Research On The Algorithm Of Passive Locating And Target Tracking

To deal with the wide application of electronic-counter, electronic jamming and anti-radiation missile, it is required to enhance the viability and antijamming ability of each electronic reconnaissance equipment. As an important part of the traditional electronic reconnaissance equipments, active locating system has not met the requirements of modern war because of its own drawbacks. Hence, single station passive locating system has gradually received extensive attention for the advantages of being hard to be detected, strong antijamming ability and long detecting range. Based on the single station bearings-only passive locating system (SSBOPLS), the system's observability and filtering divergence are researched in this thesis. The main works can be summarized as follows.Firstly, the observability of linear and nonlinear systems and the basis of parameter estimation are discussed. Particularly, the observability of single station bearings-only passive locating systems was analyzed and the influence of the moving state parameters of observer and target on the system locating accuracy is also researched. Based on that, the conclusion about the relationship between different moving state parameter and locating accuracy is made.To solve the problem of filtering divergence in the case of low observability and large estimating error to the target's original state in the SSBOPLS, this thesis presents a two-stage multiple mode extended Kalman filtering (EKF) algorithm. At the beginning of filtering, system often has large estimating error. In order to efficiently avoid filtering divergence, the algorithm uses a mode-set which contains multiple modes to approach the real target state mode. When filtering comes to steady-state, a single mode is applied in the algorithm to reduce the calculation complexity and the system storage. Simulation results indicate that the locating accuracy of the algorithm is similar to that of EKF when the original estimation is accurate. When the system has large original state estimating error, the proposed algorithm can still accurately track the target while EKF diverges.