摘要
异构多传感器采集的数据容易存在冲突,使用传统D-S证据理论融合会得到与事实相悖的结果,针对该问题,提出了一种结合支持度和确定度的异构数据融合方法。利用Jaccard系数改进冲突系数K和余弦相似度,并结合两者计算证据之间的支持度,第一次修正证据;使用修正后证据焦元区间的兰氏距离衡量证据确定度;基于支持度和确定度共同确定证据的权重,处理原证据,得到加权平均证据后,按Dempster融合得到结果。算例及应用结果证明,所提方法在处理冲突证据和正常证据时,对比传统证据理论、经典改进方法和相似改进方法,均有更好的聚焦性能和收敛性能,证明了该方法的有效性,为多传感器数据融合提供了一定的参考价值。
The data collected by heterogeneous multi⁃sensors are prone to conflict,and the fusion using traditional D-S evidence theory will yield results that are contrary to the facts.To address this problem,a heterogeneous data fusion method that combines support and certainty is proposed.Jaccard coefficient is used to improve the conflict coefficient K and cosine similarity,which are combined to calcu⁃late the support between the evidence,and the evidence is corrected for the first time.The certainty of the evidence is measured by u⁃sing the Lance and Williams distance between the focal element of the corrected evidence,the weight of the evidence is determined based on the support and certainty together,and the results are obtained according to Dempster’s rule after processing the original evi⁃dence to obtain the weighted⁃average evidence.The algorithm examples and application results prove that the proposed method has bet⁃ter focusing performance and convergence performance when dealing with conflicting evidence and normal evidence compared with tradi⁃tional evidence theory,classical improvement methods and similar improvement methods,proveing the effectiveness of the method and providing a certain reference value for multi⁃sensor data fusion.
作者
项新建
李可晗
曹楹
黄炳强
郑永平
XIANG Xinjian;LI Kehan;CAO Ying;HUANG Bingqiang;ZHENG Yongping(School of Automation and Electrical Engineering,Zhejiang University of Science and Technology,Hangzhou Zhejiang 310023,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2023年第8期1243-1249,共7页
Chinese Journal of Sensors and Actuators
基金
浙江省自然科学基金项目(LY19F030004)
浙江省重点研发计划项目(2018C01085)。
