| Passive location has many advantages, including better concealment and longer range,and distributed sensor nodes layout can effectively improve the tracking system performance,and the system can be geratly enhance the larger coverage as well as accuracy by using radio over fiber. A novel distributed passive locations system based on WDM technology is proposed. Considering the widely arranged of sensor nodes in the detection area, using WDM technology to distribute the signals received by each sensor node on the separated optical carrier, and then focusing on integration processing in the center station after transmitted over fiber. Which effectively improves the location accuracy and simplifies the system design without complex clock synchronization and parameter estimation in the sensor nodes. In addition, the system theoretically applies to the pulse signal and non-pulse signal source. Therefore, the research about the theoretical realization and experimental verification of the location system based on the TDOA location algorithm and its error analysis is carried out, furthermore, a appropriate model is established based on the state space, and a tracking algorithm based on Kalman filter is proposed for the tracking of a maneuvering target based on the state space.The main research work focus on the following two aspects:(1) Based on the hyperbolic model of TDOA location technology, Chan algorithm,TDOA estimation and the main sources of system error are studied. For the three-station two-dimensional target location, at first, TDOA obtention in a periodic with pulse modulation signal is simulated, and then the spatial distribution as well as the improvement method of GDOP are analyzed, finally, the conceptual verification of the 433MHz signal source is carried out by experiment and achieved the relative error of positioning accuracy less than 51%.(2) Based on the distributed node passive location system proposed in this paper, the tracking of maneuvering target based on "current" statistical model and curve turning model are studied. The interacting multiple Kalman filter algorithm with excellent tracking performance is used to estimate the state of the target, and the performance is compared with that of the filterless Chan localization algorithm. The simulation results show that the IMM Kalman filter can improve the tracking performance of the system with more excellent application value. |
