Evaluation and Analysis of Urban Power Grid Operation Status Based on Online Sequence Extreme Learning Machine and Self-Coding Network

The study aims to improve the accuracy and speed of the operating state assessment of urban power grids. The hierarchical analysis and self-coding network are combined to reduce the model input, and the improved online sequence limit learning machine (OSELM) algorithm is adopted to improve the real-time response ability of the model to dynamic data. The results show that the prediction accuracy of the weighted online sequence limit learning machine algorithm is significant, and the prediction average of 0.0994 is similar to the actual value of 0.0991, and better than the prediction value of 0.0946 and the standard 0.0986 algorithm. Moreover, the experimental weighted online sequence limit learning machine algorithm converges faster and improves the prediction accuracy than the traditional support vector machine algorithm after 450 iterations. It can be found that this study confirmed that the accuracy of urban power grid status assessment can be effectively improved by the combination of dimension reduction technology and dynamic update learning mechanism. This research provides a new technical scheme for the real-time monitoring and fault prediction of the operation status of the urban power grid, which is of great significance to ensure the safe operation of the power grid.