Whereas the “current” statistical model for tracking target suffers from the shortcomings of lower tracking accuracy when tracking no-maneuvering or maneuvering weakly target and heavy depending on the limit acceleration of target, a novel variance adaptive filtering algorithm is presented in this paper. In this paper, the relationship between innovation and variation of acceleration is analyzed and described, and the equation of real time estimating acceleration variance is deduced and used in the filtering algorithm. This algorithm can carry out the self adaptation of acceleration variance with the information of innovation, and overcome the shortcomings of lower tracking accuracy of “current” statistical model when tracking no-maneuvering or maneuvering weakly target and avoid setting the limit acceleration. Theoretics analysis and simulation results indicate that the algorithm in this paper has a better performance in tracking heavily maneuvering or no-maneuvering target. In simulation test, the RMSE of position is decreased by about 28％, and the RMSE of velocity is decreased by about 45％ and above 25％ when tracking no-maneuvering and maneuvering heavily target respectively, and the accuracy of estimated acceleration is better compared with the traditional “current” statistical model adaptive filtering algorithm. The simulation results show the algorithm is valid.%针对"当前"统计模型自适应滤波算法跟踪非机动或弱机动目标时存在跟踪精度不高以及受目标机动加速度极限值先验设置的制约等问题,提出了一种方差自适应滤波算法,该方法利用新息与加速度变化量之间的关系,实现了加速度方差的实时估计和自适应调整,能更好地适应目标非机动和强机动的各种情况,避免了"当前"统计模型在目标非机动或弱机动时加速度方差过大而带来的跟踪精度不高的问题;同时也避免了对目标加速度的极限值设定.理论分析和仿真结果表明,相对于"当前"统计模型,无论是跟踪常速运动目标还是跟踪强机动目标,本算法都具有较高的跟踪精度.仿真结果验证了算法的有效性.
展开▼
