基于最大相关雷尼熵与相空间重构的航空发动机复合故障信号特征提取方法

针对低信噪比（SNR），复杂噪声工况下，复合故障信号特征难以提取的问题。提出基于相空间重构融入最大相关雷尼熵解卷积的信号特征提取方法，该方法以雷尼熵为敏感特征范数，以最大相关雷尼熵解卷积为基本方法，并在其中融入具有噪声抑制特性和分解特性的相空间重构技术。结果表明：雷尼熵与峭度相比，在故障灵敏度相当并略好的情况下，对偶发噪声敏感度仅为峭度的18.4%。通过仿真验证，实验数据验证以及台架实验验证，证明了本文方法与现有的对比方法相比，在提取复合故障信号特征方面具有优势。

Composite fault signal feature extraction method for aero-engine based on maximum correlation Rényi entropy and phase space reconstruction

In order to solve the problem of complex fault signal feature extraction under the condition of low signal-to-noise ratio (SNR) and complex noise, a feature extraction method based on phase space reconstruction and maximum correlation Rényi entropy deconvolution was proposed. Rényi entropy was taken as the performance index, and the maximum correlation Rényi entropy deconvolution was taken as the basic method, and the phase space reconstruction technique was incorporated with the characteristics of noise suppression and decomposition. Results showed that the sensitivity of Raney entropy was only 18.4% of the kurtosis when the fault sensitivity was equal to and slightly better than that of kurtosis. Through simulation, experimental data and bench test, this method was proved superior to existing comparison methods in extracting the features of composite fault signals.