Application of Big Data Analysis in Abnormal Detection of Power Grid Business Data

Abnormal detection of power grid business data is very important in ensuring data security. Use strong encryption algorithm to protect the transmitted data and ensure that the data will not be stolen or tampered with during transmission. When analyzing data, anonymization and desensitization techniques are used to protect user privacy. Firewall, intrusion detection system and other network security measures are adopted to protect the system from malicious attacks. In this paper, the application of big data analysis in abnormal detection of power grid business data is deeply studied, aiming at improving the reliability and security of power system. Through case study, this paper takes a 220 kV main transformer as an example by using the realtime data of a substation, and shows the remarkable effect of big data analysis in anomaly detection. In the research, this paper first collects multidimensional realtime data of main transformer, including current, voltage, temperature and other information. Then, AutoEncode (AE), a deep learning method, is used for feature extraction to learn the hidden features of equipment in normal operation. Combined with the anomaly detection algorithm based on K-means clustering, we successfully captured the abnormal points in the operation of power grid equipment, and provided intuitive abnormal information for power system managers. This multidimensional data analysis and comprehensive judgment enable us to understand equipment anomalies more comprehensively, which is helpful to improve the timely discovery and handling efficiency of power grid equipment faults. Through accurate feature extraction and multidimensional analysis, big data analysis provides a more reliable and intuitive anomaly detection means for power system managers.