摘要： 在多传感器融合系统中，时间配准的动态噪声和观测噪声不易确定。为此，提出一种基于自适应卡尔曼滤波的时间配准算法。采用滤波过程的自适应性来估计系统的动态噪声和观测噪声，并对采样数据进行配准前的预处理。实验结果表明，该算法具有数值稳定性好、动态范围大的优点，且自适应性较强，能提高时间配准的精度。
关键词:
多传感器,
数据融合,
卡尔曼滤波,
时间配准,
运动模型
Abstract: For the system of multi-sensor data fusion, it is difficult to know the properties of kinematic noise and measurement noise in the time registration. To solve these problems, time registration algorithm based on adaptive Kalman filtering is proposed. This algorithm uses the adaptation to estimate the properties of kinematic noise and measurement noise, and thus preprocesses the sampling data before time registration. Experimental results show that this algorithm has better numerical stability, larger dynamic range, and stronger adaptability. It improves the accuracy of the time registration.
Key words:
multi-sensor,
data fusion,
Kalman filtering,
time registration,
motion model
中图分类号: