基于深度学习的配电网工程数据融合与智能分析方法研究

Research on data fusion and intelligent analysis method of distribution network engineering based on deep learning

配网工程数据的快速增长给数据的处理与分析带来了新的挑战。针对配网工程中数据量大、信息处理困难等问题,开展了基于深度学习的配网工程数据融合与智能分析方法研究。通过建立改进的FP-FRCNN检测模型以及压缩连接金字塔结构,进行池化层调整。采用全智能分析技术进行大数据降噪处理,利用聚类分析的稀疏自编码数据融合算法进行数据融合,将配网终端采集到的现场数据进行数据挖掘及可视化分析。以某区域配网工程数据来进行算例分析,实验结果表明,文中所提方法在不同分辨率下的AP值达到了95.6%,平均相对误差为3.53%。

The rapid growth of distribution network engineering data brings new challenges to data processing and analysis.Aiming at the problems of large amount of data and difficult information processing in distribution network engineering,the data fusion and intelligent analysis method of distribution network engineering based on deep learning is studied.Through the establishment of improved FP-FRCNN detection model and compressed connection pyramid structure,the pooling layer is adjusted. The full intelligent analysis technology is used for big data denoising,and the sparse self coding data fusion algorithm of cluster analysis is used for data fusion.The field data collected by the distribution network terminal is used for data mining and visual analysis.The experimental results show that the AP value of the proposed method in different resolutions reaches 95.6%,and the average relative error is 3.53%.