Abstract
The present invention teaches a system (10) to detect anomalies in a steel wire rope (SWR) of an elevator based on changes in raw magnetic flux leakage (MFL) data, the system comprises: at least one MFL signals sensor placed on a targeted SWR of the elevator; a pre-processing module (14) configured to process the MFL data; at least one pre-trained deep learning neural network model (16) configured to train the processed data and generate outputs (18); and a feedback module configured to record and feed the outputs back into the at least one pre-trained deep learning neural network model (16) for parameters updates and re-training and its method
thereof.
thereof.
Original language | English |
---|---|
Patent number | HK30089194 |
Filing date | 8/06/23 |
Publication status | Published - 2024 |