【作者】 陆颖；

【导师】 陆森林；

【作者基本信息】 江苏大学 ， 车辆工程， 2005， 硕士

【摘要】 发动机排气噪声是汽车的主要噪声源,具有频带宽,主要噪声频率随工况变化的特点。传统有源消声器的消声特性,无法随发动机工况变化,因而难以在全工况范围内取得良好的消声效果。另外传统消声器还有排气阻力大,功耗和能耗大的缺点。采用主动控制技术能很好的克服传统消声器的上述缺点。 在主动控制技术中,自适应前馈控制方法是最常用的方法。但是将这种方法运用于排气消声时,将面临参考信号的选择、误差通道的辨识、次级源的配置和主动控制的算法的通用性差等问题。因此,在实际应用或开展实验研究前,有必要先通过仿真来确定系统参数和选择参考信号等。 本论文通过仿真计算,对排气噪声有源消声进行了以下几个方面的研究。 1.录取了发动机不同工况下的噪声,对其进行了频谱分析。指出了发动机的排气噪声的能量主要集中在低频段,有源控制在低频段效果较好。 2.根据Filtered-XLMS算法,编写了仿真程序。通过对所录取的发动机噪声的仿真控制的研究,指出了参考信号改变后,有源消声控制效果的也相应的变化,说明了参数选择的范围与参数优化的可能性。 3.介绍了系统辨识的相关理论,比较了离线辨识和在线辨识的不同点,最后完成了自适应在线辨识控制器的仿真程序,取得了4dB的宽带降噪效果。该仿真程序可移植到DSP上。 4.实测了发动机机体上不同振动点的振动信号。分析了这些信号与发动机排气噪声的相关性。仿真结果表明,参考信号的选择不仅与信号的相关性有关,还与振动信号的能量分布有关,为选择振动信号作为参考信号提供了依据。更多还原

【Abstract】 Engine exhaust is main source of automobile noise. The exhaust noise has wide frequency band and its primary noise frequency is variable with engine working conditions. Traditional exhaust silencer can not match to the variable working conditions, thus it has difficulty to reduce noise substantially in all working conditions. Moreover, the traditional has the disadvantage of high resistance, high energy consumption and power dissipation. Active noise control can compensate above insufficiencies.Among active control techniques, adaptive feed-forward control technique is popular. But this method exists some difficulties when used to control exhaust noise, e.g., the selection of reference signal, the system identification for error acoustic path, the radiation power of secondary sound source, and the poor generality of AANC. Therefore it is necessary to check the parameters and select the reference signal by simulation before the experiment.The following problems are studied.1. Engine exhaust noise in various working conditions is recorded, then their frequency character is analyzed .It is indicated that energy of engine exhaust is concentrated in low frequency band and a good performance of ANC can be obtained in low frequency.2. Simulation program based on Filtered-XLMS algorithm is accomplished. Numerical simulation result shows that the performance of ANC alters with the reference signal. It is feasible to confine and optimize the scope of system parameters.3. The theory of system identification is introduced. On-line identification is compared with off-line identification. Simulation program using adaptive identification and adaptive filter is written . 4dB noise decrease in wide frequency band is influenced by on-line identification. This program can be transplanted to DSP.4. Vibration signal on the various positions of engine is acquired .The correlation between vibration signal and noise signal is studied. Simulation results show that the selection of reference signal depends not only on the correlation ,but also on the distribution of vibration energy, which can be used to guide the signal selection.更多还原