Research Of Data Fusion Technology Based On Multi-Vector Hydrophones

Vector hydrophone and Vector hydrophone array are more widely used in the field of underwater acoustic. The vector hydrophone itself has the ability of directional estimation, but its accuracy of estimation is limited and its ability of anti-interference is poor. In order to improve the accuracy of directional estimation of vector hydrophone and array, in this paper, data fusion technology will be used for combined treatment of multi-vector hydrophone and array, and for increasing the detection performance of vector hydrophone.First, the paper researched the decision fusion algorithm of multi-vector, and derived for optimally weighted least squares algorithm, finite window weighted least squares algorithm and adaptive weighted least squares algorithm of directional estimation of vector hydrophone, compared with the characteristics and application of the three algorithms. As the adaptive weighted least square algorithm takes into account the performance of hydrophone and combined effects of environmental noise, the detection performance of vector hydrophone can be more truly reflected by the algorithm. The paper also researched data fusion algorithm of sound intensity. Because the algorithm uses more redundant information of sound intensity, it can increase the accuracy of estimation. The paper researched the fusion algorithm of wideband signal based on the method of histogram statistics. The result showed that selecting the appropriate number of vector hydrophone can obtain the optimal fusion performance when SNR is fixed.Second, the paper researched the decision fusion algorithm of vector hydrophone arrays. Using the differences of different positions of multi-array, and get the fusion results of the estimation of each array. The algorithm can increase the directional accuracy of array at different incident angles. The paper also researched the data fusion algorithm of vector hydrophone arrays. The transformation matrix is be used for transforming covariance matrix of other arrays to the spatial coordinates of reference array, then weighted fusion the multi-array covariance matrix. The result showed that because the correlation between array elements is considered by the data fusion algorithm of multi-array, the gain of array is increased and the noise is suppressed. The paper also discussed the method of directional estimation of vector array with baffle. Signal processing model of vector hydrophone with baffle is established, and derived for beamforming algorithm of vector array with baffle. Analysis showed that the array can obtain more gain of array with baffle than in the free field, when array and rigid baffle spacing, signal frequency and incident angle meet certain conditions.Finally, the paper dealt with the data of experiments. The result showed that the fusion algorithm of multi-vector hydrophone not only can effectively improve the accuracy of directional estimation, but also can increase the anti-jamming capability of vector hydrophone. The fusion algorithm of vector array can improve the accuracy of directional estimation in a certain extent, and can eliminate the sensitivity of direction of vector array.