计算机科学与探索 ›› 2017, Vol. 11 ›› Issue (10): 1672-1680.DOI: 10.3778/j.issn.1673-9418.1607049

• 人工智能与模式识别 • 上一篇    下一篇

不确定性网络连续高斯协同局部聚类更新方法

张克柱1+,杨  忆2,张  勇1   

  1. 1. 宿州职业技术学院 计算机信息系,安徽 宿州 234101
    2. 淮北师范大学 计算机学院,安徽 淮北 235000
  • 出版日期:2017-10-01 发布日期:2017-10-20

Gauss Collaborative Local Clustering Updating Method of Continuous Frontier for Uncertainty Network

ZHANG Kezhu1+, YANG Yi2, ZHANG Yong1   

  1. 1. Department of Computer Information, Suzhou Vocational Technical College, Suzhou, Anhui 234101, China
    2. College of Computer Science, Huaibei Normal University, Huaibei, Anhui 235000, China
  • Online:2017-10-01 Published:2017-10-20

摘要: 为提高不确定性无线传感器网络(wireless sensor network,WSN)模型的危险边界局部演化特性感知精度,提出了一种基于局部聚类的不确定性WSN模型网络局部前沿协同更新算法。首先,给出基于高斯的WSN感知距离不确定性模型和速度不确定性模型,并给出封闭形式的考虑WSN节点有限处理能力和能量约束的连续贝叶斯局部前沿速度更新模型;其次,基于局部聚类更新算法对WSN网络主节点、列表、辅助列表进行更新,实现危险连续局部前沿的实时更新,实现复杂危险演变特征的分布式准确预测;最后,通过实验对比,所提方法对于传感器节点故障和通信链路故障具有强大的鲁棒性。

关键词: 不确定性模型, 无线传感器网络, 连续前沿, 局部聚类, 协同更新

Abstract: In order to improve the sensing accuracy of the local evolution of the risk boundary of uncertain wireless sensor network (WSN) model, this paper conducts a Gauss collaborative local clustering updating of continuous frontier for uncertainty WSN. Firstly, this paper presents the distance uncertainty model and velocity uncertainty model, and also gives the closed form of the continuous Bayesian local front velocity update model, which considers the limited processing power of WSN node and energy constraints. Secondly, the local clustering update algorithm is used to update the master node, the list and the auxiliary list for WSN, which realizes the real-time update of the continuous local front of the danger, and also realizes the distributed accurate prediction of the complicated risk evolution. Finally, the experimental results show that the proposed method is robust to sensor node failures and communication link failures.

Key words: uncertainty model, wireless sensor network, continuous front, local clustering, cooperative update