High-order Hydrophone Array Signal Processing

The DOA estimation of underwater far-field targets with low SNR is an urgent issue need to be solved for our country in the field of underwater acoustics.Among the existing solutions,the dominant one is the target orientation estimation method based on pressure array and vector array.However,it is difficult to get a big breakthrough only by optimizing algorithm based on the mainstream sensors.We will now focus on higher order sensor and provide a new method to solve such problems.Higher-order physics has more semaphores and includes more information of acoustic field.Recently,this issue has drawn much attention of domestic and foreign scholars.Some achievements on research about high-order acoustic field have been published.This thesis discussed the potential ability of arbitrary order generalized acoustic sensors constructed by higher-order sound field physical components in DOA estimation in the high-order physics field and analyzed the spatial correlation in arbitrary noise fields based on the summarization of the existing research results at home and abroad by using of the Taylor series expansion.The analysis results show that second-order sensors have higher research value than higher order ones.However,the existing acoustic dyadic sensor(ADS)DOA estimation algorithms are all based on the same processing methodology that utilizes the vibration gradient information and velocity information and pressure information of ADS as an independent element(i.e.the vibration gradient is regarded as an analog of the pressure information and velocity information).Hence these algorithms do not give full play to the advantages of ADS.First of all,the thesis deducted the output spatial correlation,single-channel directivity and multi-channel combined directivity of the ADS.These theories are the basis of array design,DOA estimation and performance evaluation of the ADS.Second,the thesis studied AVS signal processing method to ADS and research the ADS signal processing method.Two kinds of beamformer CBF and MVDR to simulate DOA estimation algorithm are used in this thesis based on ADS and ADS uniform linear array(ULA).The estimation performance of ADS(beamwidth,port and starboard ambiguity,signal-noise ratio threshold,noise resisting ability,spatial multi-sound source resolution ability)were analyzed.In this part,the joint processing idea was applied to the ADS DOA estimation for the first time,and obtained a significant gain,which verified the advantages of the joint processing idea.Finally,the results of Monte Carlo experiments prove that the conclusions above is correct.Furthermore,the potential performance of the ADS array in discarding grating lobes and suppressing port and starboard ambiguity is analyzed in this thesis based on the beampattern product theorem of designing the combined directivity of ADS beampattern.Finally,comparing with the principal direction of stress theory in analog mechanics,a method of spatial orientation estimation based on the tensor of ADS of sound field is proposed at the first time.The physical meaning of the algorithm is clear and intuitive.With the proposed algorithm,single sensor DOA estimation can be achieved.The thesis analyzed the principle of this method,and the theoretical estimation error of the method,proved that the method is feasible.All works of this thesis is based on rigorous theoretical derivation.The sensor here mainly refers to the hydrophone because the research in this thesis is based on underwater acoustic,but the method proposed in this thesis can also apply to the air acoustics.