基于BP网络的注塑成型模具硬态高速铣削力研究

Milling force of plastic mold parts in high speed hard milling based on BP neural network

用AlTiN涂层硬质合金立铣刀对4Cr5MoSiV1钢注塑成型模具进行硬态高速铣削,研究切削加工参数对切削力的影响;通过多因素法进行正交试验,利用改进的BP神经网络建立切削力的神经网络模型并对工艺参数进行优化,将网络预测结果与现场加工实践数据进行对比。研究结果表明:人工神经网络能准确地预测铣削力,模型具有较强的泛化能力和自适应能力;在高转速、小切深、合适的进给速度以及微量切削液状态下铣削力较小,为优化模具硬态铣削的切削参数并对其实际生产应用提供了较好的依据。

Taking the hard milling of the plastic mold parts（cavity） as an example,using a hardened 4Cr5MoSiV1 steel and AlTiN coated carbide end-milling tool,the empirical model of the cutting force model was carried out with the improved BP neural network and a multi-factorial orthogonal test,the accuracy of the cutting force network prediction was examined by field-processing practice（practice of hard milling of the mold parts）.The results show that the artificial neural network can predict the size of milling force well and has strong generalization ability and adaptive capacity;the smaller milling forces can be attained by using higher spindle speed,appropriate feed rate,lower axial and radial depth of milling parameters as well as minimum quantity of lubricant（MQL）,which provide a better basis for cutting parameters optimization of hard milling of the plastic mold parts and its manufacture.