一种多传感器数据时空融合估计算法

当采用分布在不同空间位置上的多传感器观测值对测量噪声干扰下的参数进行融合估计时,数据融合存在时间性与空间性。为了提高测量精度,基于参数估计理论,提出一种多传感器数据时空融合算法。该算法将数据融合分解为两次估计,第一次是基于时间的递推融合估计,第二次是基于空间的自适应加权融合估计。该算法不要求知道测量数据的任何先验概率分布知识,编程简单,计算量小。计算机仿真表明,该算法在减少测量误差方面优于目前已有的基于时间或基于空间的多传感器数据融合算法。

Spatial-temporal estimation algorithm for multisensor data fusion

When the survey data sampled by multiple sensors are adopted to estimate the parameter under the interference of measurement noise, the multisensor data fusion demands timeliness and spatiality. In order to improve the measurement precision, based on the parameter estimation theory, a spatial-temporal estimation algorithm for multisensor data fusion is presented. The process of this algorithm can be divided into recursive fusion estimation based on time and adaptive weighted fusion algorithm based on space. The algorithm needs no prior probability distributing knowledge of measurement data, and is easy to realize with simple programming and calculation. The simulation results show that this algorithm is better than those separately based on time or space in reducing measuring error.