基于机器学习的镍基单晶高温合金蠕变寿命预测模型研究

Research on Prediction Model of Creep Life of Nickel-based Single Crystal Superalloys Based on Machine Learning

构建合适的镍基单晶合金蠕变寿命预测模型,对于我国航空发动机叶片设计、强度分析和寿命预测具有重要意义。采用多项式回归、最近邻回归、支持向量机回归、决策树回归四种机器学习算法,建立镍基单晶高温合金蠕变寿命与合金成分、微观组织和蠕变工艺参数的关系模型,为镍基单晶高温合金的蠕变性能调控提供了新方法。基于蠕变寿命预测模型,系统地比较了四种算法和特征选择对模型性能的影响。结果表明,支持向量机回归模型的预测结果最优,相关性较高的四个特征依次为γ′固溶温度、Ta、W、Re。研究结果可为获得更有效的镍基单晶高温合金蠕变性能预测方法提供参考。

Building the suitable creep life prediction model of nickel-based single crystal superalloys is of great significance for the design,strength analysis,and life prediction of aircraft engine blades in China.In this paper,four machine learning algorithms are used,including polynomial regression,nearest neighbor regression,support vector machine regression and decision tree regression.The relationship models between the creep life of nickel-based single crystal superallays and the alloy composition,microstructure and creep process parameters are established,which provide a new method for the control of creep properties of nickel-based single-crystal superalloys.Based on the creep life prediction model,the effects of algorithm selection and feature selection on the comprehensive performance of the model are systematically compared.The results show that the SVR regression model has the best prediction,and the four features with relatively high correlations are in order as follows:γ′solution temperature,Ta,W,and Re.The research results can provide a reference for the prediction method of nickel-based single crystal superalloys with more effective creep properties.