Line Loss Prediction Method for Urban Distribution Networks Based on Optimized RBF and K-means++ Algorithm Under Double Carbon Targets

The application of digital technology to optimize urban distribution networks and lower carbon emissions has drawn more attention as a result of the dual carbon targets and the quick development of artificial intelligence. In order to reduce the carbon emissions of urban distribution networks under the dual carbon target, a strategy for urban distribution networks based on K-means++algorithm has been proposed. The study combines improved particle swarm optimization algorithm for parameter optimization of radial basis function network. It also combines the K-mean++ algorithm to classify and manage the user line loss by time period. In this way, the goal of indirect emission reduction can be achieved. The outcomes revealed that the line loss prediction model has significant advantages in prediction accuracy compared to the comparative model, and can better capture the dynamic changes in electricity demand. After being applied in State Grid Chengdu Power Supply Company, the carbon emissions of the distribution network from July to December were significantly reduced, with the total carbon emissions and carbon emissions caused by line losses reduced to 60000 tons and 51000 tons, 55000 tons and 46750 tons, respectively. This provides new theoretical support for achieving the dual carbon goals and optimizing carbon emission management in power enterprises.