Abstract:

In view of the limitations of the heuristic method based on experience in the power grid partitioning problem,considering the matching characteristics between the new energy and the load in the power grid partitioning problem,the measurement index of the power grid optimal partitioning is established.Based on the source-load matching degree,theoptimal partitioning of power grid is realized by solving the source-load matching degree coefficients of each region.In the traditional K-means algorithm,the sum of error squares is introduced to screen the best clustering value to avoid the adverse impact of the clustering model on the analysis results.The calculation example uses the operation data of six 500 kV substations in a area to analyze and calculate the source-load matching coefficient under each main transformer.According to the coefficient,the power grid is partitioned.The results show that the proposed algorithm can effectively balance the short circuit capacity between partitions.The improved algorithm has significantly improved the overall clustering effect.

Key words: power grid partitioning, source-load matching, optimal partitioning, improved K-means algorithm