广义径向基函数神经网络在热误差建模中的应用

发布时间：2018-04-11 20:30

本文选题：广义径向基函数 + 神经网络　；参考：《光学精密工程》2015年06期

【摘要】：针对现有的热误差建模方法建模效率低,模型预测精度不理想等问题,提出了广义径向基函数神经网络(RBF)建模方法并将其应用于数控机床热误差建模中。讨论了采用广义RBF神经网络进行热误差建模的原理及步骤。以数控导轨磨床主轴箱系统为例,布置了12个主轴热误差的关键温度测点,测得了2组独立的主轴箱系统热误差数据。将测得的数据分别用于建立主轴箱系统热误差广义RBF神经网络预报模型和验证模型的准确性。研究结果表明,热误差广义RBF神经网络模型具有预测精度高及泛化能力强的优点;与传统的RBF神经网络建模方法相比,提出的广义RBF神经网络建模方法建模效率更高,模型鲁棒性及预测性能更好,是一种可以用于数控机床热误差实时补偿的有效建模方法。
[Abstract]:Aiming at the problems of low modeling efficiency and poor prediction precision of existing thermal error modeling methods, a generalized radial basis function neural network (RBFN) modeling method is proposed and applied to the thermal error modeling of NC machine tools.The principle and steps of thermal error modeling based on generalized RBF neural network are discussed.Taking the spindle box system of NC guide grinding machine as an example, 12 key temperature measuring points of spindle thermal error are arranged, and two groups of independent spindle box system thermal error data are measured.The measured data are used to establish the generalized RBF neural network prediction model for the thermal error of spindle box system and verify the accuracy of the model.The results show that the generalized RBF neural network model with thermal error has the advantages of high prediction accuracy and strong generalization ability, and compared with the traditional RBF neural network modeling method, the proposed generalized RBF neural network modeling method is more efficient.The model is more robust and predictive. It is an effective modeling method which can be used for real-time compensation of thermal error of NC machine tools.
【作者单位】：湖南大学汽车车身先进设计制造国家重点实验室;
【基金】：国家自然科学基金资助项目(No.51175161;No.51475152) 国家科技重大专项资助项目(No.2011ZX04003-011)
【分类号】：TG596;TP183

【参考文献】