用BP人工神经网络(artificial neural net简称ANN)算法分 别对飞机结构材料、1Cr17不锈钢腐蚀损伤数据进行学习训练，建立了腐蚀损伤与环境条件 的映射模型，并预测腐蚀损伤值.分析了三种预测方法的预测精度.得到了ANN预测的精度比 灰色GM(1,1)模型及Logistic模型的预测精度高，且对数据有较好的适应能力的结论；采用A NN技术定量预测飞机结构腐蚀损伤是一种较好工程方法.
A prediction model of corrosion damage for aircraft structure and 1Cr17 stainless steel under a varied corrosion environment based on artificial neural net was developed and the nonlinear relationship between a verage corrosion rate,average corrosion loss weight,corrosion depth,corrosion te mperature,immersion duration,concentration was established based on BP learning algorithm.The corrosion characteristic quantity (CCQ) such as surface area,depth and so on can be predicted by means of the trained neural net from the data.The results show that,the model has relative good prediction accuracy and flexibili ty than the gray theory GM(1,1) model and the Logistic model.The prediction mode l based on BP learning algorithm of corrosion damage for aircraft structure is f easible and effective.Thus,by virtue of the prediction model,the future corrosio n status and service duration of engineering structure can be evaluated.
参考文献
| [1] |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%