CAJ | 학술논문

针对传统切削经验公式无法精确预测微细铣削零件表面粗糙度的问题,提出了一种基于人工神经网络的表面粗糙度预报方法. 利用试验选择不同切削参数组合进行铣削试验,将试验结果分为两部分,一部分数据用作BP神经网络的训练样本并最终建立预报模型,另一部分用作测试样本,与相同切削参数条件下的神经网络预测值进行对比. 从而证明BP神经网络对于微细铣削表面粗糙度值具有很高的预测精度.
침대전통절삭경험공식무법정학예측미세선삭령건표면조조도적문제,제출료일충기우인공신경망락적표면조조도예보방법. 이용시험선택불동절삭삼수조합진행선삭시험,장시험결과분위량부분,일부분수거용작BP신경망락적훈련양본병최종건립예보모형,령일부분용작측시양본,여상동절삭삼수조건하적신경망락예측치진행대비. 종이증명BP신경망락대우미세선삭표면조조도치구유흔고적예측정도.
In view of the problem that the traditional cutting experience formula can not accurately predict the micro-milling roughness of parts, a prediction method of roughness which is based on ANN is proposed. The different cutting parameters of milling combination are selected while testing. The test results are divided into two parts:one part of the data is used as training samples of BP neural network to set up the forecast model eventually;the other part is used as the test sample which is compared neural network predictive value with the same cutting parameters to prove that the BP neural net-work for micro milling roughness value prediction has a high precision.