动态数据融合算法改进仿真研究  

Improved simulation research of dynamic data fusion algorithm

作　　者：刘文龙[1] 杨辉[1] LIU Wen-long;Yang Hui(School of Information E ngineering,Suihua University,Suihua,Heilongjiang,152000)

机构地区：[1]绥化学院信息工程学院,黑龙江绥化152000

出　　处：《计算机仿真》2020年第4期294-297,共4页Computer Simulation

基　　金：2018黑龙江省高等教育教学改革研究(SJGY20180580);黑龙江省省属高校基本科研业务费科研项目(KYYWF10236180205)。

摘　　要：为了有效滤除传感器网络动态数据携带的冗余信息,提高网络数据准确性与网络节点生命周期,提出了基于时间间隔与数据间隔双重增量的自适应加权动态数据融合算法。获取一段时间内的数据,结合数据间隔得出基准数据,利用其它数据与基准数据的偏差进行数据融合处理,有利于减少网络的负载压力;并根据递推估计将同一类型数据采取多次融合计算,引入自适应理论,利用相对方差对各个数据加权做相应调节,同时对数据修正后的估计权值做融合处理,得到最终的二次加权融合结果,进而提高动态数据融合的精度。通过仿真结果,验证了双重增量自适应加权算法在网络动态数据融合方面的有效性,显著降低了数据冗余程度,提高了数据准确性。In order to effectively filter the redundant information carried by the dynamic data of sensor network and improve the accuracy of network data and the life cycle of network nodes,an adaptive weighted dynamic data fusion algorithm based on double increment of time interval and data interval is proposed.Aquiring the data within a period of time,Benchmark data were obtained by combining data intervals Data fusion processing was conducted by using the bias of other data and benchmark data,helping to reduce network load pressure.And the same type of data were taken for multiple fusion calculation according to recursive estimation.The adaptive theory was introduced to adjust the weights of each data with relative variance.At the same time,the estimated weight after data correction was processed by fusion.The final quadratic weighted fusion results were obtained,further improving the precision of dynamic data fusion.Through the simulation results,the effectiveness of dual incremental adaptive weighting algorithm in network dynamic data fusion was verified,which can significantly reduce the degree of data redundancy and improve the data accuracy.

关 键 词：传感器网络 双重增量 自适应加权 动态数据融合 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]