多传感器配准估计方法与实现

采用地心地固坐标系作为统一坐标系,研究了多传感器组网中的配准估计问题。首先论述了传感器配准的现实意义和传感器偏差的客观来源。随后给出了多传感器配准问题的数学模型描述,并据此给出了配准参数状态矢量。然后利用单目标在不同传感器探测中的量测,在地心地固坐标系下得出偏差方程,采用一阶泰勒展开进行近似,给出各传感器距离、方位角、仰角偏差的线性化公式,并使用最小二乘法估计出各传感器的实时配准参数。最后通过一个Matlab仿真实例,验证了上述方法的有效性。

Bias estimation in multi-sensor registration

We studied the biases estimation and sensor registration of sensor network by using Earth Centered Earth Fixed(ECEF) coordinates as unified coordinates.The practical meaning of sensor registration and the sources of sensor biases are discussed at first.Then a mathematical model of multi-sensor registration is put forward,based on which the condition vectors of registration parameters are presented.By using the measurement of a single target in different sensors,the equation of biases is obtained in ECEF coordinates,together with the linear equation of sensor distance,azimuth and elevation.The least square algorithm is used to estimate the biases of the sensors online.Matlab simulation results showed the validity of the algorithm.