A novel hybrid strategy for dynamic electricity pricing

Abstract

This invention relates to the field of energy trading, specifically to a hybrid dynamic pricing scheme for electricity. Significant fluctuations in electricity demand have long been a problem for power grid companies, often increasing their costs. With the development of smart grids, real-time dynamic pricing for electricity, as a demand-side management technology, has attracted widespread attention in both academia and industry due to its ability to achieve peak shaving. To address the problem of significant electricity demand fluctuations and improve trading utility, this invention provides a hybrid dynamic pricing scheme based on Stackelberg game theory. Specifically, this invention designs a user dissatisfaction function arising from participation in peak shaving, an incentive cost model for power companies, utility functions for both power companies and users, and constructs an electricity demand fluctuation cost based on electricity demand. This cost is incorporated into the utility function of power companies, and the Sarima-Ann hybrid forecasting model is used to predict electricity demand. This invention can reduce peak electricity demand, improve the utility of both trading parties, and simultaneously improve the accuracy of electricity demand prediction and pricing results.

Description

Technical fields:

[0001] This invention belongs to the field of energy trading technology and relates to a dynamic electricity pricing method based on Stackelberg game theory. Background technology:

[0002] Dynamic pricing is a common method in demand-side management. It can achieve peak shaving by encouraging users to adjust their electricity consumption and shift load. Power companies can incentivize users to shift load by raising prices during peak hours and lowering prices during off-peak hours, or by providing incentives to users during peak or off-peak periods to encourage participation in peak shaving.

[0003] Game theory is often applied in dynamic pricing. When a game reaches Nash equilibrium, unilaterally changing one's strategy will not improve one's utility. By constructing a game model and solving for the Nash equilibrium, the optimal strategy that maximizes the utility of each player can be obtained. Therefore, using game theory for dynamic pricing can yield the price that maximizes the utility of each trading party. In constructing the game model, a reasonable representation of the utility function is the foundation and key to the pricing process, and can better improve transaction utility. Current research largely focuses on improving the utility function, but there is still room for optimization. For example, the design of the user dissatisfaction function and the power company cost function does not fully consider all factors. This paper considers the impact of price and incentives on electricity consumption, represents costs and user satisfaction (and dissatisfaction), and designs utility functions for the power company and users. Summary of the Invention:

[0004] The purpose of this invention is to provide a novel hybrid strategy for dynamic electricity pricing, which includes steps such as electricity demand forecasting, designing cost functions and satisfaction (dissatisfaction) functions, constructing a game model, and solving for Nash equilibrium. The specific process is as follows:

[0005] 1) The power company uses the Sarima-Ann prediction model to predict the electricity demand based on the user's historical electricity transaction dataset L, and obtains the prediction result dataset D;

[0006] 2) Design the marginal cost C of the power company. k Load fluctuation cost flu(l) k Incentive cost E k and user satisfaction function s k And the diss function k ;

[0007] 3) Design the utility function U1 of the power company and the utility function U2 of the users, and construct a game theory model;

[0008] The utility function of the power company:

[0009] User utility function U2:

[0010] Objective functions: (p',l') = argmaxU1, (p',l') = argmaxU2

[0011] Constraints: l min ≤l k ≤l max k = 1, 2, ..., N, l max =min{d max ,z max}, p k ≥C k

[0012] 4) Solve the Nash equilibrium of the game model to obtain the optimal price p' and the optimal transaction quantity l';

[0013] This invention can reduce peak electricity demand, improve the utility of both trading parties, and improve the accuracy of electricity demand forecasting and pricing results. Attached Figure Description

[0014] Figure 1 This describes the structure of the Ann model for the electricity demand forecasting part of this scheme. Detailed implementation method:

[0015] 1) The power company uses the Sarima-Ann prediction model to predict the electricity demand based on the user's historical electricity transaction dataset L, and obtains the prediction result dataset D;

[0016] (1.1) Using the historical electricity transaction dataset L as the input to the Sarima model, we obtain the electricity demand dataset D and the prediction error dataset E.

[0017] (1.2) Using the electricity demand dataset D predicted by the Sarima model and the prediction error dataset E as inputs to the Ann model, the future prediction error dataset FE is obtained.

[0018] (1.3) The future electricity demand forecast results of the Sarima model are combined with the future error forecast results of the Ann model to obtain the future electricity demand forecast value {d} of the hybrid model Sarima-Ann. k}, where d k This represents the electricity demand in stage k on a certain day.

[0019] 2) Design the marginal cost C of the power company. k Load fluctuation cost flu(l) kIncentive cost E k and user satisfaction function s k And the diss function k ;

[0020] Marginal cost C k :C k =a1l k 2 +a2l k +a3

[0021] Load fluctuation cost flu(l) k ):

[0022] Incentive cost E k :

[0023] User satisfaction function s k :

[0024] User dissatisfaction function diss k : eco k =θ(l k -orl k ) 2 , α>0, θ>0

[0025] Among them l k d represents the actual electricity consumption in stage k. avg Represents the average daily electricity demand, ir represents the incentive rate, and orl represents the incentive rate. k Let z represent the electricity consumption in the original k-th stage, μ represent the load fluctuation coefficient, η represent the user preference coefficient, and z represent the user preference coefficient. max This indicates the user's maximum electricity consumption. Represents economic weights, Eco represents the comfort weight. k Represents the dissatisfaction arising from the economy in stage k, com k Let θ represent the dissatisfaction arising from comfort level in stage k, θ represent the coefficient of influence of demand fluctuations on dissatisfaction, α represent the absolute value of the price elasticity coefficient, and r represent the price elasticity coefficient. k This represents the electricity price in the original k-th stage.

[0026] 3) Design the utility function U1 of the power company and the utility function U2 of the users, and construct a game theory model;

[0027] The utility function of the power company:

[0028] User utility function U2:

[0029] Objective functions: (p',l') = argmaxU1, (p',l') = argmaxU2

[0030] Constraints: l min ≤l k ≤l max k = 1, 2, ..., N, l max =min{d max ,z max}, p k ≥C k

[0031] 4) Solve the Nash equilibrium of the game model to obtain the optimal price p' and the optimal transaction volume l'.

Claims

1. A novel hybrid strategy dynamic electricity pricing method, characterized in that: The process includes electricity demand forecasting, designing cost and satisfaction functions, dissatisfaction functions, game model construction, and Nash equilibrium solution. The specific steps are as follows: 1) The power company uses the Sarima-Ann forecasting model to forecast electricity demand based on the user's historical electricity transaction dataset L, obtaining the forecast result dataset D; 2) Designing the power company's marginal cost. Load fluctuation costs Incentive costs User satisfaction function and dissatisfaction function , ;in This represents the actual electricity consumption in stage k. This represents the average daily electricity demand. ir Indicates the incentive rate. Indicates the original number k Electricity consumption during the period Indicates the load fluctuation coefficient. Represents the user preference coefficient. This indicates the user's maximum electricity consumption. Represents economic weights, Represents the comfort weight. Representing the k Dissatisfaction arising from the economy at this stage Representing the k Dissatisfaction arising from comfort level during a certain stage The coefficient representing the impact of demand fluctuations on dissatisfaction. Represents the absolute value of the price elasticity coefficient. Indicates the original number k 3) Design the utility function of the power company; and user utility function Construct a game theory model; The utility function of the power company: User utility function : Objective function: Constraints: 4) Solve for the Nash equilibrium of the game model to obtain the optimal price. and optimal trading volume .

Images