Kalman Filter Tracking Algorithm Based On Adaptive Node Cluster

Target tracking technology in wireless sensor network(WSN) has a widely application inindustrial production, environmental monitoring and military field. The Kalman filter, whichis one of the main algorithms in optimal estimation, has numerous applications in targettracking technology. Research on the Kalman filter using in the target tracking in wirelesssensor network is of great significance. The Kalman filter operates recursively on streams ofnoisy input data to produce a statistically optimal estimate of the underlying system state.This paper focus on the target tracking technology of wireless sensor network, which is ahot spot information technology, and a kind of Kalman filter based on adaptive node cluster isproposed to achieve good tracking performance. This algorithm bases on the adaptive nodescheduling strategy, using one point observation extended Kalman filtering to track the targetin wireless sensor network. A real-time location feedback control system based on wirelesssensor network is designed for testing the tracking algorithm. The experiment result validatesthe good performance of the Kalman filter based on adaptive node cluster and shows that ithas advantages of shortening measurement period, reducing calculation and energyconsumption, and improving tracking precision.Kalman filter also has important application in trajectory prediction. This paper discussesthe trajectory prediction in satellite tracking using radar equipment, and puts forward a kindof external acceleration adaptive Kalman filter. Because of the complex external force, theacceleration of the satellite is unstable and complicated. With the method of linear estimationand parameter identification, the adaptive Kalman filter estimates the external acceleration ofthe satellite synchronously with the process of recursive operation in Kalman filter, so as toreduce the system model error and increase the prediction accuracy. The experiment result,which is the simulation of a set of real radar measurement data, shows that this adaptiveKalman filter has better Prediction accuracy than the traditional Kalman filter and generalpolynomial fitting method.