Single Observer Passive Location Based On A Number Of Outside Radiation Source And The Target Tracking Algorithm

Passive location system can be grouped into two types: the multi-station system and the single station system. The multi-station system needs the stations working synchronously and requires number of data communications, it also needs fusion algorithm to process the data, so the system gets complex and its flexibility is limited. For this reason, the single station passive location method is getting increasing attentions nowadays.In this dissertation, the passive location and tracking algorithm using single observer based on Non-Cooperative Emitters is studied. The main work of the dissertation is as follows:1. Location algorithm in the condition that the receiver can receive the LOS from the emitters is studied. In the three-dimensional space, the azimuth and elevation of the reflected signal formed a direction surface and a conic surface respectively, and the TDOA between the LOS and reflected signal formed an ellipsoid surface. The intersection of the three surfaces is the position of the target. In the condition that there are more than one emitters, the location accuracy can be improved by using the data fusion algorithm.When the number of the emitters is enough, the doppler shifts are measured to improve the location accuracy furtherly.2. Location algorithm in the condition that the receiver can't receive the LOS from the emitters is studied. The doppler shifts of the reflected signal include the position and velocity information of the target, so the position and velocity of the target can be jointly estimated by the AOA and doppler shifts of the reflected signal in certain conditions. Because the velocity of the target is also need to be estimated, this method needs a number of emitters. When the number of the emitters is less, to location the target, continuous observation is required, and the target must be moving in the mode of invariable velocity or invariable acceleration movement, so there are some limitations.3. Tracking algorithm based on kalman filtering is studied. In last years, the extended kalman filter (EKF) algorithm has become the usual algorithm to deal with the nonlinear dynamic systems. Unfortunately, because of the linearization, the EKF produces large errors and easily lead to divergence. To overcome this problem, the unscented kalman filter (UKF) algorithm is introduced in the dissertation. This algorithm estimates the posteriors with some typical sigma point. It amends flaws of the EKF, and have better properties such as robustness and accuracy. In the two conditions that the receiver can or can't receive the LOS, the EKF and UKF algorithms are both studied, the tracking performance is compared by the computer simulation.