基于声发射技术的砂轮磨损状况在线检测

摘要

为了实现砂轮磨损状态在线检测，提高砂轮磨损状态检测的准确性，研究了法向磨削力与砂轮磨损的对应关系；利用小波分解系数统计法对声发射(AE)信号进行了分析；把法向磨削力和统计小波分解系数的特征作为识别砂轮磨损状态的参数指标，建立了基于神经网络(BP)的砂轮磨损状态识别模型。实验结果表明，该方法可以辨识出砂轮的磨损状态，并且具有较高精度。%In order to realize the on-line detection of the grinding wheel wear state and to improve its accuracy,this paper studies the relationship between normal grinding force and grinding wheel wear,analyzes acoustic emission signal by wavelet decomposition coefficient statistics method and establishes the classification model of grinding wheel wear state based on neural network by regarding the characteristics of the normal grinding force and the statistical wavelet decomposition coefficient as the parameter indexes of the recognition of grinding wheel wear state.The experimental results show that this method can assess the grinding wheel wear state with high accuracy.