Shanjing Zhang
Computer numerical control, local faults, intelligent detection, iterative filtering, feature extraction
Computerised numerical control machine tools are prone to rolling bearing failures under long-term high loads and complex working conditions. Thus, an intelligent detection method is proposed for local faults. Firstly, the fault characteristics of computerised numerical control machine tool bearings are analysed, an adaptive local iterative filter (ALIF) is introduced to extract and reconstruct the fault features, and then the frequency-weighted energy operator (FWEO) demodulates and analyses the reconstructed signal to extract the fault results. Considering the efficiency of fault diagnosis, composite multi-scale scatter entropy is introduced to express features, and finally, graph theory clustering is used to achieve fault-type diagnosis. In the analysis of bearing vibration signals in computerised numerical control, the proposed model improved the extraction accuracy of fault features by 32.6% compared to the time– frequency analysis combined with the FWEO model. In the local fault detection of computerised numerical control machine tools, the proposed model accurately identified four common bearing faults, with a fault classification accuracy of 98.65%, which was superior to the comparative model. At the same time, the comprehensive performance of the fault diagnosis model was compared, and the proposed model had detection accuracy rates higher than 0.950 in bearing fault detection of computerised numerical control, which was superior to the comparison model. The proposed local fault detection model for computerised numerical control machine tools has good application results in practical fault diagnosis scenarios, with higher detection accuracy and adaptability to complex working conditions. The research content will provide technical references for the application and improvement of intelligent detection technology in the field of industrial manufacturing.
Important Links:
Go Back
