CO_2激光三维方位切割的人工神经网络分析

Analysis of Artificial Neural Network on 3D Orientation CO_2 Laser Cutting

三维激光切割在汽车、航空航天工业等领域得到越来越广泛的应用。在三维激光切割过程中 ,由于三维工件的复杂性 ,切割头与工件表面保持垂直的同时 ,不可避免地要进行一些方位切割。为了定量地描述激光头的方位对切割质量的影响 ,采用 10分制对切割质量进行量化 ,然后用人工神经网络的方法对结果进行分析。结果表明 ,用人工神经网络可以通过网络的训练和自学习 ,获得很多新的知识 。

D laser cutting has been widely used in automobile and aerospace industry. In 3D laser cutting it is inevitable for laser head to move up and down, keeping the laser head vertical to the surface of workpiece. The orientation of laser head influences cutting result. In order to quantitatively analyse the relationship between cutting result and cutting parameters, the 10 marks quality-evaluation system of cutting result and ANN(artificial neural network) have been setup. The new knowledge can be introduced by ANN. These new knowledges have been used to guide actual cutting to realize intelligent laser processing.