The Hanning self-convolution window (HSCW) is proposed in this paper. And the phase difference correction algorithm based on the discrete spectrum and the HSCW is given. The HSCW has a low peak side lobe level, a high...The Hanning self-convolution window (HSCW) is proposed in this paper. And the phase difference correction algorithm based on the discrete spectrum and the HSCW is given. The HSCW has a low peak side lobe level, a high side lobe roll-off rate, and a simple spectrum representation. Hence, leakage errors and harmonic interferences can be considerably reduced by weighting samples with the HSCW, the parameter estimation by the HSCW-based phase difference correction algorithm is free of solving high order equations, and the overall method can be easily implemented in embedded systems. Simulation and application results show that the HSCW-based phase difference correction algorithm can suppress the impacts of fundamental frequency fluctuation and white noise on harmonic parameter estimation, and the HSCW is advantageous over existing combined cosine windows in terms of harmonic analysis performance.展开更多
为提高车窗电机异常噪声识别的准确性,提出一种以改进的谱减法为基础、以优化的梅尔频率倒谱系数(Mel Frequency Cepstrum Coefficient,MFCC)为特征值的电机异常噪声辨识方法。结合电机声音信号和工厂背景噪声信号特点,通过优化的谱减...为提高车窗电机异常噪声识别的准确性,提出一种以改进的谱减法为基础、以优化的梅尔频率倒谱系数(Mel Frequency Cepstrum Coefficient,MFCC)为特征值的电机异常噪声辨识方法。结合电机声音信号和工厂背景噪声信号特点,通过优化的谱减法进行消噪处理。针对频谱泄漏,用汉宁自卷积窗代替汉宁窗,获得优化的MFCC。实验结果表明,该方法能够有效判别电机是否存在异响,准确率达到91%。展开更多
基金Supported by the National Natural Science Foundation of China (Grant No.60872128)
文摘The Hanning self-convolution window (HSCW) is proposed in this paper. And the phase difference correction algorithm based on the discrete spectrum and the HSCW is given. The HSCW has a low peak side lobe level, a high side lobe roll-off rate, and a simple spectrum representation. Hence, leakage errors and harmonic interferences can be considerably reduced by weighting samples with the HSCW, the parameter estimation by the HSCW-based phase difference correction algorithm is free of solving high order equations, and the overall method can be easily implemented in embedded systems. Simulation and application results show that the HSCW-based phase difference correction algorithm can suppress the impacts of fundamental frequency fluctuation and white noise on harmonic parameter estimation, and the HSCW is advantageous over existing combined cosine windows in terms of harmonic analysis performance.
文摘为提高车窗电机异常噪声识别的准确性,提出一种以改进的谱减法为基础、以优化的梅尔频率倒谱系数(Mel Frequency Cepstrum Coefficient,MFCC)为特征值的电机异常噪声辨识方法。结合电机声音信号和工厂背景噪声信号特点,通过优化的谱减法进行消噪处理。针对频谱泄漏,用汉宁自卷积窗代替汉宁窗,获得优化的MFCC。实验结果表明,该方法能够有效判别电机是否存在异响,准确率达到91%。
