hence they can be used in real world industrial scenarios.