Research on the Design, Simulation and Analysis of Residual Current Monitoring System Combined with Data Communication Technology

This paper proposes a residual current monitoring system based on improved wavelet transform and deep learning (WT-DL), combined with broadband carrier communication (PLC) and LoRa wireless communication technology, to build an efficient and low-cost intelligent monitoring network to solve the limitations of current residual current monitoring. This system uses wavelet transform (WT) to extract features of residual current signals, and uses CNN-LSTM deep learning model to improve fault classification accuracy. At the same time, the K-means clustering algorithm is used to optimize the layout of monitoring points to improve data collection efficiency. The experimental data comes from the rural low-voltage distribution network, and is simulated and analyzed by MATLAB and TensorFlow, and compared with traditional WT, CNN, and LSTM methods. Experimental results show that WT-DL algorithm achieves 91.3% fault location accuracy, 13.1% higher than traditional wavelet transform, F1 abnormal leakage identification value 0.82% higher than CNN algorithm by 10.8%, and data transfer packet loss rate decreases by 1.5%. This study shows that the intelligent residual current monitoring system combined with data communication technology has significant advantages in improving fault detection capabilities, optimizing monitoring coverage, and improving data transmission reliability, providing innovative technical support for rural power grid safety management.