Tool Condition Monitoring of Single-Point Dresser Using Acoustic Emission and Neural Networks Models
Tool Condition Monitoring of Single-Point Dresser Using Acoustic Emission and Neural Networks Models
复制标题
DOI：
10.1109/tim.2013.2281576
复制
发表时间：
2014-03-01
影响因子：
5.6
通讯作者：
Bianchi, Eduardo Carlos
中科院分区：
文献类型：
--
作者：
Martins, Cesar H. R.;Aguiar, Paulo R.;Bianchi, Eduardo Carlos
对梳妆台磨损的识别和在线监测是保证砂轮表面质量和提高磨削效率的必要条件。然而，刀具磨损是一个复杂的现象，发生在几个和不同的方式在切削过程中，并有一个缺乏的分析模型，可以代表刀具的条件。另一方面，神经网络被认为是解决缺乏分析或经验模型的一种很好的方法。提出了一种利用声发射信号表征梳妆台磨损状态的方法。为了实现这一点，提出了一些神经网络模型。最初，对原始AE信号的频率内容进行了研究，以确定与信号和梳妆台磨损相关的特征。从声发射谱中选取了9个频段，利用功率统计量的均方根和比值得到了信号的特征。两个频带的组合进行了评估，作为八个神经网络模型的输入，它们的分类能力进行了比较。可以验证的是，28-33和42-50 kHz的频带的组合最好地表征梳妆台磨损状况。一些模型产生了非常好的结果，因此可以确保地面部分符合项目规范。
Identification and online monitoring of the dresser wear are necessary to guarantee a desired wheel surface and improve the effectiveness of grinding process to a satisfactory level. However, tool wear is a complex phenomenon occurring in several and different ways in cutting processes, and there is a lack of analytical models that can represent the tool condition. On the other hand, neural networks are considered as a good approach to resolve the absence of an analytical or empirical model. This paper describes a method to characterize the dresser wear condition from acoustic emission (AE) signal. To achieve this, some neural network models are proposed. Initially, a study on the frequency content of the raw AE signal was carried out to determine features that correlate the signal and dresser wear. The features of the signal were obtained from the root mean square and ratio of power statistics at nine frequency bands selected from AE spectra. Combinations of two frequency bands were evaluated as inputs to eight neural networks models, which have been compared with their classification ability. It could be verified that the combination of the frequency bands of 28-33 and 42-50 kHz best characterized the dresser wear condition. Some of the models produced very good results and can therefore ensure the ground part will be within project specifications.