论文部分内容阅读
为更好地识别运动声源并解决声源识别中的虚假声源问题,基于运动声源短时信号的Doppler频移特性,建立运动声场的短时波叠加关系,利用波束形成方法对声源点进行预估,基于预估建立起多运动声源的动态叠加方程,进一步通过波叠加方程的求解进行声源的计算,从而创建一种可以用于运动声源识别的动态波叠加方法.该方法可以有效识别运动声源,将波叠加方法扩展到了运动声源测量领域,并在不增加传声器数量以及改变阵列形式的条件下有效抑制运动声源重建中的旁瓣效应,解决运动声源识别中的虚假声源问题.仿真及实际运动声源的测量试验结果证明了该方法的有效性.
In order to recognize the moving sound source better and solve the problem of false sound source in the sound source recognition, based on the Doppler frequency shift characteristics of the moving sound source short-time signal, the short-time wave superposition of the moving sound field is established. By using the beamforming method, Point estimation, the dynamic superposition equation of multi-motion sound source is established based on the estimation and the sound source is further calculated by solving the wave superposition equation so as to create a dynamic superposition method which can be used for motion sound source recognition. The method can effectively identify the motion sound source and extend the wave superposition method to the field of motion sound source measurement. The method can effectively suppress the sidelobe effect in the reconstruction of the motion sound source without increasing the number of microphones and changing the array form, In the problem of false sound source.The results of the simulation experiment and measurement of the actual exercise sound source prove the effectiveness of the method.