基于BP神经网络的超声磨料混粉电火花加工模型  

A Ultrasonic Abrasive Mixed Powder EDM model Based OnBP Artificial Nerve Network

作　　者：张云鹏[1] 任中根[1] 迟恩田[1] 张安洲[1] 

机构地区：[1]西北工业大学,陕西西安710072

出　　处：《电子工艺技术》2003年第3期126-129,共4页Electronics Process Technology

摘　　要：由于超声磨料混粉电火花加工是一个复杂的过程,难于建立精确的加工理论模型。应用BP神经网络实现了超声磨料混粉电火花加工的模型建立,并通过该模型对超声磨料混粉电火花加工试验研究的零件表面粗糙度和加工速度进行了预测,预测结果与实际试验数据有较好的一致性,可用于指导实际超声磨料混粉电火花加工。On account of the ultrasonic abrasive mixed powder EDM is a complex process,it is difficult to build a precise mathematical model of the processing.Build the model for ultrasonic abrasive mixed powder EDM based on the BP Artificial Nerve Network (ANN).Finally predicte the roughness of machined surface and processing speed by the model,the results of prediction are in accord with the test results.The model may be used as the reference model for the theory study and practical application of the ultrasonic abrasive mixed powder EDM.

关 键 词：超声磨料混粉电火花加工 模型 BP神经网络 建模 表面粗糙度 

分 类 号：TG661[金属学及工艺—金属切削加工及机床]