数控内齿强力珩齿珩削力的预测方法研究

Research on prediction method of honing force of CNC internal gear power honing

为获得内齿强力珩齿加工中的珩削力数值,文章引入反向传播(back propagation,BP)神经网络对其进行预测。以主轴转速、珩轮径向进给量、珩轮轴向进给速度为设计因子开展珩削试验,采用机床内部Kistler测力仪测量获取珩削力;采集试验样本后利用BP神经网络进行训练,得到珩削力的BP神经网络模型,并与相同试验样本下的指数型模型进行了对比。结果表明,BP神经网络模型可以更精准地预测内齿强力珩齿加工中的珩削力,BP神经网络普遍误差均不超过5%,而指数型预测模型误差最大达到了18.94%。但指数型模型在预测珩削力上具有操作简便性的特点,因此2种模型对内啮合强力珩齿珩削力的研究都具有一定的参考价值。

In order to obtain the honing force in the process of internal gear power honing,back propagation(BP)neural network is introduced to predict the honing force.The honing experiments were carried out with the spindle speed,the honing wheel radial feed rate and the honing wheel axial feed rate as the design factors,and the honing force was measured with the Kistler force measuring instrument inside the machine tool;the experimental samples were trained with BP neural network to obtain the BP neural network model of honing force,and compared with the exponential model under the same experimental samples.The results show that BP neural network model can more accurately predict the honing force in the process of internal gear power honing,the general error of BP neural network is less than 5%,the maximum error of exponential prediction formula is 18.94%,but the exponential model has the characteristic of simple operation in predicting honing force.Therefore,the two models have certain reference value for the study of honing force of internal gear power honing.