传感器数据存在随机一步延迟的线性无偏估计

传感器数据存在随机一步延迟的线性无偏估计

在网络环境中，当信号或数据经网络传输时，时常会出现具有随机性的时滞或延迟现象。针对这一情况，当观测数据存在有不超过一个采样周期的随机延迟时，本文研究了动态系统的滤波器设计问题。首先，通过状态进行扩维的方法将已建立的动态系模型改写成新的模型，从而克服了已有方法中存在的不足；然后，基于新模型开展了滤波器设计的研究，建立的新滤波器除了能有效地解决原有滤波器设计中存在的局限性，通过计算机仿真实验能发现，在相同条件下，新建立的滤波器在性能上还具有一定的优势。

In network, randomly time-delayed or time-tagged situation often occurs, when date or signal is transmitted through network. On the basis of this case ,this paper studied the filter design for a dynamic model with one-step sensor delay .First, rewrite the previous dynamic model into a new one by using the method of augmenting state ,in order to compensate for the deficiency existing in previous algorithms ,Then ,based on this new model ,establish new filters and thus effectively solve the problem of filter design in original paper; Also, it is showed through simulation that in same conditions ,the modified filter has superiority in precision.