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 ,用于指导实际切割 Three-dimensional laser cutting in the automotive, aerospace industry and other fields have been more widely used. In the 3D laser cutting process, due to the complexity of the 3D workpiece, the cutting head and the workpiece surface remain vertical at the same time, it is inevitable to make some azimuth cutting. In order to quantitatively describe the influence of laser head orientation on the cutting quality, the cutting quality is quantified by 10-point scale, and then the results are analyzed by the method of artificial neural network. The results show that using artificial neural network training and self-learning through the network, get a lot of new knowledge, used to guide the actual cutting