基于UKF的无线传感器异步数据融合优化算法

提出了一种基于无迹卡尔曼滤波(UKF)的无线传感器异步数据融合算法,利用RNAT机制识别无线传感器网络中的冗余节点,构造数据冗余树来实现冗余数据的去除.根据重复数据消除的结果,在每个传感器检测范围半径相等的环境下,采用四圆定位法,任意选择2个检测目标信息的节点,计算2个圆形检测区域边界的交点,根据迭代法找到并近似目标.设定了不同传感器的原始传感器相互独立、同一传感器不同原始量测量值相互独立的前提条件,计算出各通道的测量值,利用未测量卡尔曼滤波器以滤波的形式更新测量值,引入卡尔曼滤波增益矩阵,并结合异步数据定位结果实现数据融合.实验结果表明,融合后的数据利用率高于现有结果,算法耗时短、能耗低,且具有较高的数据融合精度,整个融合的准确率在90％以上.

UKF-based optimization algorithm for asynchronous data fusion of wireless sensor

A UKF-based algorithm for asynchronous data fusion of wireless sensor is proposed. The RNAT mechanism is used to identify redundant nodes in the wireless sensor network, and a data redundancy tree is constructed to implement redundant data removal. According to the result of deduplication, in the environment where the radius of each sensor''s detection range is equal, two nodes that detect the target information are arbitrarily selected by the four-circle positioning method to calculate the intersection of the boundaries of two circular detection regions, finding and approximating the target according to the iterative method. The preconditions that the original sensors of different sensors are independent of each other and the different original quantity measurements of the same sensor are independent of each other are set, and the measured values of each channel are calculated. The unmeasured Kalman filter is used to update the measured value in the form of filtering, the Kalman filter gain matrix is introduced, and the asynchronous data positioning result is combined to realize data fusion. The experimental results show that the data utilization after fusion is higher than the current results, and the algorithm has short time-consuming, low energy consumption, and high data fusion precision. The accuracy of the whole fusion is more than 90%.