神经网络模拟螺旋槽形加工的研究与实现

Research and Practice of Helical Groove Machining Simulation with Neural Network

根据铣刀切削刃绕自身轴线旋转的同时工件作螺旋运动形成包络面的原理,应用BP神经网络建立铣刀刃形与螺旋槽形之间的非线性关系模型,根据铣刀轴向截形模拟仿真出工件螺旋槽端面截形.仿真结果表明,采用神经网络的非线性逼近特性,能精确、快速地仿真出螺旋槽形;此模型能够验证不同螺旋面参数下螺旋槽形的变化情况,以便对螺旋槽端面廓形的形状、精度和加工情况进行分析评估,从而对铣刀刃形的正确性给出一个生产先期的客观评价,为铣刀刃形的修正和最大刃形重磨量提供理论依据.

An envelope surface will be machined when the milling cutter rotates around its axis and the workpiece moves helically, according to this principle and based on a nonlinear model of relationship between the milling cutter＇ s profile and the workpiece＇ s helical groove which was established with BP neural network, the workpiece helical groove＇ s profile was successfully emulated in accordance with the milling cutter＇ s cutting edge. The simulation result shows that the helical groove＇ s profile can be emulated accurately and conveniently with neural network＇ s nonlinear simulation. The change of helical groove＇ s profile with various helicoid parameters can be verified by this method, so as to analyze the helical groove＇ s profile and precision, and to verify whether the milling cutter satisfies the requirements before machining, and finally, the theory evidences for milling cutter profile＇ s modification and the most grinding capacity of cutting edge abrasion will be supplied.