Energy entropy-based vibration anomaly detection system of time series data using machine learning

摘要

The present invention relates to a vibration anomaly detection system based on energy entropy of time series data using machine learning, and more particularly, to a vibration anomaly detection system based on energy entropy of time series data using machine learning, which can perform preventive maintenance on an accurate target at accurate time more efficiently and effectively than an existing technique for detecting anomalies in a large amount of sensor data by analyzing energy entropy of abnormal vibration in a corresponding area by separating an abnormal area from sensor data generated more than 100 times per second in real time. The vibration anomaly detection system based on energy entropy of time series data using machine learning comprises: at least two vibration sensor parts; at least two heat sensor part; an entropy-based anomaly analysis server; and a monitoring terminal.