Research On Revenue Promotion Strategy Of Electricity Retailer Based On Data And Model Hybrid Driven

China’s energy structure has entered a period of swift development dominated by renewable energy.The intermittence and fluctuation of renewable energy on the supply side and the uninterrupted integration of random load on the demand side have become a great challenge for the power system.In other words,the contradiction between the development trend of power system and the serious shortage of renewable energy consumption capacity and rapid balance capacity is intensifying.In order to promote the balance between supply and demand,and elevate the utilization of resources,demand side technologies are widely considered,such as demand response,energy storage system,and load aggregation.By virtue of the above technologies,"unidirectional balance the demand side from the supply side" will progressively transform into“coordinated interaction of generation,grid,load and energy storage”.Based on the perspective of electricity retailer,this paper will aggregate consumers and tap the potential of the demand side resources to participate in demand response project by reasonably adjusting real-time price,configuring energy storage system and participating in the demand side energy sharing market,which can not only enrich the profitability measures of retailer,but also promote the balance between power supply and demand for grid.Firstly,an improved CNN-LSTM algorithm is proposed to model power consumption behavior of consumers under the real-time price mechanism.A large number of historical accumulated data are trained to establish the demand response dynamic characteristic function with consumption as input and real-time price as output.On this basis,a real-time pricing method is proposed with the goal of maximizing the net market revenue in the context of the retailer participating in the demand response project.However,the quantity and type of consumers served by retailer are constantly changing,which causes the difficulty in achieving ideal demand response precision only through adjusting RTP.For above consideration,the retailer opts to configure energy storage system for coordinative optimization with RTP to enhance his overall demand response precision and economy.Therefore,a bi-level optimization model is proposed to configure energy storage and coordinate the energy storage operation strategy with the real-time pricing strategy.In recent years,the sharing economy has been proved to effectively promote the optimal allocation of resources.Based on the sharing economy,this paper proposes a mechanism for multiple retailers to form an alliance to share the demand response capability.Further,an optimization model of demand response cost of retailers based on cooperative game is proposed to reduce the cost of each participant and improve the overall economy of demand response project.Through the simulation of the proposed model and mechanism,the conclusions are as follows: by paralleling the CNN-LSTM model with a linear part,the improved one has higher precision,which can satisfy the demand of retailer for quantitative evaluation of DR program.The configuration of ESS can help retailer reduce the loss of electricity sales revenue and improve the overall profit.Similarly,through cooperation and sharing of demand response capability,the total revenue of demand response projects of each retailer has been improved.The research proves that the models and mechanisms proposed in this paper can help retailer improve profit.