基于过程神经网络的航空发动机排气温度预测
Prediction of aeroengine exhaust gas temperature using process neural network
-
摘要: 从泛函分析的角度出发,将航空发动机排气温度预测问题转换为一种泛函逼近问题.利用过程神经网络对任意连续泛函的逼近能力,提出了一种基于过程神经网络的航空发动机排气温度预测方法.为克服过程神经网络学习速度慢的问题,为其开发了一种基于正交基函数展开的Levenberg-Marquardt学习算法.将所提出的预测方法应用到某型航空发动机的排气温度预测中,取得了满意的结果.Abstract: From the perspective of functional analysis,the prediction of the aeroengine exhaust gas temperature could be seen as a functional approximation problem in nature.Utilizing the functional approximation capability of the process neural network,an aeroengine exhaust gas temperature prediction method was proposed based on the process neural network.In order to overcome the problem that the learning speed of the existing learning algorithm for process neural network is slow,a Levenberg-Marquardt learning algorithm based on the expansion of the orthogonal basis functions was developed.Finally,the proposed prediction method with the corresponding learning algorithm was used to predict the exhaust gas temperature of some aeroengine,and the results were satisfactory.
点击查看大图
计量
- 文章访问数: 1561
- HTML浏览量: 2
- PDF量: 460
- 被引次数: 0