| It is of engineering importance to detect low altitude moving targets such as helicopters, due to the blindness of the traditional radar. The passive acoustical direction finding methods is the one of the key technologies of the detection system for low altitude targets. Based on the current research status and the main problems, this dissertation not only improved the methods now available, but also carried out a deep study on new ones for direction finding of low altitude moving targets. The main contributions are as follows:1. The characteristics of target sound sources and sound field are reviewed, respectively. The mechanism of the produce of target noise, its transmission characteristics in air and several main factors that influence the sound transmission are analyzed. By the analysis of the noise data radiated by a helicopter, the noise characteristics of low altitude targets are presented.2. The theory and algorithm for low altitude target localization based on time delay estimation (TDE) is studied. Considering the problems facing by the localization method for widely used four-element cross array, a five-element cross array scheme is proposed. The localization equation is derived, followed by the location accuracy analysis. The influence of TDE error, effective sound speed error and array size measurement error to target location accuracy is discussed respectively, both in theoretical and numerical sense.3. TDE algorithms based on frequency domain adaptive filtering and higher-order spectrum are investigated, respectively. (1) The performance of the frequency domain adaptive TDE algorithm is analyzed, and its convergence condition is also derived. (2) A Wiener weighting based frequency domain adaptive TDE approach is proposed. In order to improve the performance of the new algorithm in low signal to noise ratio (SNR), a method for weighting the phase data is also given. The effectiveness of the algorithms is evaluated by the simulation experiments on real helicopter noise both in static and dynamic environments. (3) A bispectrum slice based least squares TDE algorithm is proposed. The simulation results show that the method outperforms the bispectrum based TDE method both in uncorrelated Gaussian noise and correlated one, and the computation load for the new one is less than that of bispectrum based TDE method.4. This dissertation studied firstly the passive acoustic direction finding approaches based on pressure difference vector sensor (PDVS), it is a beneficial explore in accurate direction finding with a small size array under condition of low frequency band. (1) A two-dimensional PDVS based direction finding algorithm is proposed, followed by the presentation of the procedures for engineering realization.The direction finding accuracy error due to sound field calculation error, channel variances in amplitude and phase characteristics are analyzed, respectively. Numerical simulations and analysis with real helicopter noise are studied. The relationships between the direction finding accuracy and sensor size, time of integration and SNR are presented, respectively. Under the same array size, the performance of proposed method is compared with that of the TDE based one. (2) A new DOA (Direction of Arrival) estimation algorithm based on complex intensity measurements using a three-dimensional PDVS is proposed. The influence of sound field calculation error to the direction finding accuracy is analyzed. The relationships between the direction finding accuracy and sensor size, time of integration and SNR are investigated by the simulations on real helicopter noise data, respectively. The influence of channel variances in amplitude and phase characteristics to the direction finding accuracy are also analyzed. The theoretical analysis show that the performance of proposed method outperforms that of the TDE based one under the same small size.5. The acoustic vector sensor array (VSA) based wideband passive direction finding problem is studied. (1) Based on the established wideband array signal model for VSA, A VSA based wideband coherent signal subspace optimal beamforming algorithm is proposed. The wideband focusing is firstly introduced to VSA array processing, and the relationship between the focusing matrixes of pressure sensor array and VSA is derived. The spatial spectrum for the VSA wideband optimal beamforming is derived herein, this method need no pre-estimation for the direction of sources by wideband focusing with spatial re-sampling technology. The performance of proposed beamformer is assessed by simulation results. The influence of array amplitude and phase error to performance of the wideband optimal beamformer is analyzed. (2) A new two-dimensional DOA estimation algorithm based on VSA is proposed. This method suffers no difficult problems of two-dimensional parameter search and paring usually encountered by some two-dimensional DOA estimation approaches. The simulation results show that the proposed method can decide the number of sources reliably and has a better DOA estimation accuracy, even there are variances in amplitude and phase characteristics among array channels.
|
PDF Full Text Request