Multi-source And Multi-dimensional Information Data Fusion Technology Based On Passive Sonar

Underwater acoustic target detection has always been one of the hot research fields in sonar field.In recent decades,due to the rapid development of "acoustic stealth" technology,the radiation noise intensity of underwater acoustic targets have been significantly reduced.However,the Marine environment has become increasingly complex,which not only greatly reduces the detection range of sonar,but also causes a large number of interference targets in the measured data.In addition,active detection is easy to expose the position of the platform.Therefore,how to make use of passive multi-platform detection has important application significance.Therefore,in this paper,the full text content passive acoustic target detection based on multi-platform tasks as the core,according to the weak target detection problem,with the aid of target tracking method and data fusion method,filter the interference target information of the measured data,realization of underwater acoustic target state estimation,provide solutions for practical engineering applications.Among them,it mainly includes the following aspects:In order to solve the problem of mass jamming targets in contact level measurement data,the Gaussian Mixture Probability Hypothesis Density algorithm with strong tracking performance and high computational efficiency was discussed.To solve the problem of passive sonar target tracking with only the frequency of the target line spectrum,a multi-target tracking fusion algorithm based on the frequency characteristics of the line spectrum is proposed in this paper.This method filters the line spectrum frequency at the platform level,and then fuses the line spectrum frequency estimation of the multi-platform target by using the data fusion method and the marking method to roughly estimate the target’s motion trajectory,which provides a solution to realize the large range and long-time passive multi-target detection.Simulation results show the effectiveness of the proposed method.To solve the problem of target tracking using bearing-only information under strong interference,this paper proposes a multi-platform fusion positioning method and further proposes a multi-platform passive target tracking algorithm combined with the data compression theory.The simulation results show that the proposed method can filter out the "outliers" effectively and obtain the target’s reliable trajectory information.Compared with the dual-platform tracking results,the number of estimated real targets in the multi-platform tracking results increased,while the false alarm rate and the number of false targets decreased.When the detection probability is low,the data fusion method can effectively increase the number of real targets in the estimation results.Aiming at the problem of passive sonar target detection in high false alarm environment,this paper presents a multi-dimensional information fusion tracking method for underwater targets based on azimuth-line spectrum frequency.This method firstly solves the "outlier" problem of traditional cross-location algorithm by using the multi-dimensional information joint location method,and then performs tracking filtering through GMPHD filter based on marker association to obtain the target track state.The estimation accuracy of the target position is improved obviously by this method,and most of the interference in the measurement can be removed.The experimental results in the South China Sea in 2020 prove that this method has the ability to solve the problem of passive underwater acoustic target detection.