摘要
采用当前方法控制网络数据采集质量时,处理采集的网络数据所用的时间较长,采集得到的网络数据质量较低,存在数据处理效率低和数据质量控制效果差的问题.提出网络数据多通道采集质量跨层控制方法,分析采集得到的网络数据时间序列中的线性关系,根据分析结果建立一元线性回归模型,分析设定的阈值与采集数据之间存在的变化趋势,根据分析结果实时调整在网络中采集数据的间隔,通过更新模型的方法实现网络数据的多通道采集.建立采集数据的时间序列模型,通过内循环和外循环对时间序列中存在的残差进行修正,对采集的网络数据做清洗处理,去除采集数据中存在的冗余信息和噪声点,实现网络数据多通道采集质量的跨层控制.仿真结果表明,所提方法的数据处理效率高、数据质量控制效果好.
For the control of the network data collection quality,the current methods need long time to process the collected network data,and the quality of the collected network data is low. Therefore,a cross-layer control method for multi - channel network data collection quality was proposed. At first,the linear relationship in the collected time series of network data was analyzed. According to the analysis result,the univariate linear regression model was established to analyze the tendency between the set threshold and the collected data. Based on the analysis result,the interval of collecting data in network was adjusted in real time,so that the multi - channel collection for network data was achieved by updating the model. Moreover,the time series model of collection data was built. And then,the residuals in the time series were corrected through the inner loop and the outer loop. The collected network data was cleaned to remove the redundant information and noise points,and thus to realize the cross - layer control for multi - channel collection quality of network data. Simulation results show that the proposed method has high data processing efficiency and good data quality control effect.
作者
姚俊萍
范守祥
李晓军
YAO Jun-ping;FAN Shou-xiang;LI Xiao-jun(Rocket Force University of Engineering,Xi’an Shanxi 710025,China)
出处
《计算机仿真》
北大核心
2019年第10期289-292,共4页
Computer Simulation
关键词
网络数据采集
数据质量控制
数据清洗
Network data collection
Data quality control
Data cleaning
