论文部分内容阅读
多传感器组网系统中,分立传感器之间存在着时空基准上的差异,即使经过校准,仍然会不同程度地残存系统误差。在这种有偏观测环境中,传统的异步序贯融合算法会出现显著的性能退化。针对上述问题,该文提出有偏观测下的主从融合算法。系统误差对目标位置估计的影响显著,而对速度估计的影响并不明显。因此,新算法采用主从两级结构,应用不同策略对位置和速度的估计分别处理,通过从滤波器的速度融合,为主滤波器的状态估计提供参考信息,以改善估计的性能。仿真结果表明:主从融合算法能够更有效地调和平滑度和机动跟踪能力的矛盾,可显著提高对目标速度估计的精度,改善对机动目标的跟踪性能。
In the multi-sensor networking system, there are differences in space-time standards between discrete sensors, and the system errors will still remain in varying degrees even after calibration. In this biased observational environment, the traditional asynchronous sequential fusion algorithm has significant performance degradation. In view of the above problems, this paper proposes a master-slave fusion algorithm. The influence of systematic error on target position estimation is significant, but the influence on velocity estimation is not obvious. Therefore, the new algorithm adopts the master-slave structure, and applies different strategies to deal with the estimation of the position and velocity separately. The fusion of the velocity from the filter provides the reference information for the state estimation of the main filter to improve the performance of the estimation. The simulation results show that the master-slave fusion algorithm can more effectively reconcile the contradiction between smoothness and maneuver tracking ability, which can significantly improve the accuracy of the target velocity estimation and improve the tracking performance of maneuvering targets.