Abstract:

Abstract： Based on the acquisition of acoustic emission signal in the grinding process, two feature extraction methods for establishing grinding overload criterion were studied. The main indicators of sensitivity, stability and algorithm efficiency were proposed to evaluate the performance of different antioverload criterions. The effects of different signal processing methods on identifying the overload point between wheel and workpiece were analyzed with grinding experiments. The results show that the proposed feature extraction method of signaltonoise ratio which has high sensitivity, high algorithm efficiency and good stability, can predict the overload to some extent and can be used as the prior overload diagnosis criterion.
 

Key words: grinding anti-overload, feature extraction, acoustic emission, diagnosis