同类多传感器加权最小二乘数据融合

Weighted Least Square Data Fusion of Homogeneous Multisensor

针对同类多传感器测量中含有噪声,提出了多传感器加权最小二乘融合估算法。该算法不要求知道传感器测量数据的任何先验知识,而是通过传感器的测量数据进行方差在线学习估计,及时调整参与融合的各传感器的权系数,使融合结果的均方误差始终最小。对实际系统的采样数据的仿真结果表明了本算法的有效性,其融合结果在精度、容错性方面均优于传统的平均值估计算法。

Aimed at the existing noise during the homogeneous multisensor measurement, a weighted least square fusion estimation algorithm of homogeneous multisensor is presented. It doesn＇t need any prior knowledge on sensor measurable data, but carries on the variance estimation online-by these data and timely adjusts weights of various fusion sensors in order to make the mean-square error of fusion results least all the time. Simulation of actual sampling data shows availability of the algorithm and its results is superior to the traditional average estimation algorithm in aspects of accuracy and fault-tolerance function.