摘要
研究现有多传感器的加权融合算法,针对时变非线性系统状态估计的有效融合问题,提出了一种基于动态加权的一致性多传感器数据融合算法。对于多传感器量测,首先利用Unscented卡尔曼滤波器得到局部状态估计值,然后基于层次分析法思想,构建反映局部状态估计结果相互支持程度的一致性矩阵,充分提取数据中蕴含的冗余和互补信息,最后通过对一致性矩阵的求解实现对权重的合理选择。经过蒙特卡罗仿真验证了该算法的有效性。
Existing multi-sensor weighted fusion algorithm briefly is introduced. For the effective fusion of time-varying nonlinear system state estimation, a novel consensus multi-sensor data fusion algorithm based on dynamic weighted is proposed. Firstly, Unscented Kalman filter is used to estimate nonlinear system state. Then, according to the principle of analytic hierarchy process, consensus matrix which can measure mutual support degree of the local state estimation is established, and makes fully use of the complementary and redundancy information among the data. Finally, reasonable choice of weight is realized by means of calculating consensus matrix, the Monte Carlo simulation results verify the effectiveness of the proposed method.
出处
《火力与指挥控制》
CSCD
北大核心
2008年第8期75-78,共4页
Fire Control & Command Control
基金
广东省科技厅资助项目(2007B030402001)
