Data Mining and Prediction Technology for Power Internet of Things Based on Big Data Analysis
Weiqing Yao, Xian Luo, Ronghao Yang, Xiangyu Yuan, Xin He · Procedia Computer Science · 2025
The growing demand for high efficiency and high reliability in power system has pushed the power Internet of Things to face new challenges. Based on solving the challenges it faces, this study analyzes data mining and forecasting technology to develop the decision support ability of power system. Firstly, this study summarizes its application status and the role of big data technology in promoting the intelligence of power system. Then this study describes the processing flow of big data analysis, including data acquisition and preprocessing, feature selection and extraction, and data mining algorithm selection and optimization. In this study, the long-term and short-term memory network (LSTM) is particularly emphasized, and the performance of LSTM in terms of energy utilization, response time and system failure rate is evaluated by comparing with the generated countermeasure network (GAN) and the deep confidence network (DBN). The results show that the highest energy utilization rate under LSTM is 98%, the lowest response time is only 88ms, and the highest failure rate is 25%. In conclusion, this study summarizes the superior performance of LSTM in data mining and forecasting tasks, and discusses its application potential in improving power system fault forecasting, load forecasting and energy management.