内燃机噪声源识别的小波相关系数方法研究

Wavelet correlative coefficient method for identifying internal combustion engine noise source

为了分析内燃机(ICE)噪声信号的时频特性和识别主要噪声源,研究了小波变换中尺度与频率之间的关系,重新定义了连续小波变换,并基于不同小波对同一信号分解时小波系数之间存在极大相关性,提出用规范化相关系数时频图分析噪声信号和识别内燃机噪声源的新方法.新方法能够准确地对信号进行时间和频率定位,且频域结果与信号功率谱相当吻合.对发动机声学信号进行了时频分析,同传统连续小波变换相比较,该方法能够更好地反映信号能量的时频域分布状况.结合声强结果,声学信号时频图能够直接地显示不同噪声源的时频特征.

The relationship between scale and frequency was investigated and the definition of continuous wavelet transfer was redefined in order to analyze the time frequency characteristic of internal combustion engine （ICE） noise signal and identify the main noise source. A new method was presented for the noise signal analysis and ICE noise source identification by the time frequency spectrum of normal correlative coefficients based on the strong relativity between the wavelet transfer coefficients when analyzing the same signal by using different wavelets. The method can exactly determine the time and frequency location and the frequency domain result accords with the power spectrum. The method can show the energy distribu- tion of signal better than continuous wavelet transfer through the time frequency analysis of the engine acoustic signals. The time frequency spectrum of acoustic signal combined with the acoustic intensity test results can directly show the time frequency characteristics of different noise sources.