Research On Passive Location Technologies Of Multiple Moving Observers

It is proved that the ability to obtain information is the key factor to determine the success or failure of the modern warfare. How to access the battlefield information stealthily, timely and accurately has been a vital problem. In view of this, in the background of Network Centric Warfare, the research on passive localization using multiple moving observers (MMO) is necessary and important. In order to quickly and accurately locate the emitters, some key technologies have been studied systematically and deeply in this dissertation, including the localization precision analysis and sensor placement optimization, the estimation of observed parameters, the passive localization algorithm, and the systematic error calibration. The research enriches the framework of passive localization, and strongly supports the passive localization engineering using MMO. The main contributions can be briefly described as follows:For the localization precision analysis and sensor placement optimization,the observed parameters and corresponding passive localization systems (PLSs) related to this dissertation is firstly given. Secondly,the existing precision analysis results are introduced. And then considering the navigation parameter errors, by joint navigation parameters and emitter kinetic state variables estimation, the precision analysis is carried out in other three cases.The first one is the case without constraint but with bias.The second one is the case with equality constraint but without bias.The last one is the case with equality constraint and bias. Based on the above precision analysis, the sensor placement optimization is researched. For the problem, the average accuracy is regarded as the objective function and the differential evolution algorithm is introduced, which can effectively solve the nine parameter optimization problem appeared in the placement optimization.And the optimal sensor placement for the triple aircraft passive localization is given.The estimation of observed parameters are investigated, mainly about time difference of arrival (TDOA) and Doppler difference of arrival (DD). Here DD usually is represented by frequency difference of arrival (FDOA).Firstly, the principle and fast algorithm of joint TDOA and FDOA estimation algorithm based cross ambiguity function (CAF) are presented. Secondly, for the narrowband pulse train , the H-TDOAs-MCAF algorithm is put forward, which comprehensively utilizes the histogram method (H for short) , the TDOA sequence method (TDOAs for short) and the modified CAF method (MCAF for short).This algorithm can quickly and steadily abstracts the TDOA and FDOA of the narrowband pulse train. Thirdly, for the problem of TDOA and/or FDOA estimation ambiguity caused by the periodicity of received signals, two disambiguation solutions are presented. One is based on localization and inversion using unambiguous observed parameters. The other is based on multi-hypothesis test. Finally, for the non-narrowband pulse train, frequency domain partition is utilized. And then, two algorithms for joint TDOA and time domain DD estimation are advanced. One is named separate estimation method. The other is called frequency domain cumulate cross ambiguity function (FCCAF) method.The passive localization algorithm is studied.Firstly, for the MMO passive localization using angle and TDOA rate measurements, two algebraic solutions are presented. The first one is for the case with all the observers having the direction-finding ability, which is based on two-step linear, weighted least squares (LWLS). The second one is for the case with only one observer having the direction-finding ability, which is based on passive ranging (PR) .The solution transforms the three-dimensional position estimation into one-dimensional range algebraic estimation, which simplified the problem. Moreover, this algebraic solution is extended to MMO passive localization using DD measurements and a method named angle multi-hypothesis and iterative modification is given.Secondly,for the MMO passive localization using TDOA sequence, a computationally efficient solution named two-stage localization method(TSLM).Thirdly, the transform way for different location systems is researched. Two methods are proposed, including virtual sensor algorithm and virtual direction finding algorithm. Based on these algorithms, the solution of certain specified location system could be applied to other location systems.The systematic error calibration for MMO passive localization is discussed. Without calibration emitters, two ideas of systematic errors calibration are put forward. One is based on multiple times observation information fusion. The other is based on multiple emitter observation information fusion. The QPSO-GNI algorithm is brought forward, which comprehensively utilizes the quantum-behaved particle swarm optimization (QPSO) and Gauss-Newton iteration (GNI). Taking passive localization using TDOA and DD for example, the two ideas and the QPSO-GNI algorithm is validated by numerical analysis and Monte Carlo simulation.