Research on Power Outage Sensing Technology of Medium and Low Voltage Distribution Network Based on Deep Learning of Multi-source Data Fusion

Abstract

Aiming at the application of multi-source data fusion in intelligent operations of medium and low voltage distribution network system, a deep learning algorithm based on multi-source data fusion is proposed in this paper. Firstly, combining the stop/restart signals of a large number of users’ digital devices with the existing measurement information of the grid, two indicators of heartbeat and handshake signals have been constructed. Secondly, the convolutional neural network is combined with the improved D-S evidence theory, and an improved depth learning algorithm is proposed. Finally, taking a community as an example, the proposed algorithm is applied in practice and analyzed by simulation. The test results show that the model has good convergence and robustness. It is more intuitive and comprehensive in power failure evaluation. It has a wide power failure perception range and high accuracy and improves the quality of power supply service.