A Distributed Adjustable Resource Clustering Method Based on Covariance Proxy

Abstract

This paper involves a method based on covariance agent to cluster distributed adjustable resources, including photovoltaic, wind power, energy storage and other potentially adjustable resources, including the following specific steps: collecting DAR data and external characteristic data such as solar radiation intensity, wind speed, ambient temperature and humidity; Analyze the correlation between the external features and the DAR set, select the external features with the highest correlation as the correlation coefficient, replace the correlation coefficient with the covariance and multiply the variance of the DAR distribution; With the goal of minimizing the maximum variance of all DAR clusters, the clustering model and characterization parameters are determined to form a faster and more reliable clustering method; Compared with the variance of brute force calculation, the reliability and timeliness of multi-resource clustering are verified by Python simulation. This method does not need to enumerate and calculate all DAR combinations, and has the advantages of simplicity and accuracy, and is more guaranteed than brute force calculation.