Study On Time Delay Estimation Methods In Torpedo Trajectory Tracking System

Time delay estimation(TDE) by using three- element linear array is an important problemin torpedo trajectory tracking system which tTacks target according to strictly geometricalposition. It is the basis of measurement of target distance, bearings etc. The error of TDEinfluences directly the accuracy of trajectory tracking. The more efforts have been made in orderto acquire high accuracy estimation. Based on analysis of many references, several TDE methodshave been studied. Using these methods, the simulation data by computer and the real datacollected in the lake experiments are analyZed. Meanwhile, their performances are discussed. Itis very useful for the design of real tracking system. The research results in this theises are1isted as fo1lowst1. According to the presented sensor array form, the corresponding mathematical mode1 isestablished, and factors which influence TDE accuracy have been analyZed in detail. Theinfluence of TDE accuracy to the estimation of tafget distance and bearings estimation isdiscussed also. Meanwhile, The limitation of passive tracking methods and the differencebetWeen passive positioning and passive tracldng which based on three-element array arediscussed.2. Time delay estdriation methOds based on cross-correlation and generalised CTOsscorrelation(GCC), are investigated, and their relationship is discussed. Then thepefformance of several time delay estboation methods is theoretically analyZed.3. AdaPtive LMS methOd of time delay estdriation and time delay estAnation method ofsingle vallable adaPtive based on sinc function are Studied. And their perfOrmances arealso analyzed. In order to improve accuracy of time delay estimation, many interpolationmethods are investigated.4. The theory and method of bispecmim time delay estimation are discussed, and itsperformance compared with cross correlation method is studied. In order to improveaccuracy of time delay estimation and reduce the variance, a combination method of meanfilter and median filter (BM) is investigated, and its performance is also analyZed.5. Base on the simu1ation model of signal and correlated noise model, the performance ofdifferent method is compared under different signal to noise ratio (SNR).6. The perfOrmance of differCni time delay estimation methOd is evaluated based on theanalysis of real data. It is very usefiJl to the design and study of time delay estimationsystem in practice.