群智协同感知网络中数据质量管理方法研究

Research on Data Quality Management Methods in Group Intelligence Collaborative Perception Networks

针对群智协同感知网络在采集数据时出现的用户移动非均匀性和采集数据质量不可控问题,提出了群智协同感知网络中数据质量管理方法。该方法从非均匀采样数据间的数据结构映射关系和数据恢复残差重构两个维度重构感知薄弱区域数据质量,通过重构感知薄弱区域的数据质量,实现融合用户位置信息的数据结构特征提取、数据映射及数据恢复模型,从而提高群智感知网络的数据质量。实验表明,相较于传统方法,所提方法能够实现感知薄弱区域的高质量数据增强,提升数据的可用性。

This paper proposes a data quality management method in group intelligence collaborative perception networks to address the issues of user mobility non-uniformity and uncontrollable data quality in data collection.The method reconstructs the perceived weak area data quality from two dimensions:data structure mapping relationship between non-uniformly sampled data and data recovery residual reconstruction.By reconstructing the data quality in the perceived weak area,the method achieves the data structure feature extraction,data mapping,and data recovery model that integrate user location information,thereby enhancing the data quality of the swarm intelligence perception network.Experimental results indicate that the proposed method is capable of achieving high-quality data augmentation in areas of weak perception and improving data availability compared to conventional methods.