水下航行器动力装置振动频率自动识别方法

An Automatic Recognition Method of Vibration Frequency for Underwater Vehicle Power System

为满足水下航行器动力装置振动试验研究对频谱分析的及时性要求,本文以分析对象振动信号的频谱特征和离散傅里叶变换的数学原理为基础,将频谱细化算法和峰值搜寻算法有机结合,构造了一种振动频率自动识别准确率高的自动频谱分析方法,并设计编制了模块化架构的程序。该方法利用频谱细化算法准确计算动力装置周期性激励源的测试频率,利用峰值搜寻算法有效提取振动信号优势成分,以筛分出的峰值频率和周期性激励理论频率的一致性程度来确定振源,通过仿真信号和工程实际振动信号进行应用分析。分析表明,该方法可明显提高试验数据的分析效率和准确性。

To satisfy the timely demand of vibration spectrum analysis for underwater vehicle power system in vibration test,on the basis of the signal spectrum feature and discrete Fourier transform（DFT）principle,an automatic vibration fre-quency recognition method is presented via zoom-FFT（fast Fourier transform）combining with peak search algorithm,and the corresponding modularized program is coded.The method uses zoom-FFT to calculate the exact measuring fre-quency of the power system periodic motivation,and utilizes peak search algorithm to extract the dominant frequency component in order to confirm the vibration source with the consistent degree of two kinds of frequencies.The applica-tion of the present method to simulated signal and real signal shows the analysis efficiency and accuracy are improved markedly.