摘要
当前,铣床主轴加工产品容易受到热误差的影响,造成产品精度下降。对此,采用模糊神经网络模型预测铣床主轴热误差,并对预测结果进行比较和分析。建立神经网络径向基函数的表达式,给出了模糊推理系统和控制规则,创建了模糊RBF神经网络预测模型,对铣床主轴进行热误差验证。结果显示:铣床主轴采用RBF神经网络模型预测误差较大,其Y轴和Z轴输出最大误差分别为5.9μm和7.1μm;铣床主轴采用模糊RBF神经网络模型预测误差较小,其Y轴和Z轴输出最大误差分别为3.5μm和2.9μm。同时,模糊RBF神经网络模型预测误差跳动幅度较小。采用模糊RBF神经网络预测模型,可以补偿铣床运行时产生的热误差,提高铣床主轴加工精度。
At present,machine tools spindle processing products are vulnerable to thermal errors,resulting in product accuracy decrease.In this regard,the fuzzy neural network model is used to predict the thermal error of the machine tools spindle,and the prediction results are compared and analyzed.The expression of radial basis function of neural network is established,the fuzzy reasoning system and control rules are given,and the prediction model of fuzzy RBF neural network is established to verify the thermal error of machine tools spindle.The results show that the maximum error of Y and Z axes is 5.9μmand 7.1μm,respectively.The maximum error of Y and Z axes is 3.5μm and 2.9μm,respectively.At the same time,the prediction error jump of the fuzzy RBF neural network model is smaller.The prediction model based on fuzzy RBF neural network can compensate the thermal error caused by machine tools running and improve the machining accuracy of machine tools spindle.
作者
李耀贵
伍先明
LI Yao-gui;WU Xian-ming(Guangdong Polytechnic College.Zhaoqing Guangdong 526100,China)
出处
《井冈山大学学报（自然科学版）》
2019年第4期77-80,91,共5页
Journal of Jinggangshan University (Natural Science)
基金
2016年广东省人才培养模式创新实验区项目(项目编号:30)
关键词
径向基函数
神经网络
模糊推理
铣床
热误差
radial basis function
neural network
fuzzy reasoning
machine tools
thermal error
