Self-Adaptive Identification Model of Multi Electricity Consumption Under the Market Competition Mechanism

Studying users' electricity usage behavior in the market environment is beneficial for optimizing the electricity market mechanism; the results of such a study lay a foundation for the data-based operation of the market and promote- the safe and stable operation of the power grid. In this paper, first, the response mechanism of different users in the electricity market environment to the electricity price and demand response policy is analyzed, and a market behavior evaluation index system is built based on the aspects of electricity purchase potential, electricity price sensitivity, demand response potential, etc. Based on the index quantification of the initial user samples, the users' initial class is determined with the secondary clustering method. Second, based on the LVQ neural network and SOM neural network, a self-adapting identification model with a class incremental learning function is proposed. In this paper, the user data of Foshan City is used for model verification, and the example results show that the self-adapting identification model has accurate identification results and can effectively identify new user types; moreover, it also has great superiority in the update of the identification model.