Air Conditioning Load Cluster Electricity Aggregation Model Based on Multi-Dimensional Clustering and Double Decomposition

The participation of multi-type air conditioning (AC) load aggregation in power grid demand response can better help the power grid to suppress the peak-valley difference and absorb renewable energy generation. Considering the strong correlation between AC power consumption and environmental data, an AC load power aggregation model based on multi-dimensional clustering and double decomposition is proposed in this paper. First of all, data preprocessing is carried out on the obtained AC power consumption data and use environment data. Secondly, grey relation analysis (GRA) is used to select the factors with strong correlation with the power consumption of AC load as input variables, and then partitioning and selecting clusters (PSC) -Kmeans ++ clustering method is used to aggregate the original data after screening. Then IVMD-PE-CEEMDAN, an innovative algorithm synthesizing the improved variational mode decomposition, permutation entropy, and complete ensemble empirical mode decomposition with adaptive noise methods was used for secondary decomposition to ensure the decomposition effect, mining the data characteristics, and improving the aggregation effect of the model. This method aims to improve the accuracy and reliability of the air conditioning load aggregation model for smart grid applications.