Open Cut » Maintenance & Equipment
Conveyor belts play a pivotal role in the operation, storage, and transportation of bulk materials in the mining industry. Wear and tear of mining conveyor belts as well as monitoring their integrity in real-time mandate a critical concern for safety, security, and maintenance. Zipper failure and belt rips cause many injuries and fatalities in the coal mines. They also result in a significant maintenance and financial issues. Hence, the integrity testing of conveyor belts has become absolutely critical. Crack monitoring of coal mining conveyor belt using radio frequency identification (RFID) technology can provide optimisation and improvement in productivity.
The project aims to develop RFID based sensing systems with integration of machine learning (ML) techniques to predict and monitor the health condition of conveyor belts. The development of the new monitoring system based on RFID technology involves a novel RFID sensor based crack detection, a new monitoring system based on signal processing techniques, and an efficient machine learning algorithms based failure prediction.
Design and analysis of a novel passive Ultra High Frequency (UHF) RFID based sensor for crack detection in coal mining conveyor belts was presented and a dielectric characterisation of the belt was performed which helps to obtain the material parameters of the belt and test the performance of the proposed sensor. Both simulated and experimental studies cater to a robust sensor performance for detecting cracks at different crack widths, orientations, and locations in the conveyor belt environment.
The project proposes a robust and low-cost passive UHF based RFID crack sensing system which allows the processing and detection of the presence of crack from the sensors' data in real-time. This includes developing of a graphical user interface (GUI) and integrating it into a UHF RFID commercial reader to collect and process the received signal strength indicator (RSSI) from the sensors. This allows the site operator to make a decision in preventing damage and possible failure in the conveyor transport system. It also minimises the risk of delay or interruption in productivity. Experimental results demonstrate that the proposed monitoring system can offer highly accurate early detection of a crack having a width of as little as 0.5mm at the moving conveyor belt, which is in a high motion of 4 m/s.
The developed sensor based on crack detection and the new monitoring system are integrated with a multi-class ML algorithm for conveyor belt health monitoring. The ML model is tested with different input features and training algorithms, and their performances are compared and analysed to identify the most superior input feature and training algorithm. The simulation results show that the proposed detection system based on ML modelling could detect cracks with 100% accuracy. The proposed system can also identify crack orientations with an accuracy of 83.9%, and has a significant identification rate of 84.4% accuracy for detecting crack width of as narrow as 0.5mm. Moreover, the model can predict the region of the crack with an accuracy of 95.5%. Overall, the results show that the proposed model is very robust and can perform SHM of conveyor belts with high accuracy for a range of parameters and classification scenarios.