基于小波和神经网络的流型辨识研究

Research on Discrimination Gas-liquid Two Phase Flow Patterns Based on Wavelets and BP Neural Networks

针对航空发动机轴承腔润滑中存在的气液两相流问题,应用小波分解中的MALLAT算法将航空发动机轴承腔内各流动状态下的压差信号转换为一组特征量,并将该特征量作为BP神经网络的输入信号,对腔内流型进行识别。研究表明,该方法具有较高的辨识率,可作为轴承腔内流型辨识的一种有效工具。

Based on the problem of the gas-liquid two-phase flow in aero-engine bearing chamber lubrication, a wavelet analysis of MALLAT arithmetic was adopted to analyze the dynamic pressure-difference fluctuation signals of the flow patterns in aero-engine bearing chamber, the relevant characteristics were extracted. The wavelet energy was treated as an input to the BP neural network, an intelligent discrimination of the flow patterns was accomplished. The research results show that the method under discussion has a very good discrimination effect, thus the method can be used as an effective approach on the discrimination of flow patterns in bearing chanmber.