基于NN与GA的曲面加工适应学习控制研究

Adaptive learning control of sculptured surface machining based on NN and GA

提出了适用于曲面加工条件的适应学习控制方法．在回避加工振动的前提下以实现高效率加工为目的，将遗传算法（ＧＡ）的优化机能与神经网络（ＮＮ）的学习机能相融合，建立了由开放式ＣＮＣ加工机床，多维传感器，实时控制器，学习模型及优化模型组成的曲面加工适应学习控制系统．切削实验结果表明所提出的加工条件适应控制系统可有效地抑制曲面切削振动，提高了切削加工效率．

The adaptive learning control for on-line optimization of sculptured surface machining process was presented. An adaptive learning neural network (NN) is used for the nonlinear relationship between process input and process state to optimize the cutting parameters by genetic algorithms (GA). It consists of a machine tool with open-architecture CNC, multi-sensors, real-time controller, learning model and optimizing model. The experimental results show that the proposed system can avoid chatter vibration and achieve high quality and high efficiency.