径向基神经网络在三维激光切割中的应用  

Application of RBF Neural Network in Three Dimensional Laser Cutting

作　　者：吴艳华[1] 张永强[1] 陈武柱[1] 张旭东[1] 阎启[2] 

机构地区：[1]清华大学,北京100084 [2]宝山钢铁股份有限公司,上海201900

出　　处：《中国机械工程》2006年第12期1234-1237,共4页China Mechanical Engineering

摘　　要：分析了RBF神经网络的特点,并研究了它在三维激光切割中的应用,利用该神经网络及其变形可以预测切缝的宽度,也可以根据输入参数对试样的挂渣与否进行判定。比较了三维激光切割中切割倾角和切割方向对切割质量的影响,从而实现切割质量的预测和优化工艺参数的选取以指导实际的切割过程。The application of radial basis function （RBF） neural network in three dimensional laser cutting was investigated. RBF neural network is competent for both the prediction of the kerr width and the judgment of the dross condition. Based on the RBF neural network model, the effect of laser beam attitude, including cutting obliquity and cutting direction, on the quality of three dimensional laser cutting was also studied. The method can be used to predict the cutting quality and optimize processing parameters in three-dimensional laser cutting process.

关 键 词：激光 三维切割 径向基函数 RBF神经网络 

分 类 号：TG485[金属学及工艺—焊接]