Big Data Analysis and Intelligent Prediction Technology Based on Intelligent Digital Power System

The increasing complexity of power system structure and the continuous access of new energy sources put forward higher requirements for real-time monitoring, fault diagnosis and optimal scheduling of power system. The whole main design of this study is to build an advanced intelligent forecasting model to improve the operation efficiency and reliability of power system. Through the analysis of massive data of power system, data patterns and laws are obtained, which lays a scientific foundation for optimal operation and intelligent decision-making of power system. Considering the superior characteristics of long-term and short-term neural network in processing time series data, it is selected as the core model. The research process is divided into big data preprocessing, feature extraction, intelligent prediction model establishment and parameter optimization. Pretreatment of big data can ensure the high quality and accuracy of data, and some feature selection is used to select more meaningful features for prediction. Using long-term and short-term neural networks to build an intelligent prediction model, the performance of the model is optimized by adjusting the parameters of the model. The data results show that the predicted energy efficiency can reach 95%, the load forecasting accuracy is $\mathbf{7 5. 5 \%} \sim \mathbf{9 4. 6} \%$, and the frequency deviation can reach 2 Hz. The above results show that the long-term and short-term neural network model has high energy efficiency, high precision load forecasting and system stability, and can be well applied to intelligent digital power systems, providing strong technical support for intelligent operation and maintenance of power grids.