Application of Artificial Intelligence in Power Data Governance and Quality Control

• Published in: 2025 International Conference on Electrical Drives, Power Electronics & Engineering (EDPEE) pp. 135 - 139
• Main Authors: Li, Baohai, Song, Jihong, Li, Hui
• Format: Conference Proceeding
• Language: English
• Published: IEEE 26.03.2025
• Subjects: Artificial intelligence; Data center; Data governance; Deep learning; Deep learning algorithms; Electricity data governance; Fault prediction and diagnosis; Long short term memory; Power system stability; Prediction algorithms; Predictive models; Protective relaying; Quality control; Relay protection; Smart grids
• DOI: 10.1109/EDPEE65754.2025.00028

With the development of technologies such as the Internet, big data, Internet of Things and communications, smart grids can collect massive power data through smart meters. Therefore, the power supply quality of the grid can be effectively improved by studying the massive data. However, in the process of power supply, human or unintentional attacks lead to abnormal power consumption, such as "non-technical losses". These illegal practices not only affect the normal operation of the power grid, but also bring great losses to the power grid. At the same time, unauthorised alteration of lines or instruments can easily lead to power cuts, fires and other accidents, bringing great harm to the safety of the grid concerned. With the development of smart grids, the power system generates a large amount of data. Faced with the management and analysis needs of this data, traditional methods have shown bottlenecks. The article discussed in detail the application of deep learning algorithms, especially Long Short-Term Memory (LSTM) Network, in power relay protection, including data preprocessing, model building, and fault prediction and diagnosis. Through experimental verification, the effectiveness and generalization ability of the model on different datasets were demonstrated (with a training set loss of 0.005 and an accuracy of 0.97), demonstrating its potential in improving power system security and management efficiency. These research results have important theoretical and practical significance for optimizing the operation of the power system and improving the value of data assets.