论文部分内容阅读
研究了一类时变线性动态系统的多速率多传感器数据融合状态估计问题。首先,在不同传感器以不同采样率对同一目标进行观测的情况下,提出了一种多速率建模方法,该方法可将多采样率的融合估计问题转化为同采样率的状态估计问题。随后,利用Kalman滤波对目标状态进行了在线估计,并利用有反馈分布式融合结构对上述估计进行了有机融合,从而获得了目标状态的最优融合估计值。该方法不需要对状态或观测进行扩维,计算量适当,保证了算法的实时性。以景象匹配辅助GPS/INS组合导航为例,在两种采样关系下,分别做了仿真,仿真结果验证了算法的有效性。
A multi-rate multi-sensor data fusion state estimation problem for a class of time-varying linear dynamic systems is studied. First, when different sensors observe the same target at different sampling rates, a multi-rate modeling method is proposed, which can transform the multi-sampling rate fusion estimation problem into the same sampling rate state estimation problem. Subsequently, Kalman filter was used to estimate the target state online, and the feedback fusion was used to fuse the above estimates. The optimal fusion estimation of the target state was obtained. This method does not need to expand the state or observation, the amount of calculation is appropriate, to ensure the real-time algorithm. Taking the scene matching assistant GPS / INS integrated navigation as an example, the simulation is done under the two sampling relations, and the simulation results verify the effectiveness of the algorithm.