基于隐马尔科夫模型的刀具磨损连续监测

Continuous Tool Wear Monitoring Based on Hidden Markov Model

与离散的刀具磨损状态的分类识别相比,人们更希望得到连续的刀具磨损值,从而为最终的控制过程提供更准确的信息。为了监测连续的刀具磨损值,采用易于采集的振动和声发射信号作为监测信号,提取信号的时域特征、频域特征和时频域特征,从中筛选出对刀具磨损敏感的特征,并采用隐马尔科夫模型建模,最后通过概率计算得到连续的磨损值。通过比较采用切削力、加速度和声发射信号的监测模型和仅采用加速度和声发射两种信号的监测模型,发现在没有切削力信号的情况下,仍能够准确地预测刀具磨损值。

Instead of distinguishing tool wear states into various discrete classes,we would like to predict the continuous tool wear and provide more accurate information for the final control process. To monitor the tool wear continuously,firstly,extract the time domain feature,frequency domain feature and time-frequency domain feature from the signal collected by accelerometers and acoustic emission sensors which are easy in use,then select the features which are sensitive to tool wear,finally,train the model based on Hidden Markov model and predict the continuous tool wear by a probabilistic approach. By comparing the model using force,vibration and acoustic emission signals with the model only using vibration and acoustic emission,found the model without force signal can also predict the tool wear accurately.