摘要
以工件厚度、峰值电流、脉冲宽度、脉冲间隔为输入参数 ,以加工速度和表面粗糙度为输出参数 ,分别用神经网络技术与非线性回归理论建立了快走丝线切割加工的工艺模型。经与实验数据比较 ,得出两种模型均能较好地预测出给定条件下的加工速度和表面粗糙度 ,反映了该机床的加工工艺规律 。
Taking the workpiece thickness, peak value current, pulse width and pulse interval as input data, and the processing speed and surface roughness as output data, to establish technological models for fast Wire Cut processing respectively with neural network technology and nonlinear regression theory. After comparing the experimental data, it turned out that both of the two models could well forecast the processing speed and surface roughness under the given condition, which reflected the processing law of the machine. While the neural network model possesses a higher forecasting precision.
出处
《模具工业》
2000年第5期50-52,共3页
Die & Mould Industry
基金
上海市青年科技启明星计划资助项目!(96QF14006)
关键词
电火花线切割加工
建模
加工工艺
EDM Wire Cut machining, neural network, nonlinear regression, model_establishing
