摘要
针对车床加工具有多因素非线性,利用神经网络提出解决数控机床系统的温度误差补偿控制问题的办法。方法是在现有的数控系统内部嵌入一个神经网络的小型系统,利用主元分析进行温度误差补偿数据压缩及特征提取,通过应用粗集理论对温度误差属性进行约简,用一个BP(back propagation)神经网络进行补偿控制,在网络的训练过程中利用蚁群算法对BP网络参数进行全局搜索,实现网络快速收敛到全局最优值。试验仿真表明,此方法可有效提高系统补偿精度和实时性。
A new method using neural networks to solve numerical control machine tool system for temperature compensation control problem was proposed in this paper. The study primarily focused on the development of integrated intelligent computation approach to improve compensation precision on machine tools. Data compression and feature extraction was re- alized by way of application of RST（ rough set theory） and principal component analysis. A dynamic BP（ back propagation） neural network embeded in a digital control system of a machine tool was presented which takes advantage of ant colony al- gorithm on BP network parameters to do the global search so that BP network convergence to get a global optimum. The re- suits obtained shows that this approach can effectively improve compensation precision and real time of error compensation on machine tools.
出处
《重庆邮电大学学报（自然科学版）》
北大核心
2010年第5期656-659,共4页
Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)
基金
国家高技术研究发展计划("863"计划)第四批课题(2008AA11A134)
四川省教育厅基础应用研究课题(2009ZX002)~~
关键词
BP网络
粗集理论
主元分析
蚁群算法
BP network
rough set theory
principal component analysis
ant colony algorithm
