Details

Autor(en) / Beteiligte

• Jiang, Nan
• Li, Xin Bo
• Liu, Nan Nan
• Long, Xiao Bo
• Jiao, Xiao Yang

Titel

Second-Order Statistics-Based Multi-Parameter Estimation of Near-Field Acoustic Sources

Ist Teil von

• Applied Mechanics and Materials, 2013-11, Vol.461 (Advances in Bionic Engineering), p.977-983

Ort / Verlag

Zurich: Trans Tech Publications Ltd

Erscheinungsjahr

2013

Link zum Volltext

Quelle

Alma/SFX Local Collection

Beschreibungen/Notizen

• In this paper, a new approach based on the second-order statistics (SOS) and acoustic vector sensor (AVS) array is proposed, for localization estimation of near-field acoustic narrowband sources. Firstly, we choose the centrosymmetric uniform linear-array as the AVS arrangement, and the array is consistent with the coordinate axis direction of the acoustic vector-sensor. This estimation method makes good use of the acquisition information from the AVS, such as one-dimensional sound pressure and three-dimensional particle velocity, and has shown preferable performance for the parameter estimation of direction-of-arrival (DOA) and range of target acoustic sources in the near field. The estimation algorithm expands the near-field array manifold of one single acoustic vector sensor to the acoustic vector-sensor’s uniform linear-array, and the near-field acoustic vector sensor linear array output model is deduced. The autocorrelation and cross-correlation function of the velocity field and the pressure field are used to construct the rotational invariance frame, which helps to extract the expected information. Consequently, the closed-form solutions of the incident source’s DOA and range are derived explicitly through the parameter pairing operation. The proposed method reduces the computational burden and has good spatial recognition ability and high resolution in the case of limited array elements. It also has better engineering application prospect. Eventually, the performance of the method is verified by Monte Carlo simulation experiments.