Demand-side adjustable resource pricing method for incomplete information game

Abstract

The invention discloses a demand side adjustable resource pricing method for an incomplete information game, and belongs to the technical field of adjustable load resource pool optimization scheduling, and the method comprises the following steps: S1, calculating an incomplete information variable fluctuation range; s2, establishing an incomplete information game model; and S3, solving the incomplete information game model. The demand side adjustable resource pricing method for the incomplete information game is used for supporting the construction and scheduling of an adjustable load resource pool, the participation of a plurality of market subjects is considered, the income change condition among different themes is effectively reflected, and the stability of power transaction is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of optimized scheduling of adjustable load resource pools, and more particularly to a pricing method for demand-side adjustable resources with incomplete information games. Background Art

[0002] Currently, methods for evaluating the potential of demand response users mainly include pricing methods such as marginal cost pricing and nodal pricing. However, for power companies, the number of their power users is huge, and some users do not have a deep understanding of demand response (DR). The method of questionnaire survey has problems such as long time consumption, large workload, strong subjectivity, and low reliability. There are also some defects: first, there is no fully market-based pricing mechanism, making it difficult to effectively reflect the income changes between different entities; second, in the power market, the load and electricity price fluctuate greatly, with large uncertainties, and this will directly affect the planned income of each entity. Summary of the Invention

[0003] The object of the present invention is to provide a pricing method for demand-side adjustable resources with incomplete information games, which is used to support the construction and scheduling of adjustable load resource pools. Considering the participation of multiple market entities, it can effectively reflect the income changes between different entities and improve the stability of power trading.

[0004] To achieve the above object, the present invention provides a pricing method for demand-side adjustable resources with incomplete information games, including the following steps:

[0005] S1. Measuring the fluctuation range of incomplete information variables: Based on historical data, predict the clearing price and the response ability of demand-side adjustable resources in the power market on the operating day, and calculate the random fluctuation range of the clearing price and the response ability of demand-side adjustable resources;

[0006] S2. Establishing an incomplete information game model: The aggregator represents the demand-side adjustable resources to participate in the power market interaction facing incomplete information, where the incomplete information includes the clearing price and the response ability of demand-side adjustable resources;

[0007] S3. Solving the incomplete information game model: Use the column constraint generation algorithm C&CG to solve it; convert the original problem model into a two-layer optimization problem including a master problem and a sub-problem for processing, and use the master problem and the sub-problem to alternately iterate to solve the answer.

[0008] Preferably, in S1, the clearing price of the power market is estimated according to the historical clearing price, specifically as follows:

[0009]

[0010] In the formula: is the predicted value of the clearing price at time period t; is the clearing price for time period t on date d; N com is the number of historical data used;

[0011] Then, the maximum and minimum values of the historical clearing prices are used as the fluctuation ranges of the clearing price on the operating day, specifically as follows:

[0012]

[0013] In the formula: λ com,t represents the real-time market clearing price, and are the upward and downward fluctuation ranges of the clearing price in time period t;

[0014] Under the condition that the baseline load is known, the response ability of the demand-side adjustable resources is determined by the actual load on the operating day and estimated from the historical response data, specifically as follows:

[0015]

[0016] In the formula: is the predicted value of the actual load of the demand-side adjustable resource i in time period t on the operating day; is the load of the demand-side adjustable resource i in time period t on date d; N DR is the number of historical user response data used;

[0017] Then, the maximum and minimum values of the load in the historical response data are used as the fluctuation ranges of the actual load of the demand-side adjustable resources on the operating day, specifically as follows:

[0018]

[0019] In the formula: is the actual response power of each demand-side adjustable resource, and are the upward and downward fluctuation ranges of the load after response of the demand-side adjustable resource i in time period t on the operating day.

[0020] Preferably, a robust model is used to find a solution with good performance for all uncertain inputs under a given conservative parameter of the decision maker to solve the uncertainty problem. The detailed model is divided into the day-ahead bidding stage and the expected response stage.

[0021] Preferably, in the day-ahead bidding stage, first analyze the revenue situation of users participating in demand response under the aggregator. When there is incoordination within the aggregator, the aggregator formulates an internal trading price according to the on-grid electricity price to balance the supply and demand relationship of electricity users; the goal of the aggregator is to control the risk level while minimizing the operating cost;

[0022]

[0023]

[0024] In the formula: is the price at which electricity users purchase electricity from the aggregator, is the price at which electricity users sell electricity to the aggregator, is the price at which the aggregator sells electricity to the market, is the price at which the aggregator purchases electricity from the market, is the price at which electricity users purchase electricity from the aggregator, is the price at which electricity users sell electricity to the aggregator, is the minimum price at which electricity users purchase electricity from the aggregator, is the maximum price at which electricity users purchase electricity from the aggregator, is the minimum price at which electricity users sell electricity to the aggregator, is the maximum price at which electricity users sell electricity to the aggregator, is the average price at which electricity users purchase electricity from the aggregator, is the average price at which electricity users sell electricity to the aggregator, is the amount of electric energy purchased by electricity users from the aggregator, is the amount of electric energy sold by electricity users to the aggregator, is the amount of electric energy sold by the aggregator to the electricity market, is the amount of electric energy purchased by the aggregator from the electricity market, is the amount of electric energy sold by electricity users to the aggregator, π i,w is the moment of PV output, ξ i , η i,w are both auxiliary variables to help calculate CVaR; N is the number of users, Ω is the PV output size, β is the weighting parameter, and the larger the β value, the stronger the aggregator's preference for risk aversion;

[0025] According to the game situation between the aggregator and users, based on the established market clearing price and the response ability of adjustable resources on the demand side, with the goal of maximizing the demand response utility of the aggregator response entity, the optimal adjustment capacity is determined; the objective function in the day-ahead bidding stage is specifically as follows:

[0026]

[0027] In the formula: represents the expected demand response utility of the aggregator response entity in the day-ahead bidding stage; λ com,t is the value obtained by solving the expected response stage model in the current iteration, and its initial value is λ pun,tPenalty cost for the unit's uncompleted regulated power, T A Demand response period, α re Ratio of the upper limit of the effective response power specified in the electricity market to the reported regulation capacity, α pun Ratio of the lower limit of deviation assessment specified by the market to the reported regulation capacity Regulation capacity The day-ahead planned response power of each demand-side adjustable resource; the day-ahead bidding stage model also includes regulation capacity constraints and response capacity constraints of demand-side adjustable resources; among them, Also the value obtained by solving the expected response stage model in the current iteration, and its initial value is Among them, And Represent the economic cost and comfort cost of the demand-side adjustable resource participating in market interaction in the period

[0028] Finally, the obtained regulation capacity Is passed to the expected response stage model

[0029] Preferably, the model in the expected response stage is a two-layer model. The upper-layer model aims to minimize the demand response utility of the aggregated response entity, and obtains the market clearing price and the worst-case scenario of the random fluctuation of the response capacity of the demand-side adjustable resource; the lower-layer model aims to maximize the demand response utility of the aggregated response entity, and decides the optimal demand-side adjustable resource response plan

[0030] The objective function of the expected response stage is specifically as follows

[0031]

[0032] In the formula: Is the actual load of the demand-side adjustable resource in the period Is the actual response load of the demand-side adjustable resource in the period Is the value obtained by solving the expected response stage model in the current iteration

[0033] The expected response stage model includes response capacity constraints of demand-side adjustable resources and fluctuation range constraints of uncertainty variables; specifically as follows

[0034]

[0035] To meet the definition of the worst-case scenario, the following conversion needs to be performed on formula (18)

[0036]

[0037] In the formula: Γcom and is the robustness coefficient for controlling the random fluctuation conservatism of the incomplete information variable; B com,t is the variable for controlling the worst degree of random fluctuation of the market clearing price, and its value range is 0 to 1; L com is the adjustment parameter for controlling the worst degree of the market clearing price on the execution day as a whole;

[0038] Finally, the worst-case scenarios of the market clearing price and the random fluctuation of the demand-side adjustable resource response ability obtained by solving, namely λ com,t and are passed to the day-ahead bidding stage model;

[0039] According to the day-ahead bidding stage model and the expected response stage model, the two-stage robust demand response model of the demand-side adjustable resources is written in the following compact form:

[0040]

[0041] In the formula, B is the variable for controlling the worst degree of the uncertainty variable, and its value range is 0 to 1; A, M, Y, K, G, and H respectively represent the coefficient matrices of the corresponding constraint conditions; this optimization model is divided into 2 layers; the outer max problem is the day-ahead bidding stage model, and the decision variable x is and the inner min-max problem is the expected response stage model, and the decision variable y is the uncertainty variable u is λ com,t and Solving this model to obtain the optimal bidding strategy of the aggregator, namely

[0042] Preferably, the detailed steps of the column constraint generation algorithm C&CG are as follows:

[0043]

[0044] In the formula: l represents the number of the current iteration, k represents the maximum number of iterations, θ represents the lower bound of the objective function; a is the coefficient matrix in the objective function, c T is the transpose of the solution of the master problem, and F, G, M, K, Y respectively represent the coefficient matrices of the corresponding constraint conditions, y l is the solution of the sub-problem after the l-th iteration; u l is the value of the incomplete information variable u in the worst-case scenario after the l-th iteration; the sub-problem is expressed as:

[0045]

[0046] Where: \(U\) is the value space of the uncertainty variable. According to the duality principle, the min problem in the inner layer model of the sub-problem is dual-transformed into a max problem, and then combined with the max model of the outer layer model to obtain the following form:

[0047]

[0048] Where: \(\gamma\) and \(\pi\) are the dual variables of the constraints in the fourth row of the formula respectively; through derivation and transformation, the two-stage robust demand response model is finally transformed into two independent models.

[0049] Preferably, in the upper layer of the two-layer optimization configuration model, phased solution is carried out according to multiple application scenarios; the column constraint generation algorithm C&CG algorithm is used to split the robust optimization model of the shared energy storage into a master problem and a sub-problem, and the iterative solution of the two-stage upper and lower bounds is realized to form a demand response plan under the worst scenario of the clearing price in the electricity market;

[0050] In the lower layer of the two-layer optimization configuration model, the column constraint generation algorithm C&CG algorithm is used to split the robust optimization model of the response plan of the demand-side adjustable resources into a master problem and a sub-problem, and the iterative solution of the two-stage upper and lower bounds is realized to form the response plan for each time period under the worst scenario of the operation of the demand-side users.

[0051] Preferably, the specific steps of the solution are as follows:

[0052] Step 1. Initialization: Set the initial worst-case scenario of the incomplete information variable, the upper bound of the initial objective function is \(UB = +\infty\), the lower bound is \(LB = -\infty\); the number of iterations is \(k = 1\), and the convergence threshold is \(\varepsilon = 0.005\);

[0053] Step 2. Solve the master problem through the worst-case scenario \(u\) l to obtain the optimal solution \((x\) l , \(y\) l ), and update the lower bound of the objective function to \(LB=\theta\) l ;

[0054] Step 3. Based on \(x\) obtained in Step 2 l , solve the sub-problem to obtain the worst-case scenario \(u\) of the incomplete information variable l+1 , and update the new objective value to the upper bound

[0055]

[0056] Step 4. Judge whether it converges: If \(UB - LB\leqslant\varepsilon\), stop the operation and output the day-ahead declaration strategy \(x\) l and the demand response plan under the worst-case scenario Otherwise, add the cut-plane constraint Add it to the main problem; let l = l + 1, and go to solution step 2 to continue the solution until the convergence condition is met.

[0057] Therefore, by adopting the above method for pricing demand-side adjustable resources in an incomplete information game, the present invention can help electricity users form a more reliable pricing scheme and provide a technical solution for the construction of a resource pool.

[0058] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Description of the Drawings

[0059] Figure 1 is a flowchart of an embodiment of the method for pricing demand-side adjustable resources in an incomplete information game according to the present invention. Detailed Embodiment

[0060] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0061] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains.

[0062] Embodiment 1

[0063] As Figure 1 shown, the present invention provides a method for pricing demand-side adjustable resources in an incomplete information game, including the following steps:

[0064] S1. Measuring the fluctuation range of incomplete information variables:

[0065] Based on historical data, predict the clearing price of the electricity market on the operating day and the response ability of the demand-side adjustable resources on the operating day, and calculate the random fluctuation ranges of the above two incomplete information variables to support the incomplete information day-ahead declaration decision of the aggregator.

[0066] 1) Clearing price

[0067] The clearing price of the electricity market can be estimated based on historical clearing prices, as follows:

[0068]

[0069] Where: is the predicted value of the clearing price at time period t; is the clearing price at time period t on date d; N com is the number of historical data used;

[0070] Then, take the maximum and minimum values of the historical clearing prices as the fluctuation ranges of the clearing price on the operating day, respectively, as follows:

[0071]

[0072] Where: λ com,t represents the real-time market clearing price, and are the upward and downward fluctuation ranges of the clearing price in period t;

[0073] 2) Responsive ability of adjustable resources on the demand side

[0074] Under the condition that the baseline load is known, the responsive ability of adjustable resources on the demand side is determined by the actual load on the operating day and can be estimated from historical response data, as follows:

[0075]

[0076] Where: is the predicted value of the actual load of adjustable resource i on the demand side at period t on the operating day; is the load of adjustable resource i on the demand side at period t on date d; N DR is the number of historical response data of users adopted.

[0077] Then, the maximum and minimum values of the load in the historical response data are used as the upward and downward fluctuation ranges of the actual load of adjustable resources on the demand side on the operating day, as follows:

[0078]

[0079] Where: is the actual responsive power of each adjustable resource on the demand side, and are the upward and downward fluctuation ranges of the post-response load of adjustable resource i on the demand side at period t on the operating day.

[0080] S2. Establish an incomplete information game model;

[0081] When aggregators act on behalf of adjustable resources on the demand side to participate in power market interactions, they face incomplete information such as clearing prices and the responsive ability of adjustable resources on the demand side. The incomplete information variables (uncertainty variables) mentioned in S1 usually have "Knightian uncertainty", that is, aggregators have relatively limited knowledge of their "incomplete" characteristics and it is difficult to measure or describe them through methods such as probability distribution functions and membership functions. Therefore, it is impossible to directly use stochastic programming theory for processing. Therefore, the present invention uses a robust model to find a solution with good performance for all uncertain inputs under a given conservative parameter of the decision maker to solve the above uncertainty problem. The detailed model is divided into two stages:

[0082] 1) Day-ahead bidding stage

[0083] First, analyze the benefits of users participating in demand response under the aggregator. As a professional "electricity intermediary", the aggregator has the obligation to manage electricity users. When there is disharmony within the aggregator, the aggregator formulates internal trading prices based on the on-grid electricity price to balance the supply and demand relationship of electricity users. The goal of the aggregator is to control the risk level while minimizing the operating cost.

[0084]

[0085] In the formula: is the price at which electricity users purchase electricity from the aggregator, is the price at which electricity users sell electricity to the aggregator, is the price at which the aggregator sells electricity to the market, is the price at which the aggregator purchases electricity from the market, is the price at which electricity users purchase electricity from the aggregator, is the price at which electricity users sell electricity to the aggregator, is the lowest price at which electricity users purchase electricity from the aggregator, is the highest price at which electricity users purchase electricity from the aggregator, is the lowest price at which electricity users sell electricity to the aggregator, is the highest price at which electricity users sell electricity to the aggregator, is the average price at which electricity users purchase electricity from the aggregator, is the average price at which electricity users sell electricity to the aggregator, is the amount of electricity purchased by electricity users from the aggregator, is the amount of electricity sold by electricity users to the aggregator, is the amount of electricity sold by the aggregator to the electricity market, is the amount of electricity purchased by the aggregator from the electricity market, is the amount of electricity sold by electricity users to the aggregator, π i,w is the moment of photovoltaic output, ξ i and η i,w are both auxiliary variables to help calculate CVaR; N is the number of users, Ω is the size of photovoltaic output, β is the weighting parameter, and the larger the β value, the stronger the aggregator's preference for risk aversion;

[0086] According to the game situation between the aggregator and users, based on the established market clearing price and the response ability of adjustable resources on the demand side, with the goal of maximizing the demand response utility of the aggregation response entity, the optimal adjustment capacity is determined; the specific objective function in the day-ahead bidding stage is as follows:

[0087]

[0088] In the formula: represents the expected demand response utility of the aggregation response entity under the aggregator in the day-ahead bidding stage; λ com,t The value obtained by solving the model for the expected response stage in the current iteration, and its initial value is λ pun,t is the penalty cost per unit of uncompleted regulation electricity, T A is the demand response period, α re is the ratio of the upper limit of the effective response electricity specified by the electricity market to the reported regulation capacity, α pun is the ratio of the lower limit of the deviation assessment specified by the market to the reported regulation capacity is the regulation capacity is the day-ahead planned response electricity of each demand-side adjustable resource; the day-ahead bidding stage model also includes a regulation capacity constraint and a response capacity constraint for demand-side adjustable resources; among them, the in the current iteration is also the value obtained by solving the model for the expected response stage, and its initial value is Among them, and represent the economic cost and comfort cost of the demand-side adjustable resource participating in market interaction in the period

[0089] Finally, the obtained regulation capacity is passed to the model of the expected response stage. It should be noted that the present invention assumes that the declaration strategy of the aggregator will not affect the clearing price of the electricity market

[0090] 2) Expected response stage

[0091] The model of the expected response stage is a two-layer model. The upper-layer model aims to minimize the demand response utility of the aggregation response entity, and obtains the "worst" scenario of the clearing price of the market and the random fluctuation of the response capacity of the demand-side adjustable resources; the lower-layer model aims to maximize the demand response utility of the aggregation response entity, and decides the optimal response plan of the demand-side adjustable resources

[0092] The objective function of the expected response stage is specifically as follows

[0093]

[0094] In the formula is the actual load of the demand-side adjustable resource in the period is the actual response load of the demand-side adjustable resource in the period is the value obtained by solving the model for the expected response stage in the current iteration

[0095] The expected response stage model includes the response capacity constraint of adjustable resources on the demand side and the fluctuation range constraint of uncertainty variables; specifically as follows:

[0096]

[0097] To conform to the definition of the "worst" scenario, the following transformation needs to be performed on formula (18):

[0098]

[0099] In the formula: Γ com and are the robust coefficients for controlling the conservative degree of random fluctuations of incomplete information variables; B com,t is the variable for controlling the worst degree of random fluctuations of the market clearing price, and its value range is 0 to 1; L com is the adjustment parameter for controlling the worst degree of the market clearing price on the execution day as a whole;

[0100] Finally, the obtained market clearing price and the worst scenario of random fluctuations in the response capacity of adjustable resources on the demand side, that is, λ com,t and are passed to the day-ahead bidding stage model;

[0101] According to the above day-ahead bidding stage model and expected response stage model, the two-stage robust demand response model of adjustable resources on the demand side can be written in the following compact form:

[0102]

[0103] In the formula, B is the variable for controlling the worst degree of uncertainty variables, and its value range is 0 to 1; A, M, Y, K, G, and H respectively represent the coefficient matrices of the corresponding constraint conditions; this optimization model is divided into 2 layers; the outer max problem is the day-ahead bidding stage model, and the decision variable x is and The inner min-max problem is the expected response stage model, and the decision variable y is The uncertainty variable u is λ com,t and By solving this model, the optimal bidding strategy of the aggregator can be obtained, that is

[0104] S3. Solve the incomplete information game model;

[0105] For the above model, the column constraint generation algorithm (C&CG) is used for solution. The original problem model is converted into a two-layer optimization problem including the master problem and the sub-problem for processing, and the solution is obtained by alternately iterating the master problem and the sub-problem. The detailed steps are as follows:

[0106]

[0107] In the formula: l represents the number of current iterations, k represents the maximum number of iterations, and θ represents the lower bound of the objective function; a is the coefficient matrix in the objective function, and c T is the transpose of the solution of the master problem, and F, G, M, K, and Y respectively represent the coefficient matrices of the corresponding constraint conditions, and y l is the solution of the sub-problem after the l-th iteration; u l is the value of the incomplete information variable u in the worst-case scenario after the l-th iteration; the sub-problem can be expressed as:

[0108]

[0109] In the formula: U is the value space of the uncertainty variable. According to the duality principle, the min problem of the inner model in the sub-problem can be dual-transformed into a max problem, and then combined with the max model of the outer model, the following form can be obtained:

[0110]

[0111] In the formula: γ and π are the dual variables of the constraints in the fourth row of the formula respectively; through derivation and transformation, the two-stage robust demand response model is finally transformed into two independent models.

[0112] In the upper layer of the two-layer optimization configuration model, it is solved in stages according to multiple application scenarios; the column constraint generation algorithm C&CG algorithm is used to split the robust optimization model of the shared energy storage into a master problem and a sub-problem, and the iterative solution of the two-stage upper and lower bounds is realized to form a demand response plan in the worst-case scenario of the clearing price in the electricity market;

[0113] In the lower layer of the two-layer optimization configuration model, the column constraint generation algorithm C&CG algorithm is used to split the robust optimization model of the response plan of the demand-side adjustable resources into a master problem and a sub-problem, and the iterative solution of the two-stage upper and lower bounds is realized to form the response plan for each time period in the worst-case scenario of the demand-side user operation.

[0114] The specific steps of the solution are as follows:

[0115] Step 1. Initialization: Set the initial worst-case scenario of the incomplete information variable, and the upper and lower bounds of the initial objective function are UB = +∞ and LB = -∞ respectively; the number of iterations is k = 1, and the convergence threshold ε = 0.005;

[0116] Step 2. Solve the master problem through the worst-case scenario u l to obtain the optimal solution (x l , y l ), and update the lower bound of the objective function to LB = θ l ;

[0117] Step 3: Based on the solution of \(x\) in Step 2 l , solve the sub-problem to obtain the worst-case scenario \(u\) of the incomplete information variable l+1 , and update the new objective value to the upper bound

[0118]

[0119] Step 4: Determine whether it converges. If \(UB - LB\leqslant\varepsilon\), stop the operation and output the current declaration strategy \(x\) l and the demand response plan under the worst-case scenario Otherwise, add the cutting plane constraint to the master problem; let \(l = l + 1\), and go back to Step 2 to continue the solution until the convergence condition is met.

[0120] Therefore, the present invention adopts the above-mentioned method for pricing adjustable resources on the demand side with incomplete information game, which can be applied to the construction project of the demand-side resource pool to support the construction and scheduling of the adjustable load resource pool. Considering the participation of multiple market players, it effectively reflects the revenue changes among different players and improves the stability of power transactions.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements do not make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A pricing method for demand-side adjustable resources in an incomplete information game, characterized in that: It includes the following steps: S1. Measurement of the fluctuation range of incomplete information variables: Based on historical data, predict the clearing price and the response ability of demand-side adjustable resources in the electricity market on the operating day, and calculate the random fluctuation range of the clearing price and the response ability of demand-side adjustable resources; S2. Establish an incomplete information game model: When aggregators act as agents for demand-side adjustable resources to participate in electricity market interactions, there is incomplete information, including the clearing price and the response ability of demand-side adjustable resources; S3. Solve the incomplete information game model: Use the column constraint generation algorithm C&CG for solution; Convert the original problem model into a two-layer optimization problem including a master problem and a sub-problem for processing, and use the master problem and the sub-problem to alternately iterate to solve the answer.

2. The demand-side adjustable resource pricing method for incomplete information games according to claim 1, characterized in that: In S1, the clearing price of the electricity market is estimated based on historical clearing prices, specifically as follows: Wherein: is the predicted value of the clearing price at time period t; is the clearing price at time period t on date d; N com is the quantity of historical data adopted; Then, take the maximum and minimum values in the historical clearing prices as the fluctuation range of the clearing price on the operating day, specifically as follows: where: λ com,t represents the real-time market clearing price, and are the upward and downward fluctuation ranges of the clearing price in period t; Under the condition that the baseline load is known, the response ability of demand-side adjustable resources is determined by the actual load on the operating day and estimated from historical response data, specifically as follows: Wherein: is the predicted actual load value of the demand-side adjustable resource i at the operating day period t; is the load of the demand-side adjustable resource i at the period t of the date d; N DR is the number of historical user response data adopted; Then, take the maximum and minimum values of the load in the historical response data as the fluctuation range of the actual load of demand-side adjustable resources on the operating day, specifically as follows: Wherein: is the actual response power of each demand-side adjustable resource, and are the upward and downward fluctuation ranges of the post-response load of the demand-side adjustable resource i at the time period t of the operation day.

3. The demand-side adjustable resource pricing method for incomplete information games according to claim 2, characterized in that: Adopt a robust model to seek a solution with good performance for all uncertain inputs under the given conservative degree parameter of the decision maker, and solve the uncertainty problem. The detailed model is divided into the day-ahead bidding stage and the expected response stage.

4. The demand-side adjustable resource pricing method for incomplete information games according to claim 3, characterized in that: In the day-ahead bidding stage, first analyze the revenue situation of users participating in demand response under the aggregator. When there is incoordination within the aggregator, the aggregator formulates an internal trading price according to the on-grid electricity price to balance the supply and demand relationship of electricity users; the goal of the aggregator is to control the risk level while minimizing the operating cost; Wherein: is the price at which electricity users purchase electricity from the aggregator, is the price at which electricity users sell electricity to the aggregator, is the price at which the aggregator sells electricity to the market, is the price at which the aggregator purchases electricity from the market, is the price at which electricity users purchase electricity from the aggregator, is the price at which electricity users sell electricity to the aggregator, is the minimum price at which electricity users purchase electricity from the aggregator, is the maximum price at which electricity users purchase electricity from the aggregator, is the minimum price at which electricity users sell electricity to the aggregator, is the maximum price at which electricity users sell electricity to the aggregator, is the average price at which electricity users purchase electricity from the aggregator, is the average price at which electricity users sell electricity to the aggregator, is the amount of electric energy purchased by electricity users from the aggregator, is the amount of electric energy sold by electricity users to the aggregator, is the amount of electric energy sold by the aggregator to the electricity market, is the amount of electric energy purchased by the aggregator from the electricity market, is the amount of electric energy sold by electricity users to the aggregator, π i,w is the time of photovoltaic output, ξ i , η i,w are both auxiliary variables to help calculate CVaR; N is the number of users, Ω is the magnitude of photovoltaic output, β is the weighting parameter, and the larger the β value, the stronger the aggregator's preference for risk aversion; According to the game situation between the aggregator and the users, based on the given market clearing price and the response ability of demand-side adjustable resources, with the goal of maximizing the demand response utility of the aggregated response entity, decide the optimal regulation capacity; the objective function of the day-ahead bidding stage is specifically as follows: In the formula: represents the expected demand response utility of the aggregated response entity under the aggregator in the day-ahead bidding stage; λ com,t is the value obtained by solving the model for the expected response stage in the current iteration, and its initial value is λ pun,t is the penalty cost per unit of uncompleted regulated electricity, T A is the demand response period, α re is the ratio of the upper limit of the effective response electricity specified by the electricity market to the reported regulation capacity, α pun is the ratio of the lower limit of the deviation assessment specified by the market to the reported regulation capacity is the regulation capacity is the day-ahead planned response electricity of each demand-side adjustable resource; the day-ahead bidding stage model also includes a regulation capacity constraint and a demand-side adjustable resource response capacity constraint; among them, in the current iteration is also the value obtained by solving the model for the expected response stage, and its initial value is where and represent the economic cost and comfort cost of the demand-side adjustable resource participating in market interaction in the period Finally, transfer the obtained regulation capacity to the expected response phase model.

5. The demand-side adjustable resource pricing method for incomplete information games according to claim 3, characterized in that: The model of the expected response stage is a two-layer model. The upper-layer model aims to minimize the demand response utility of the aggregated response entity to obtain the worst-case scenario of the random fluctuation of the market clearing price and the response ability of demand-side adjustable resources; the lower-layer model aims to maximize the demand response utility of the aggregated response entity to decide the optimal demand-side adjustable resource response plan; The objective function of the expected response stage is specifically as follows: Wherein: is the actual load of the demand-side adjustable resource in the time period, is the actual response load of the demand-side adjustable resource in the time period, is the value obtained by solving the expected response stage model in the current iteration; The model of the expected response stage includes the constraint of the response ability of demand-side adjustable resources and the constraint of the fluctuation range of uncertain variables; specifically as follows: To conform to the definition of the worst-case scenario, the following transformation needs to be performed on formula (18): where: Γ com and are robustness coefficients for controlling the conservative degree of random fluctuations of incomplete information variables; B com,t is a variable for controlling the worst degree of random fluctuations of market clearing prices, and its value range is 0 to 1; L com is an adjustment parameter for controlling the worst degree of market clearing prices on the execution date as a whole; Finally, transmit the worst-case scenario of the market clearing price and the randomly fluctuating response ability of the demand-side adjustable resources obtained by the solution, that is, λ com,t and to the day-ahead bidding stage model; According to the day-ahead bidding stage model and the expected response stage model, write the two-stage robust demand response model of demand-side adjustable resources in the following compact form: where B is a variable that controls the worst degree of the uncertainty variable, and its value range is 0 to 1; A, M, Y, K, G, and H respectively represent the coefficient matrices of the corresponding constraint conditions; this optimization model is divided into two layers; the outer-layer max problem is the model for the day-ahead bidding stage, and the decision variable x is and the inner-layer min-max problem is the model for the expected response stage, and the decision variable y is the uncertainty variable u is λ com,t and By solving this model, the optimal bidding strategy of the aggregator is obtained, that is 6. The demand-side adjustable resource pricing method for incomplete information games according to claim 5, characterized in that: The detailed steps of the column constraint generation algorithm C&CG are as follows: where: \(l\) represents the number of the current iteration, \(k\) represents the maximum number of iterations, and \(\theta\) represents the lower bound of the objective function; \(a\) is the coefficient matrix in the objective function, \(c\) T is the transpose of the solution of the master problem, and \(F\), \(G\), \(M\), \(K\), \(Y\) respectively represent the coefficient matrices of the corresponding constraint conditions, and \(y\) l is the solution of the sub-problem after the \(l\)-th iteration; \(u\) l is the value of the incomplete information variable \(u\) under the worst-case scenario after the \(l\)-th iteration; the sub-problem is expressed as: Where: \(U\) is the value space of the uncertainty variable. According to the duality principle, the min problem in the inner layer model of the sub-problem is dual-transformed into a max problem, and then combined with the max model in the outer layer model to obtain the following form: Where: \(\gamma\) and \(\pi\) are the dual variables of the constraints in the fourth row of the formula respectively; through derivation and transformation, the two-stage robust demand response model is finally transformed into two independent models.

7. A pricing method for demand-side adjustable resources in an incomplete information game according to claim 6, characterized in that: In the upper layer of the two-layer optimal configuration model, phased solutions are carried out according to multiple application scenarios; the column constraint generation algorithm C&CG algorithm is used to split the robust optimization model of the shared energy storage into a master problem and a sub-problem, and the iterative solution of the upper and lower bounds of the two stages is realized to form a demand response plan under the worst-case scenario of the clearing price in the electricity market; In the lower layer of the two-layer optimal configuration model, the column constraint generation algorithm C&CG algorithm is used to split the robust optimization model of the response plan of the demand-side adjustable resources into a master problem and a sub-problem, and the iterative solution of the upper and lower bounds of the two stages is realized to form the response plan for each time period under the worst-case scenario of the operation of the demand-side users.

8. A demand-side adjustable resource pricing method for incomplete information games according to claim 7, characterized in that: The specific steps of the solution are as follows: Step 1. Initialization: Set the initial worst-case scenario of the incomplete information variable, the upper bound of the initial objective function is \(UB = +\infty\), the lower bound is \(LB = -\infty\); the number of iterations is \(k = 1\), and the convergence threshold is \(\varepsilon = 0.005\); Step 2: Solve the main problem through the worst-case scenario u l to obtain the optimal solution (x l , y l ), and update the lower bound of the objective function to LB = θ l ; Step 3: Based on the solution of \(x\) in Step 2 l , solve the sub-problem to obtain the worst-case scenario \(u\) of the incomplete information variable l+1 , and update the new objective value to the upper bound Step 4. Determine whether convergence is achieved: If UB - LB ≤ ε, stop the operation and output the current declaration strategy x l and the demand response plan under the worst-case scenario Otherwise, add the cutting plane constraint to the master problem; let l = l + 1, and go back to Step 2 for further solution until the convergence condition is satisfied.

Images