Correlation Analysis Between Multivariate External Data and Power Consumption Behaviors Based on Bayesian Network

Inhalt:
Accurate load sensing is the key basis for flexible regulation of user side resources. Traditional load sensing methods are mainly based on the electrical characteristics such as current and power. In view of this, this paper uses non-electrical characteristics to realize the identification of power consumption behaviors, and proposes a correlation analysis method between multivariate data and power consumption behaviors based on Bayesian network. Firstly, the crawler technology is used to obtain the factors that may affect the power consumption behaviors of users from the Internet. Furtherly, the main influencing factors of power consumption behaviors are selected from many influencing factors through the correlation analysis method. Finally, the Bayesian network model is constructed and trained with historical data to realize the accurate recognition of power consumption behaviors. It is shown that the proposed method can effectively mine the internal relationship between multiple external data and the power consumption behaviors of users and owns the advantages of simple model and high recognition accuracy.