论文部分内容阅读
提出了利用人工神经网络（ＡＮＮ）及模糊识别理论融合多监控参数进行刀具状态识别的方法．该方法首先对各监控参数按刀具不同状态的敏感性进行分组，并利用多个ＡＮＮ子网络建立各组参数与刀具状态的模糊隶属度关系，然后利用模糊决策法对各ＡＮＮ子网络确定的刀具状态模糊隶属度进行综合评判并按最大隶属度判定刀具状态．该方法不仅具有ＡＮＮ的并行运算特点，而且具有模糊综合评判的容错性，从而提高状态识别的实时性和正确率．结合功率信号的多个特征对大量实验数据的测试表明，该方法可将ＡＮＮ的识别正确率从平均８８％提高到９５％．
A method of tool state recognition based on artificial neural network (ANN) and fuzzy recognition theory is proposed. Firstly, the sensitivity of each monitoring parameter to the different states of the tool is grouped according to the sensitivity of the tool. Then, the fuzzy membership degree between each parameter and the tool state is established by using a plurality of ANN sub-networks. Then, the fuzzy decision- State fuzzy membership degree comprehensive evaluation and judge the state of the tool according to the maximum membership degree. This method not only has the characteristics of ANN parallel computing, but also has the fault-tolerance of fuzzy comprehensive evaluation, so as to improve the real-time performance and accuracy of status recognition. Combined with many features of power signal, a large number of experimental data tests show that this method can improve ANN recognition accuracy from an average of 88% to 95%.