Substation clustering method and device

The invention discloses a transformer substation clustering method and device, and the method comprises the steps: repeatedly obtaining a clustering index of a Gaussian mixture model through employing a K-means clustering algorithm and an expectation maximization algorithm according to the load data of a transformer substation and an updated clustering number until the updated clustering number is greater than a threshold value; according to an entropy weight method, determining the weight of the clustering index, and then according to the weight, establishing a clustering evaluation mixing index; and taking the clustering number corresponding to the maximum value of the clustering evaluation mixed index as the optimal clustering number, and obtaining a transformer substation clustering result and a transformer substation clustering center by using a K-means clustering algorithm and an expectation maximization algorithm according to the transformer substation load data and the optimal clustering number. By ad