Study Of Multi-sensor Target Tracking Algorithm

Multi-sensor data fusion technology as a new discipline in the military and civilian areas has a broad application prospects, Multi-sensor target tracking is a paradigm of data fusion technique. Multi-sensor target tracking effectively combined the information obtained from different sensors to increase the precision of target state estimation. It has more advantages than only use of single sensor.This paper concerns the problem of target tracking by data fusion technology.Firstly, we state the basic definition, application and art method of data fusion which is a burgeoning field of information processing. At the same time the application in target tracking of multi-sensor data fusion is summarized. Overview of multi-sensor data fusion filter algorithm, we use Kalman filter as data fusion algorithm, simulation results show that: use of two sensor fusion tracking better performance than a single sensor.Secondly, we concerned the essentials of target tracking, such as the setup of target maneuvering model, and some traditional filter algorithm. This article analyzed and compared extended Kalman filter, particle filter and improved particle filter.Finally, we put forward by Bar-shalom interacting multiple model - probabilistic data association algorithm, a multi-sensor will be extended to the case of the method, the formation of a multi-sensor-based interactive multi-model fusion target tracking algorithm. Simulation results show that the algorithm was effective for multiple model target tracking and use of multi-sensor fusion has better than only use of single sensor in tracking accuracy.