分布式数据库中数据存取效率管理仿真

摘要

This research focuses on the method for rapid access of data in distributed database based on principal component analysis.Our research carried out standardization processing for data and built correlation matrix of data variable,then carried out nonlinearity access of data according to variation relation of data point and worked out output data matrix with dominant mark.In addition,the research eliminated contribution value corresponding to maximum energy from original data set via accessing the output data matrix reversely and obtained data after update processing.The research carried out repeated iteration for the data and obtained data eliminating spectral leak.Simulation results show that the method can maintain precision and integrity of data.It is more beneficial for putting data in storage.%对分布式数据库中数据存取效率管理,可以有效实现分布式数据库资源的互访与共享.对数据的存取效率的管理,需要对数据进行非线性存取,计算出显性标记输出数据矩阵,完成数据的快速存取效率.传统方法先将数据实行等距变换,对变换后的数据组建数据填补模型,但忽略了输出数据矩阵的计算,导致管理效率偏低.提出基于主成分分析的分布式数据库中数据快速存取方法.对数据进行标准化处理,建立数据变量的相关矩阵,依据数据点的变化关系进行对数据非线性存取,计算出显性标记输出数据矩阵,通过对其进行反存取,从原始数据集中剔除最大能量对应的贡献值,得到更新处理后的数据,对该类数据进行反复迭代,获得消除谱泄露的数据,由此完成分布式数据库中数据快速存取.仿真结果表明,所提方法能够较好地保持数据的精度和完整性,更利于数据的入库.