灰色模型电火花磨削蜂窝环的试验研究

Research of Gray GM Model EDG Honeycomb

以电火花磨削蜂窝环(GH3536)为试验对象,设计出L16(44)正交表。采用灰色关联分析方法研究峰值电流、峰值电压、脉冲宽度、脉冲间隔四个主要电加工参数对加工速度的影响关系,基于GM(1,N)多变量灰色模型及人工神经网络原理建立电火花磨削蜂窝环的加工速度预测模型,并在MATLAB环境下进行仿真,得出模型在样本参数范围内的绝对均值误差为0.53%,在样本参数外的绝对均值误差为6.74%。结果表明该模型能映射出电火花磨削蜂窝环的工艺规律并能成功预测出其加工速度,为电火花磨削蜂窝环工艺参数的选择提供了参考和依据,具有实用意义。

EDG （electircal discharge grinding）honeycomb （GH3536） was regarded as the research object and the L16（4^4） orthogonal table was designed. The influence of four main processing parameters, peak electric current, peak voltage, pulse width and pulse intermission on processing speed was studied by gray connection analysis. A processing speed prediction model of EDG honeycomb （GH3536） was established based on GM （1,N） multivariate gray model and artificial neural networks principle, and then was simulated by Matlab software. Therefore, the prediction model with the overage absolute error 0.53% was obtained within the scope of sample parameters, while outside the scope was 6.74%. The results indicated that the model can reflect the technological law of EDG honeycomb and successfully predict its processing speed. The paper provides references and basis for selecting processing parameters of EDG honeycomb and has practical significance.