Intelligent Prediction and Data Service Ptimization Scheme of Power Grid Based on Improved Bee Colony Algorithm

作     者：Liu, Bo Zhou, Wang Ren, Haowen Fan, Zhongkai 

作者机构：China Southern Power Grid Digital Platform Technology Company Guangdong Guangzhou510000 China School of Science of Zhejiang Sci-Tech University Zhejiang Hangzhou310018 China School of Electrical Engineering Northeast Electric Power University 169 Changchun Road Jilin132012 China 

出 版 物：《Journal of Network Intelligence》 (J. Network Intell.)

年 卷 期：2024年第9卷第1期

页      面：253-272页

核心收录：

基　　金：Acknowledgement. This work supported by the "Key Technologies of Digital Power Grid" (Project No.: 2020YFB0906000) 

主　　题：Forecasting 

摘      要：In recent years, the power industry is facing more and more new cyber threats, with the Ukrainian power grid being attacked twice, triggering large-scale black-outs, and the Iranian nuclear power being attacked by virus voltage. Therefore, realizing intelligent prediction and data service optimization of complex power grid has remark-able positive significance. In this paper, we combine narrow-band filter detection method, adaptive learning algorithm and piecewise linear regression analysis technique to build the model of power grids. Then according to the characteristic parameters of the grid, combining with the link prediction algorithm and the optimized Dijkstra algorithm, an optimization method of the grid characteristic model is proposed considering outliers of the reconstructed network, so as to better provide instructive suggestions for the overall network restoration. Then, this paper analyzes some algorithms, such as the classical link prediction algorithm, the anomalous edge link prediction algorithm and anomaly-based grid intelligent prediction algorithm. Not only the multi-objective swarm optimization algorithm but the anomalous link prediction swarm intelligent optimization algorithm is proposed to solve the grid model in this article. By using the SIOA-ALP anomaly link prediction swarm intelligent optimization algorithm and other benchmark social network opinion propagation control methods, the simulation experiment of topological power grid anomaly link prediction is realized. Comparing the experimental results under different data sets, and supplemented with the statistical checkout, advantages of the model and algorithm given in the paper are fully reflected, which can provide a better intelligent prediction and data service optimization scheme for complex large-scale power grid. At the end of this paper, the relevant research directions in the future are prospected and predicted. © 2024, Taiwan Ubiquitous Information CO LTD. All rights reserved.