Load Pattern Recognition Method Based on Fuzzy Clustering and Decision Tree

Load pattern recognition is a key task to determine the power consumption behaviors of customers and classify the customers into different clusters according to the load consumption characteristics. In this paper, a load pattern recognition method based on the fuzzy clustering and decision tree was proposed based on the historical electricity data. Firstly, the fuzzy C-means clustering was used to cluster the loads with the similar curve shape into a cluster and each clustering center represents a specific load pattern. Then, the classification and regression tree (CART) was used to build the decision tree and recognize the load pattern. The effectiveness of the method proposed was tested and verified through actual data, which shows the method has a good engineering application prospect.