摘要
结合旋转机械启停阶段振动信号的特点,提出了一种基于Viterbi算法和短时傅里叶变换(STFT-VA)的瞬时频率估计算法,STFT-VA算法在高噪声、临近阶比和交叉阶比情况下有较高的精度。实现了基于STFT-VA算法的自适应Vold-Kalman阶比跟踪(AVKF-OT),和传统的以硬件实现的Vold-Kalman阶比跟踪(VKF-OT)相比,此方法具有无需转速计等硬件、用纯软件的方法实现。实验结果表明,该方法能够在时域中准确地提取幅值和复杂频率变化的阶比,适合于复杂旋转机械振动响应特征提取。
A new method of vibration signal instantaneous frequency estimation for run-up or coast down of rotating machinery based on short time Fourier transform and Viterbi algorithm(STFT_VA)was presented.STFT_VA algorithm has high precision in the situation of high noise,close orders and crossing orders.Adaptive Vold-Kalman order tracking(AVKF_OT)was carried out based on the algorithm of STFT_VA.In contrast with traditional method of Vold-Kalman order tracking,the method includes some advantages,for example,tacho and other hardwares are not necessary to be employed to do order tracking.The experimental results show that the method can accurately extract varying amplitude and frequency order in time domain,and is suitable for vibration response character extraction of complex rotating machinery.
出处
《振动与冲击》
EI
CSCD
北大核心
2008年第12期112-116,共5页
Journal of Vibration and Shock
基金
国家自然科学基金资助(50675105)
