基于串行通信不完全数据采集的融合估计算法

Data fusion estimation algorithm based on serial communication missing measurements

针对监控系统采用串行电路作为数据传输方式具有一定延迟的弊端,提出一种基于异步串行数据融合的新算法。首先将随机接收自多传感器系统多采集目标的数据统一映射到融合中心的坐标系中,在融合周期内,状态更新时采用距前一融合时刻最近的观测值作为反馈,状态和观测预计时采用扩展卡尔曼滤波,从而获得基于全局的融合估计值和误差。仿真结果表明该算法估计误差更低,且随着串行系统数目的增加估计误差的增加幅度有明显降低。

Concerning the multi-way delay problem in the serial communication monitoring system,a new algorithm of data fusion estimation based on serial communication missing measurements was proposed.Firstly,the algorithm mapped measurements randomly received from the multi-sensor system in the reference frame with fusion center.During the fusion period,the measurements,which were close to last fusion period,were used in the state update step.In state and measurements prediction step,the extended Kalman filter was used to obtain the global estimation and error.Finally,the simulation results show that the algorithm has a lower estimation error,and with the increase in the number of serial systems,estimation error rate of increase has decreased.