面向配电网边缘计算装置的两阶段 数据压缩方法

Due to the high sampling rate of the distribution phasor measurement units, massive data is generated and puts a huge strain on the communication and storage systems. Considering the limited computing and storage resources of the edge computing devices in the distribution networks, a data compression method of the distribution phasor measurements based on the exception-swing door trending and the exponential-Golomb code is proposed. First, the swing door trending criterion is modified. By using the improved exception-swing door trending algorithm, the original phasor measurements are lossily compressed. Then, the lossily compressed data is pre-processed, including differential coding and normalization. The 0-order exponential-Golomb coding is used to losslessly compressed to further reduce the data redundancy. Finally, the method is tested based on the practical distribution phasor measurements. Compared with the separate exception-swing door trending algorithm, this method achieves a higher compression ratio while maintaining the same reconstruction data accuracy.