基于RBF-FA的TC4叶片抛光材料去除率预测研究  

Research on Prediction of Material Rremoval Rate in TC4 Blade Polishing Based on RBF-FA

作　　者：陈振 李锋[1] 田国良 普亚松 CHEN Zhen;LI Feng;TIAN Guo-liang;无(Xi'an Aeronautical University,Xi'an 710089;Northwestern Polytechnical University,Xi'an 710072)

机构地区：[1]西安航空学院,西安710089 [2]西北工业大学,西安710072

出　　处：《航空精密制造技术》2024年第4期9-12,共4页Aviation Precision Manufacturing Technology

基　　金：陕西省教育厅科研计划资助项目(23JK0495);陕西省自然科学基础研究计划资助项目(2022JQ-503);陕西省“十四五”教育科学规划课题(SGH23Q0324);西安航空学院博士科研启动金(2021KY0216)。

摘　　要：为了优化钛合金TC4叶片抛光过程中的材料去除率,进而提高工艺效率和延长设备使用寿命,设计了工艺参数与材料去除率之间的抛光试验。在获得试验数据的基础上,构建了基于径向基函数(RadialBasis Function,RBF)的材料去除率预测模型,并引入萤火虫算法(FireflyAlgorithm,FA)进行参数优化。经过深入训练和验证,RBF-FA预测模型展现出了高精度的材料去除率预测能力。实验结果表明,RBF-FA预测模型准确度高,其中训练组和检验组的平均误差分别为4.97%和3.59%,验证了该模型对于TC4叶片抛光过程中材料去除率预测的有效性。文中提出的方法可为TC4合金叶片抛光工艺参数的优化提供可靠依据。In order to optimize the material removal rate during the polishing process of titanium alloy TC4 blades,thereby improving process efficiency and extending equipment service life,an orthogonal experiment between polishing process parameters and material removal rate was designed.On the basis of obtaining orthogonal test data,a material removal rate prediction model based on radial basis function(RBF)was constructed,and the firefly algorithm(FA)was introduced for parameter optimization.The RBF-FA prediction model was in-depth trained and verified based on experimental data,and achieved high-precision material removal rate prediction capabilities.The experimental results show that the RBF-FA prediction model has high accuracy,with the average errors of the training group and test group being 8.04%and 5.91%respectively,confirming the effectiveness of the model in predicting the material removal rate in the TC4 polishing process.The method proposed can provide a basis for the optimization of TC4 alloy blade polishing process parameters.

关 键 词：RBF神经网络 TC4 叶片抛光 材料去除率 

分 类 号：TH122[机械工程—机械设计及理论]