Research on Power Consumption Behavior Based on PCA and Classification Algorithm

In order to estimate and predict people's real-time heating demand and optimize the existing electric heating control mode, it is necessary to explore the correlation between electricity monitoring and user behavior. Based on principal component analysis (PCA) and classification algorithm, this paper conducts research on the user's electricity consumption behavior. Firstly, a household branch circuit electricity information collection system is established to obtain the electricity consumption data of a household in Jinan City, Shandong Province. Then, the principal component analysis method is used to reduce the dimensionality of multiple original variables of electricity consumption data and time information into a few independent new variables. Finally, decision tree, support vector machine (SVM) and nearest neighbor algorithm (KNN) are used to train the data respectively, and the accuracy and training time of each algorithm model are compared. The results show that the combination of PCA and KNN algorithm is more suitable for user behavior recognition.