An improved Current Statistical Model for maneuvering target tracking

摘要

The commonly accepted current statistical model (CSM) for tracking for maneuvering target has preferable performance, but its performance sharply decreases when target is in weak jerk maneuvering state or target actual acceleration exceeds previous given acceleration limits. An improved model is proposed for overcoming the shortcoming of CSM. The improved model is still based on all equations of original CSM, in order to eliminate the need of the given acceleration limits in the equation for adaptively adjusting the variance of target acceleration, it uses the average velocity variation rate between priori velocity estimate from (k-1)th to (k)th instant and posteriori velocity estimate in a sample interval to approximately express the acceleration disturbance from (k-l)th to (k)th instant, further obtains average statistical acceleration disturbance from (k-l)th to (k)th instant, and at last puts the acceleration disturbance and average statistical acceleration disturbance in the corresponding equation of CSM to adaptively adjust the variance of target acceleration. Simulation results indicate that improved CSM eliminates the dependence to acceleration limits, enlarges the dynamic range of maneuvering target tracking and improves tracking accuracy.