论文部分内容阅读
针对不同传感器以不同采样率、异步对同一目标进行观测的一类线性时不变动态系统,给出了一种有效的状态融合估计方法。利用该方法进行状态估计,首先根据多尺度系统理论,针对每一个传感器分别建立起相应的系统模型;然后利用Kalman滤波和有反馈分布式融合结构进行数据融合并给出状态估计。该方法避免了插值以及状态和观测的扩维,具有较好的实时性。理论分析和仿真结果均表明,融合估计结果在估计误差方差最小意义下,优于最高采样率的传感器Kalman滤波的结果,融合算法是有效的。
For a class of linear time-invariant dynamic systems with different sensors sampling at different sampling rates and observing the same target asynchronously, an effective state fusion estimation method is given. Firstly, according to the multi-scale system theory, the corresponding system model is established for each sensor. Then the Kalman filter and feedback distributed fusion structure are used to fuse the data and give the state estimation. This method avoids interpolation and expansion of state and observation, and has better real-time performance. Both theoretical analysis and simulation results show that the fusion algorithm is more effective than the Kalman filter with the highest sampling rate under the minimum estimation error variance.