基于压缩感知的互联网数据采集技术研究

Research on Internet Data Acquisition Technology Based on Compression Sensing

互联网数据量大、种类多,传统数据采集技术无法满足当前的数据采集需求。基于压缩感知理论,提出了互联网数据采集的新方法。采用K-SVD字典学习对互联网数据进行自适应稀疏表示,在此基础上进行压缩观测和信号传输。在满足有限等距性的基础上进行信号重构,从而获得高精度的互联网重构数据。将提出的数据采样技术应用于能源互联网中,同时和DCT字典、FFT字典对能源互联网数据的重构结果进行对比。结果表明,采用K-SVD对能源互联网数据采集具有比较高的数据恢复精度,这对互联网数据采集技术的发展具有一定的参考价值。

Due to a large amount and a variety of internet data,traditional data collection technologies cannot meet the current data collection needs.A new method for internet data collection is proposed based on compressed sensing theory.The K-SVD dictionary is used to learn the adaptive sparse representation of internet data,and the compression observation and signal transmission are carried out on this basis.On the basis of satisfying the requirements of finite equidistance,signal reconstruction is carried out to obtain high-precision internet reconstruction data.The proposed data sampling technology is applied to the energy internet,and the reconstruction results of the energy internet data are compared with DCT dictionary and FFT dictionary.The results show that K-SVD has relatively high data recovery accuracy for energy internet data collection,which has a certain reference value for the development of internet data collection technology.