带反馈分布式不同维传感器状态估计技术

State estimation techniques for radars with different observation dimension in one distributed data fusion system with feedback information

论述了带反馈分布式信息融合系统中传感器观测维数不同时的状态估计方法。在多传感器系统中,目标的测量是在极坐标中获得,而数据处理通常是在直角坐标中完成。由于二维传感器不能获得观测目标的俯仰信息,其估计将会产生一定的动态误差。在分布式的信息融合系统中,各传感器需要将各自的局部估计送到融合中心进行统一处理。如果不适当地减小这种动态误差,必然将导致整个信息融合系统的估计精度下降。针对上述问题,文章讨论了如何利用系统反馈信息,减小二维传感器的局部估计误差,从而提高整个系统的估计精度。最后,给出了算法的仿真分析。仿真结果表明该方法能够有效地减小系统动态误差。

The method of state estimation is discussed,when radars have different observation dimension in one distributed data fusion system with feedback.In multisensor systems,the measurements of targets are got in one polar coordinate.However,the data process is commonly completed in one Cartesian coordinate.The state estimation of the local node will produce some dynamic error,if 2-D sensor can't get the pitch information of targets.In the distributed type,every radar processes its observations locally to produce local estimation,and then communicates its estimation to a central processor,where data fusion takes place.If the pitch get bigger,the error will be bigger,which will increase the estimation error of the system and the 2-D sensor.In order to reduce the error,some measures must be used.The paper utilizes the feedback information of the system to reduce the local estimation error and improve the estimation precise of the system.At last,a Monte Carlo simulation is used to analyze the performance of the method.The simulation results show the proposed method can efficiently reduces the system error.