Load Data Compression Based on Integrated Neural Network Model
Cunlong Li, Ronghao Zheng · 2019
The smart grid has been widely developed in recent years. The data produced each day are huge, which makes it hard to transform and store data. Efficient load data compressing are required to decrease the data size. This paper analyzes the load data characteristics and divides these data into 3 types, then designs traditional compressing models for different types of data correspondingly. These models are then integrated with neural networks to achieve better details keeping ability. Finally, a well-designed encoding method are given to get higher compression ratio. Experiments have shown our method can achieve high compression ratio, and many details of original data can be well preserved.