摘要
通过对已有增长模型,即Logistic模型和Confined exponential模型的修正,建立了一个腐蚀增长模型,即修正的Logis-tic模型。并结合某飞机机翼前沿腐蚀坑的深度数据,应用SAS统计分析软件对新模型和两已有的腐蚀增长模型进行非线性回归分析,然后应用加权最小二乘法反复迭代,求解模型的参数估计值,对模型进行拟合缺适度检验来检验模型对腐蚀增长数据的拟合程度,并对模型的精度进行对比。结果显示三个模型都较好地模拟了腐蚀坑深度的增长过程。而相比之下,修正的Logistic模型的精度较高。
A new corrosion growth model (i. e. the modified logistic model )was developed through the modification of two existing growth models ( i. e. , the logistic growth model and the confined exponential growth model). In SAS system , the corrosion growth models were checked comparing with the depth data obtained from xx airplane wings to show whether the models is appropriate to show the behavior of corrosion growth . Weighted least square method (WLS) was used to obtain the parameters of the models . Then the three models were compared in terms of weighted mean square errors (WMSE) . The result shows that the modified logistic model provide the best fit of the corrosion growth data.
出处
《科学技术与工程》
2007年第19期4956-4960,共5页
Science Technology and Engineering
关键词
腐蚀增长模型
非线性回归
拟合缺适度检验
加权最小二乘法
加权均方误差
corrosion growth model nonlinear regression lack-of-fit test weighted least square method weighted mean square errors