Algorithm Analysis About Traget Tracking In Wireless Sensor Network

In recent years, along with advanced in the communication technology, MEMS, the sensor technology and knowledge fusion in multi-disciplinary, the wireless sensor network has been promoted to intelligent and diversity, which make the application scope is larger and application level is deeper. Target tracking algorithm and application based on wireless sensor network have been the research hots both at home and abroad.The target tracking is an important application of wireless sensor network. The advantages that low cost, simple operation and maintance, make it suitable for the multi-objective monitoring system in large area. Recently, particle filter theory in positioning, navigation, and tracking problem has developed raplidly, taken more attention by scholars. The particle filter that is a kind of non-linear and non-gaussian filter method based on Monte Carlo theory, broke the Kalman filter theoretical framework, its superiority is non-restriction with the system process noise and measurement noise. This paper, based on Major Science and Technology Project named "Development and Application of pollution sources emergency monitoring regulatory support platform for the pollution sources" in ZheJiang province, study cost-reference filter algorithm based particle filter, also put forward some improvement in this algorithm, apply the improved one to the simulation about target tracking algorithm in wireless sensor network. The paper consists of the fellowing jobs.(1) Analysize structure of wireless sensor network node, summarize target tracking system structure in wireless sensor network.(2) Analysize the optimal bayesian filter and the basic principle of monte carlo method, compare and analyse the kalman filter, particle filter method and price reference particle filter method, make each simulation with these algorithm under the same system model.(3) For the shortage of CRPF algorithm, it puts forward the special improvement, and then proves the rationality of improved one.(4) Put the structure of cluster and improved algorithm into the target tracking simulation.