光栅海量云数据的快速聚类方法  

Fast clustering method for raster cloud data

作　　者：邓媛劼[1] DENG Yuanjie(Huanghe Science ＆ Technology College, Zhengzhou 450000, Chin)

机构地区：[1]黄河科技学院

出　　处：《激光杂志》2018年第6期178-182,共5页Laser Journal

摘　　要：针对传统的利用非平衡M-Z干涉仪对云数据的聚类效果不理想,且稳定性偏低的问题,提出基于匹配光栅线性谐调的光栅云数据快速聚类方法。首先,采用并行拓扑结构的SDM网络,塑造云计算光纤网络的并行拓扑结构。提出了多通道云数据快速聚类的思想,通过多通道光开光选择各路FBG传感通道,调解光纤传感器的信号,针对实际光纤传感应用中存在的问题,采用基于弹性梁匹配光栅线性调谐扫描滤波的方法,实现光栅数据多通道数据准确聚类。为了验证所提方法的有效性,进行一次仿真实验,对光栅云数据应变传感特性进行分析,通过所提方法与传统方法的对比实验,进一步说明了所提方法具有更高聚类精度,聚类灵敏度高,适合实际应用。In view of the traditional use of unbalanced M-Z interferometer cluster effect of the cloud data was not ideal, and the stability is low, this grating cloud data fast clustering method based on linear harmonic matching grating. Firstly, a parallel topology based SDM network is used to shape the parallel topology of cloud computing optical fiber networks. Put forward the idea of multi channel fast clustering cloud data, through multi-channel optical switch selection of all FBG sensor channels, signal mediation of optical fiber sensors, fiber optic sensing exists in view of the actual problems in application, using the method of elastic beam matching grating linear tuning filter based on grating scanning, multi channel data accurate data clustering. In order to verify the validity of the method, a simulation experiment of grating strain sensing characteristics of cloud data analysis, by comparing the proposed method with the traditional method, further illustrates the proposed method has higher clustering accuracy, the proposed clustering method has the advantages of high sensitivity, suitable for practical application.

关 键 词：光栅 云数据 快速聚类 方法 

分 类 号：TN253[电子电信—物理电子学]