量测不确定下多传感器一致性数据融合算法

摘要

Aiming at data effective fusion process of multi-sensor information gathering system in observation uncertainty,a multi-sensor consistency data fusion algorithm is proposed.Mahalanobis distance is introduced to approximately estimate the consistency degree between two sensor observations.Combining with the properties of membership function in fuzzy theory,a new support degree function is defined to extract and utilize the local observation information,and the support degree matrix which can synthesize global observation datas is construction.By calculating support degree matrix to solve the weight of each sensor observation data in evaluation system state estimation to estimate the real value of system state by weight fusion.Theoretical analysis and simulation experimental results verify the efficiency of algorithm.%针对量测不确定下多传感器信息采集系统中数据有效融合处理,提出一种量测数据一致性融合算法.首先,引入马氏距离实现两传感器量测数据一致性的近似估计；然后,结合模糊理论中隶属度函数的性质定义一种度量局部量测信息提取和利用效率的支持度函数,进而构建综合全局量测数据间相互支持程度的支持度矩阵；最后,通过对支持度矩阵的计算求解各传感器量测数据在评估系统状态估计中的权重,并通过加权融合方式实现对系统状态真值的估计.理论分析和仿真实验验证了算法的有效性.