The Technology Of Target Positioning And Data Fusion Tracking With The Asynchronous Multiple Emitter And One Receiver System

With the continuous improvement of the maneuverability of the new carriers capable of moving flexibly underwater such as UUV,the active sonar detection system riding on the underwater unmanned vehicles has shown the advantages of flexibility and rapid response.However,the multi-active sonar detection system will face some physics problems during target detection among which are easy exposure of sound source,easy aliasing of echoes,underwater environmental noise,multi-path phenomenon,direct sound interference from the detection platform,asynchronous measurements and the fusion information transmission delay.To find solutions to the above problems,this thesis carries out research from three aspects: multi-active sonar covert detection technology,target location technology in strong direct sound interference background and target tracking technology based on asynchronous data fusion.When using multi-active sonar detection systems for target detection,the sound source is prone to exposure and the echo signals of different sonars have the problem of serious sound interference at the receiving end.To solve the above problems,this thesis studies the modeling of bio-like signal transmission waveforms based on the sound generation rules and characteristics of whale and dolphin whistle signals.Firstly,the time-frequency characteristics of underwater whale and dolphin “signature whistling” signals are analyzed and the modeling criteria for bio-like signal transmission waveforms are summarized based on this.Secondly,the sparse signal expansion theory is used to realize the modeling of single-component and multicomponent bio-like signal transmission waveforms based on LASSO and block LASSO.At the same time,a strategy of “sliding dictionary” is proposed to effectively improve the accuracy of parameter estimation and the covert performance of bio-like signals.Finally,this thesis analyzes the time-frequency characteristics,mutual correlation,resolution and concealment of the biolike waveforms to further demonstrate that the time-frequency types of the bio-like signals are rich and easy to distinguish,have good detection ability and superior concealment and therefore verifies the feasibility of using multi-access bio-like signals as the transmitting waveforms.The sound interference at the receiving end makes it difficult for the multi-active sonar detection system to accurately locate targets.To solve this problem,research is conducted on the target positioning technique in the background of strong direct sound interference.To obtain the azimuth information of the target,three interference suppression methods,namely weighted vector design,blocking matrix and zero constraint,are used to eliminate the influence of strong direct sound interference.By combining those methods with the wideband beamforming method,the accurate estimation of the target echo in the background of strong direct sound interference is realized.To obtain the distance information of the target,a weighted iterative echo delay estimation method based on interference cancellation is proposed to address the problem of the target echo energy being much smaller than the strong direct sound interference energy,which makes it difficult for traditional methods to accurately estimate the echo delay.This method achieves robust estimation of the target echo delay in the strong direct sound interference background,thereby improving the target positioning accuracy.To solve the problem that it is difficult for multi-active sonar detection system to obtain synchronous measurement information from moving targets at the same time for synchronous fusion,a target tracking method based on asynchronous data fusion is studied.To solve the problem of asynchronous fusion caused by different initial measurement time of multiple sonars,an asynchronous fusion algorithm based on sequential Bayesian smoothing is proposed,which carries out backward smoothing on the basis of sequential Bayesian filtering.By correcting the time offset of fusion time,more accurate estimation results are obtained and target tracking accuracy is improved.To solve the problem of asynchronous data fusion caused by different sampling rates of multiple sonars,a multi-active sonar asynchronous fusion algorithm based on filtering prediction is proposed.Firstly,asynchronous data is advanced to the same time,and then synchronous data is fused to obtain the fusion trajectory of the multiactive sonar detection system,which effectively reduces the target tracking error.Simulation analysis and experimental data show that the proposed algorithm in this thesis is effective and can effectively improve the echo identification ability and target tracking accuracy of the multi-active sonar detection system.