用神经网络分析法评价机械加工工艺参数

Neural Network Analysis on Evaluating the Mechanical Technological Parameters

神经网络提供了获取知识的一条新途径。在机械加工学科中，人们可以利用影响工件表面粗糙度的切削或磨削参数来建造神经网络模型，然后通过实际样本对神经网络进行自学习训练，使此模型变成切合车间不同设备的实用模型，从而可以有针对性地改变某些切削或磨削参数，以降低工件表面粗糙度和提高表面质量。本文论述了神经网络模型的建立。

Neural network supplies a new way to abtain knowledge.In machanical engineering,people can build neural network model with the parameters affecting grinding or cutting surface roughness.Trained with practical samples,the model can be changed into usable models which can satisfy variety of apparatus of workshop by adjusting approprietly some grinding or cutting parameters,so that the surface roughness could decrease and the surface quality improved.The establishment of neural network model and self training principle are introduced and a good way to decrease the surface roughness proposed in this article.