| Researching turbo-machinery aerodynamic noise generation mechanism and the way of reducing the sound noise.is based on the accurate identification of the blade’s sound source. The sound source tracks blade and keeps the same rotational condition. Using the traditional steady sound source identification method to locate the noise source, it can only identify the radial location, but not identify the relative relationship between the sound source and blade. In recent years, the researchers use the algorithm of beamforming to identifying sound source and have successfully applied it to linear movement of the sound source identification. For rotating sound source localization problem, the paper studies on the way of tracking the rotational noise source using a 60-channel audio and an array of spherical wave propagation beamforming algorithmAccording to the characteristics of the rotating blade, the surface of the target sound source is designed a grid of circular to identify sound source localization and discrete sound source movement in the circumferential direction. The sound source is applied to the angular velocity of the rotating blade and use spatial resolution as the steps at different radial positions. First of all, using the optimized 60-channel wheel array to obtain the acoustic signal and design a target sound source frequency which is the range of 50 Hz band-pass filter. Secondly, the algorithm of the near-field spherical wave beamforming will calculate selecting filter of the first ?? time domain signal to obtain the rotating sound source identifying initial radial position and angle θ, wherein the time ?? selection is based spatial resolution of the array. Finally, the result of sound source identification is based on different recognition resolution steps φ to rotate and obtain the result of the rotational movement of the sound source identification, wherein the step of selecting φ is mainly based on the rotational frequency ω and select the length of τ time. The length of τ is same as a considerable collection interval digital signal sample selection, since the trigger signal and the acoustic signal is synchronous acquisition, so as to track the time interval by the same length of time to get the real sound source rotational position.Under rotational sound source simulation test bench, simultaneous acquisition 60-channel audio signal and the phase signal is calculated by the algorithm to identify the rotational point sound source location at the different speed and sound source position. On a rotating disc sound source simulation experiments, the beepers are fixed radial distance from the rotating disk and wherein the radius is 0.20 m and 0.40 m, the rotational speed is about 80r/min and 128r/min, and the testing distance of the comb array is 1.00 m and 1.50 m and array acquired the experimental data were calculated algorithm. The result of rotation radial position of the sound source is respectively 0.23 m and 0.39 m, the sound source the initial phase angle is 174° and 188°, and the recognition result based on the identification step resolution φ is rotated to obtain the true location of the sound source. The results of the tracing algorithm recognition show that the algorithm can not only identify the radial position of rotating sound source, but also position corresponding to the rotation phase angle. The compared the method with Spectrum beamforming finds that the side lobe suppression of the algorithm is better than spectrum beamforming algorithm in the process of identifying the rotating sound source.The paper proposes the new algorithm which is the rotation point source tracing recognize in condition of the rotating sound source simulation test bench. The algorithm can track low speed rotation point sound source location and it put forward a new idea to obtain wind turbine sound source location. |
PDF Full Text Request