合金疲劳强度随温度变化关系的灰色预测模型研究  

Research on Gray Prediction Model for Fatigue Strength Changing with Alloy Temperature

作　　者：金鼎沸 吴永军[1] 罗意平[1] 宋永锋[1] 陈江[1] 

机构地区：[1]中南大学交通运输工程学院,湖南长沙410075

出　　处：《热加工工艺》2015年第20期77-79,共3页Hot Working Technology

摘　　要：合金疲劳强度随温度变化关系是金属加工过程中需要考虑的重要因素,需要合理选择方法对其进行预测。根据三次样条插值一致收敛性、精度较高等特点,提出了一种基于三次样条插值的非等间距的GM(1,1)模型灰色预测方法。该方法通过三次样条插值法处理原始非等间距采样数据,将其转变为等间距序列,继而用传统等间距GM(1,1)模型进行预测。在预测试验结果中,均方差比值为0.0014,小误差概率为1,精度等级为一级,得到了良好的预测效果。The alloy tem perature dependence of fatigue strength is an important factor that should be taken into consideration in the process of metal working. It should be forecasted in a proper way. A gray prediction method of non-equidistance GM（1,1） model which is based on cubic spline interpolation was put forward according to the cubic spline interpolation of uniform convergence and high accuracy. Using the method of cubic spline interpolation, the manipulation original sampled data was treated, which make it become equal interval sequence, thus the classic equal interval gray prediction model GM（1,1） can be used. In the test results, the ratio of mean-square deviation is 0.0014, the minor error probability is 1, and accuracy grade is 1, the good predication effect can be obtained.

关 键 词：三次样条插值 非等间距 灰色预测 

分 类 号：TG230.2[金属学及工艺—铸造]