Research On Household Electric Power Optimization Based On Hourly Spot Price

With the promotion of the market-oriented reform of the power system and the increase of new energy power access,the allocation of flexible resources in the power system according to the price signal has become an important means of regulation in the competitive market environment.At the same time,the proportion of household electricity in social electricity consumption is increasing year by year,but there is a lack of effective regulation means to guide households to use electricity reasonably.Through the introduction of real-time electricity price,the use of price signals to guide households to reasonably arrange the time of electricity consumption,reduce the peak load of household power consumption,and increase the trough load,can improve the system load rate,save power investment,and reduce users' electricity expenditure.Based on an in-depth understanding of the research status and development of electricity price forecasting and household power consumption optimization at home and abroad,this paper puts forward a complete optimization method and model to solve the problem of how to optimize household electricity consumption in the market environment of real-time electricity price.It mainly includes the following aspects:It is proposed to improve the input data quality of the forecasting model by establishing a real-time electricity price feature database.This paper makes an in-depth study of the basic theory of electricity price formation,starting with the real-time electricity price source model,this paper analyzes the measurement of the factors of electricity price formation represented by the marginal fuel and maintenance component,the marginal value component of power generation production and the marginal value component of network operation in the real-time electricity price theory.On this basis,11 real-time electricity price features are extracted,combined with the common features in electricity price forecasting,the feature database is established,and the dimension of the feature data is reduced by t-SNE algorithm,and the high-quality input feature data is obtained.In this paper,an improved Seq2seq-Attention network is proposed to forecast the short-term electricity price,which makes it suitable for the regression of sequence pairs in electricity price forecasting.After the characteristic data of the real-time price characteristic database is input into the network,the simulation results show that this method can effectively reduce the error of electricity price forecasting.A household power consumption optimization method considering users' power consumption satisfaction is proposed,and the power consumption satisfaction index and the adjustment coefficient ? of this index are established to control the optimization results to meet the needs of different customer groups.The genetic particle swarm optimization algorithm is used to analyze and verify the typical household power consumption cases.the results show that this method can meet the needs of different customer groups while meeting the needs of different customer groups.effectively reduce the peak load and cost of household electricity.