Study On Point Targets Tracking Methods By Using Passive Sensors

Target tracking usually means predicting relative position and velocity of a target to some sensors based on known angles. As an important branch of target tracking, bearing-only tracking plays an important role and has great research values and application background in improving survival ability and fighting ability of weapons in modern electronic wars. Because of its important role in the detection of sonar, infrared, laser, navigation, telescope, and electro-counter which is developed fast in the recent years, bearing-only tracking becomes a hotspot in nonlinear prediction. There are two main problems in bearing-only tracking: system's nonlinearity and inscrutability of the range of target. This paper discusses Passive Target Positioning and Passive Target Tracking, and it is organized as follows:First, we briefly introduce the application background and the development of passive target tracking, and then describe some present passive tracking methods.Second, Passive Target Positioning is discussed, and a method aiming at target cursor according to uniform motion in a straight line has been presented, the result of the experiments shows that compared with Extended Kalman Filter (EKF), this method is easier to understand and more practical because of its lower computation complexity.Third, using local linearization and Kalman Filter, we present a linear predicting passive direction tracking method aiming at target cursor according to uniform motion in a straight line, and the simulation results show that the method has the advantage of good stability quality, brief dynamic current and good tracking accuracy. It fits the target, of which the change of cursor is smooth and the initiation position is unknown.Last, we discuss Passive Target Tracking, introduce its development status, basic theory and suitable area, and compare the Particle Filter (PF) with Extended Kalman Filter in target tracking simulation in 2D space. Several experiments results demonstrate that the performance of Monte Carlo-particle filter is better than that of Kalman Filter under conditions of nonlinearity and non-gauss. Then, we apply a Bayes nonlinear filter based on the thought of Monte Carlo-particle filter into passive target tracking, In addition, combined the strongpoint of particle filter and two-point extrapolator, a particle filtering tracking method based on two-point extrapolator for track initiation is presented . The experiments results show that the effect is good in the condition of unknown target's initiation position and the non-gauss noise.