Research On Short-term Load Forecasting Of Power System Considering Price-based Demand Response

With the progress of smart grid,the global power industry is undergoing a revolutionary change.Smart grid combines information network technology with outdated traditional distribution network,which enhances the flexibility of power resources and reduces carbon footprint.With the launch of government projects and various business promotion activities,smart grid has gradually integrated into people's daily life,among which demand response projects have been widely used and studied.Under the effect of demand response,the elastic resources on the user side are effectively utilized.While both the power company and the user get benefits,the law of power load consumption of the user is also changed.This paper mainly studies the impact of price demand response on the daily load curve of users,so as to improve the traditional load forecasting model,so that it can have good adaptability and forecasting performance in the implementation of demand response environment.Firstly,this paper summarizes the research status of smart grid and short-term load forecasting at home and abroad,introduces the relevant concepts and specific classifications of demand response,and analyzes the new challenges faced by short-term load forecasting in smart grid environment.Then the demand response of time-sharing tariff and real-time tariff are studied.For the TOU price,the fuzzy membership function is used to determine the peak and valley period,and the dynamic peak and valley price is established according to the seasonal characteristics of load.The utility theory of consumers is used to simulate the user load curve under the dynamic TOU price.The sensitivity analysis of parameters shows that the model can flexibly simulate the response of users to the price.In order to accurately reflect the daily load curve rule of users under the demand response,the input quantity of RBF and BP neural network short-term load forecasting model is improved,in which demand response factor is added.The simulation results show that the prediction model with demand response factor has obvious advantages and higher prediction accuracy.For the real-time electricity price,the objective function is to maximize the satisfaction of consumers and the income of the power company,and the elasticity matrix of the user load is considered.The user demand response model under the real-time electricity price is established,and the load situation of the user under the real-time electricity price is simulated.From the simulation results,the income of the user and the power company is increased under the real-time electricity price And the real-time electricity price and the user load have the similar change trend.Through the quantification of electricity price,this paper constructs the Elman NN load forecasting model with demand response under the real-time electricity price mechanism,and analyzes the simulation results.The improved model can more accurately reflect the changes of power users under the real-time electricity price,and has better model adaptability,which provides some help for the future related research.