System and Method for a 1D-DICNN-GRU-based Deep Learning Feature Extraction Model in Non-intrusive Elevator Monitoring

The present invention discloses a non-intrusive elevator condition monitoring method based on a deep learning model, comprises the steps of: extracting multi-variant signals from non-intrusive current sensors (102); aggregating and converting the extracted multivariant signals into processable uniformed signal data (104); integrating the uniformed signal data segments into a deep learning
model (106); training the deep learning model with validation and testing (108); and monitoring the condition and detecting anomaly
of the elevator based on the deep learning model (110)