论文部分内容阅读
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。
Aiming at the noise in the same kind of multisensor measurement, a multisensor weighted least square fusion estimation method is proposed. The algorithm does not need to know any prior knowledge of the sensor measurement data. Instead, the algorithm performs on-line learning estimation of variance through the measurement data of the sensor and adjusts the weight coefficients of the sensors involved in the fusion in time so that the mean square error of the fusion result is always minimum. The simulation results of the sampled data of real system show the effectiveness of the proposed algorithm. The fusion results are superior to the traditional average estimation algorithms in accuracy and fault tolerance.