A flexible load coordinated control method considering electricity sales revenue and user utility

Abstract

The present invention provides a flexible load collaborative control method that takes into account electricity sales revenue and user utility, and belongs to the technical field of demand-side resource optimization control. It solves the problem that the current demand response control method based on electricity price cannot balance electricity sales revenue and user utility, resulting in insufficient peak-shaving capacity and poor flexibility of the power grid. The method comprises the following steps: obtaining the electricity consumption history data of the flexible load of the power user; adopting the improved hierarchical clustering algorithm to classify the load; comprehensively considering the maximization of the interests of the power sales company and the maximization of the utility of the power user, and establishing a two-layer optimization model for demand-side resource collaborative control according to the coupling relationship between the electricity sales price and the electricity consumption; converting the two-layer optimization model for demand-side resource collaborative control into a single-layer mixed integer linear programming model to form a solution method for the two-layer optimization model, and determining the specific control strategy for the flexible load cluster to participate in peak-shaving; the present invention is applied to the demand-side control of the power grid.

Description

Technical Field

[0001] The present invention provides a flexible load collaborative control method taking into account electricity sales revenue and user utility, belonging to the technical field of demand-side resource optimization control. Background Art

[0002] With the large-scale grid connection of new energy sources such as wind power and photovoltaic power, the problem of insufficient peak-shaving capacity of the power system has become increasingly prominent. In order to balance the supply and demand of electricity, alleviate the peak-shaving pressure of the power system, and improve the safety of power system operation, the regulation and control of demand-side flexible resources has become an inevitable trend in the development of the power industry.

[0003] At present, there are three main demand response control methods based on electricity prices: time-of-use electricity prices, dynamic electricity prices, and peak electricity prices. All of them use electricity sales companies as the contact point between users and electricity wholesalers. However, from different perspectives, the ultimate goals are different. For example, from the perspective of the electricity sales company, it will fully consider the electricity sales revenue and the cost of purchasing electricity in the spot market, and dynamically adjust the electricity prices in different periods to maximize the revenue from electricity sales; from the perspective of users, it will transfer as much electricity cost required for heating (cooling) as possible to the low-price period, and try not to affect the comfort level, in order to maximize the user's benefits.

[0004] Therefore, how to establish a two-layer optimization control model for demand-side resources that takes into account electricity sales revenue and user utility, which can not only determine the dynamic electricity price pricing strategy with the goal of maximizing the interests of the electricity sales company, but also determine the user's electricity consumption decision with the goal of maximizing user utility, has become an urgent problem to be solved to improve the safe operation and flexibility of the power grid. Summary of the invention

[0005] In order to solve the problem that the current demand response control method based on electricity price cannot balance the electricity sales revenue and user utility, resulting in insufficient peak-shaving capacity and poor flexibility of the power grid, the present invention proposes a flexible load collaborative control method that takes into account the electricity sales revenue and user utility.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a flexible load collaborative control method considering electricity sales revenue and user utility, comprising the following steps:

[0007] S10: Obtain the electricity consumption history data of the flexible load of the power users, and take each household as a set of load sample data sets;

[0008] S20: Use an improved hierarchical clustering algorithm to classify loads, analyze users’ electricity usage habits and preferences, and divide the classified loads into flexible load clusters with different flexibility;

[0009] S30: Taking into account the maximization of the interests of the power sales company and the maximization of the utility of the power users, a two-layer optimization model for the coordinated control of demand-side resources is established based on the coupling relationship between the power sales price and the power consumption;

[0010] S40: The two-layer optimization model of demand-side resource collaborative control is transformed into a single-layer mixed integer linear programming model, a solution method for the two-layer optimization model is formed, and a specific control strategy for the flexible load cluster to participate in peak regulation is determined.

[0011] The input part of the improved hierarchical clustering algorithm includes N Group load sample data set, target number of clusters K , the initial minimum similarity index of the mergeable clusters d The output part is K The specific steps of improving the hierarchical clustering algorithm are as follows:

[0012] 1) Initialization: Treat each load sample data set as a separate cluster X , then there are N A load sample data set;

[0013] 2) Calculate similarity: Calculate the similarity between clusters through Euclidean distance D , and the minimum similarity index between mergeable clusters d Compare;

[0014] 3) Merge clusters: If D > d , then merge the two most similar clusters; otherwise go to the next step and update the minimum similarity index d , forming an information feedback mechanism;

[0015] 4) Iteration: Update the new cluster center and repeat steps 2) and 3) until the preset target number of clusters is reached. K .

[0016] The two-layer optimization model for collaborative control of demand-side resources includes a power sales layer model and a user layer model.

[0017] The electricity sales model takes into account electricity sales revenue, spot market electricity purchase costs, positive deviation electricity purchase costs and negative deviation electricity purchase costs. The objective function f 1( x , y ) is the expected benefit of maximizing the electricity sales company, and its expression is as follows:

[0018] ;

[0019] In the above formula: x is the decision variable set of the electricity sales company, yis the set of decision variables of electricity users, T is the total time period, i.e. the optimization period; , Indicates time period t The price of electricity sold by the electricity company to customers. Indicates the time period of the power sales company t The electricity purchase price in the spot market, Indicates the time period of the power sales company t The amount of electricity purchased on the spot market; Indicates the time period of electricity users t The positive deviation of electricity consumption, that is, the part of electricity consumption that exceeds the electricity purchased in the spot market, Indicates the time period of electricity users t The negative deviation of electricity is the remaining amount after the electricity purchased in the spot market meets the actual demand of electricity users; , L F,t It is the electricity user period t The flexible load, L R,t It is the electricity user period t The rigid load capacity; It is the time period t The settlement price of positive deviation electricity in the real-time market, It is the time period t The settlement price of negative deviation electricity in the real-time market;

[0020] The constraint condition of the power sales model is the transaction price constraint, which is expressed as follows:

[0021] ;

[0022] ;

[0023] In the above formula: The minimum electricity selling price for the electricity selling company; It is the maximum electricity selling price of the electricity selling company; It is the daily average dynamic electricity sales price.

[0024] The user layer model considers both the electricity purchase cost and the degree of deviation from the comfort zone. The optimization goal is to maximize the user utility. Its objective function is f 2( x , y ) is as follows:

[0025] ;

[0026] In the above formula: x is the decision variable set of the electricity sales company, yis the set of decision variables of electricity users, T is the total time period, i.e. the optimization period, , Indicates time period t The price of electricity sold by the electricity company to customers. , L F,t It is the electricity user period t The flexible load, T COM Indicates the degree to which the indoor temperature deviates from the comfortable temperature zone; It represents the penalty coefficient for deviation from the indoor temperature comfort zone;

[0027] The constraints of the user layer model include user power constraints and temperature deviation constraints, where the expression of user power constraints is as follows:

[0028] ;

[0029] In the above formula: L t,min Indicates time period t The minimum flexible load of electricity users, L t,max Indicates time period t The maximum flexible load of the electricity user;

[0030] The expression of temperature deviation constraint is as follows:

[0031] ;

[0032] in ;

[0033] In the above formula: T R,low and T R,high They represent the lower and upper limits of the indoor temperature comfort zone, T R,t Indicates time period t indoor temperature.

[0034] The user layer model satisfies the third-order discrete state space model, and the expression is as follows:

[0035] ;

[0036] In the above formula: T R,t , T F,t , T W,t are three state variables, among which T R,t Indicates time periodt The indoor temperature T F,t Indicates time period t The floor temperature, T W,t Indicates time period t The temperature in the water tank directly connected to the heat pump; T R,t-1 , T F,t-1 , T W,t-1 Respectively in the time period t -1 Indoor temperature, floor temperature and the temperature in the water tank directly connected to the heat pump in the third-order discrete state space model; T A,t-1 is the outdoor temperature, which is a random disturbance; α , β and γ is the coefficient matrix, all of which are constants;

[0037] The third-order discrete state space model is converted into constraints, which are expressed as follows:

[0038] ;

[0039] but:

[0040] ;

[0041] ;

[0042] .

[0043] The KKT complementary relaxation condition is used to convert the user layer model into the constraints of the power sales layer model, and then a single-layer mixed integer linear programming model of the new power sales layer model is formed. Solving the single-layer mixed integer linear programming model realizes the solution of the double-layer optimization model.

[0044] The objective function of the single-level mixed integer linear programming model is:

[0045] ;

[0046] In the above formula: , , They are T R,t , T F,t , T W,t The initial values ​​of the dual variables of the three state variables, T R,0 is the indoor temperature before the period starts,T F,0 is the floor temperature before the period starts, T W,0 is the temperature in the water tank directly connected to the heat pump before the start of the period, L F,0 is the flexible load of the power user before the start of the period, , , is the dual variable that is transformed into the constraint condition for the third-order discrete state space model. , is the dual variable of the power consumption constraint, , The dual variable of the temperature deviation constraint, L min represents the minimum flexible load of the power user, L max represents the maximum flexible load of the power user, and They represent the additional cost and revenue loss per MW·h of electricity generated by the power sales company when there is a power deviation.

[0047] The constraints of the single-level mixed integer linear programming model are:

[0048] ;

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] ;

[0054] ;

[0055] ;

[0056] ;

[0057] ;

[0058] ;

[0059] .

[0060] The flexible loads include temperature-controlled loads, interruptible loads and electric vehicles.

[0061] The beneficial effects of the present invention compared to the prior art are as follows: the flexible load collaborative control method provided by the present invention can realize the aggregation control of flexible load resources and then participate in the peak-shaving application of the power grid, which is conducive to alleviating the current problem of insufficient peak-shaving capacity of the power grid after the connection of new energy to the grid, and improving the safety and flexibility of the power system.

[0062] In the process of constructing a two-layer optimization model for collaborative control of demand-side resources, the present invention not only determines a dynamic pricing strategy at the power sales company level that aims to maximize the profits of the power sales company, but also considers electricity costs and comfort at the user level, makes decisions on user electricity consumption with the goal of maximizing user utility, and comprehensively considers multiple factors to maximize the interests of both parties.

[0063] The present invention transforms the two-layer optimization model of demand-side resource collaborative control into a single-layer mixed integer linear programming model, provides a solution method for the above two-layer optimization model, determines the specific control strategy for flexible load clusters to participate in peak load regulation, and can provide a reference for power sales companies to formulate market strategies and power users to formulate electricity consumption plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The present invention will be further described below in conjunction with the accompanying drawings:

[0065] Figure 1 is a flow chart of the method of the present invention;

[0066] Figure 2 A structural diagram of a two-layer optimization model for collaborative control of demand-side resources provided by the present invention;

[0067] Figure 3 It is a schematic diagram of the spot electricity price, deviation electricity price, rigid load and ambient temperature in this embodiment of the present invention;

[0068] Figure 4 It is a schematic diagram of the electricity selling price, electricity quantity deviation and spot purchase quantity under the optimal market strategy of the electricity selling company in this embodiment of the present invention;

[0069] Figure 5 It is a schematic diagram of power consumption decisions of different types of flexible loads in different time periods under the optimal utility strategy of power users in an embodiment of the present invention. DETAILED DESCRIPTION

[0070] like Figures 1 to 5 As shown, the present invention provides a flexible load collaborative control method considering electricity sales revenue and user utility, comprising the following steps:

[0071] S10: Obtaining electricity consumption history data of flexible loads such as temperature-controlled loads, interruptible loads, and electric vehicles of power users;

[0072] S20: Use an improved hierarchical clustering algorithm to cluster loads, analyze user electricity usage habits and preferences, and divide the clustered loads into flexible load clusters with different flexibility;

[0073] S30: Taking into account the maximization of the interests of the power sales company and the maximization of the utility of the power users, a two-layer optimization model for the coordinated control of demand-side resources is established based on the coupling relationship between the power sales price and the power consumption;

[0074] S40: The two-layer optimization model of demand-side resource collaborative control is transformed into a single-layer mixed integer linear programming model, a solution method for the two-layer optimization model is formed, and a specific control strategy for the flexible load cluster to participate in peak regulation is determined.

[0075] The improved hierarchical clustering algorithm can introduce a feedback mechanism with class information; the input part of the improved hierarchical clustering algorithm includes N Group load sample data set, target number of clusters K , the initial minimum similarity index of the mergeable clusters d , the output part is the number of target clusters K ,Right now K The specific steps of improving the hierarchical clustering algorithm are as follows:

[0076] 1) Initialization: Treat each load sample data set as a separate cluster X , then there are N A load sample data set;

[0077] 2) Calculate similarity: Calculate the similarity between clusters through Euclidean distance D , and the minimum similarity index between mergeable clusters d Compare;

[0078] 3) Merge clusters: If D > d , then merge the two most similar clusters; otherwise go to the next step and update the minimum similarity index d , forming an information feedback mechanism;

[0079] 4) Iteration: Update the new cluster center and repeat steps 2) and 3) until the preset target number of clusters is reached. K .

[0080] Among them, the minimum similarity index d The absolute value of cosine similarity is used, and the calculation formula is as follows:

[0081] (1);

[0082] In the above formula: X (i) For thei The average value of the variable in each cluster, X (j) For the j The mean value of the variable for each cluster.

[0083] The two-layer optimization model for collaborative control of demand-side resources includes a power sales layer model and a user layer model.

[0084] The electricity sales model considers electricity sales revenue, spot market electricity purchase costs, positive deviation electricity purchase costs and negative deviation electricity purchase costs. The objective function f 1( x , y ) is the expected benefit of maximizing the electricity sales company, and its expression is as follows:

[0085] (2);

[0086] In the above formula: x is the decision variable set of the electricity sales company, y is the set of decision variables of electricity users, T is the total time period, i.e. the optimization period; , Indicates time period t The price of electricity sold by the electricity company to customers. Indicates the time period of the power sales company t The electricity purchase price in the spot market, Indicates the time period of the power sales company t The amount of electricity purchased on the spot market; Indicates the time period of electricity users t The positive deviation of electricity consumption, that is, the part of electricity consumption that exceeds the electricity purchased in the spot market, Indicates the time period of electricity users t The negative deviation of electricity is the remaining amount after the electricity purchased in the spot market meets the actual demand of electricity users; , L F,t It is the electricity user period t The flexible load, L R,t It is the electricity user period t The rigid load capacity; It is the time period t The settlement price of positive deviation electricity in the real-time market, It is the time period t The settlement price of negative deviation electricity in the real-time market;

[0087] The transaction price constraints of the power sales model are as follows:

[0088] (3);

[0089] (4);

[0090] In the above formula: The minimum electricity selling price for the electricity selling company; It is the maximum electricity selling price of the electricity selling company; It is the daily average dynamic electricity sales price.

[0091] The user layer model considers both the electricity procurement cost and the degree of deviation from the comfort zone. The optimization goal is to maximize the user utility. The objective function is as follows:

[0092] (5);

[0093] In the above formula: x is the decision variable set of the electricity sales company, y is the set of decision variables of electricity users, T is the total time period, i.e. the optimization period, , Indicates time period t The price of electricity sold by the electricity company to customers. , L F,t It is the electricity user period t The flexible load, T COM Indicates the degree to which the indoor temperature deviates from the comfortable temperature zone; represents the penalty coefficient for deviation from the indoor temperature comfort zone; and the user layer model satisfies the third-order discrete state space model, as follows:

[0094] (6);

[0095] In the above formula: T R,t , T F,t , T W,t are three state variables, among which T R,t Indicates time period t The indoor temperature T F,t Indicates time period t The floor temperature, T W,t Indicates time period t The temperature in the water tank directly connected to the heat pump; T R,t-1 , T F,t-1 , T W,t-1 Respectively in the time period t-1 Indoor temperature, floor temperature and the temperature in the water tank directly connected to the heat pump in the third-order discrete state space model; T A,t-1 is the outdoor temperature, which is a random disturbance; α , β and γ is a coefficient matrix, all of which are constants.

[0096] The user power constraints of the user layer model are as follows:

[0097] (7);

[0098] In the above formula: L t,min Indicates time period t The minimum flexible load of electricity users, L t,max Indicates time period t The maximum flexible load of an electricity user.

[0099] If the user's indoor temperature is higher than the upper limit of the comfortable temperature or lower than the lower limit of the comfortable temperature, they will be penalized; therefore, the temperature deviation of the user layer model must meet certain constraints, as follows:

[0100] (8);

[0101] (9);

[0102] In the above formula: T R,low and T R,high They represent the lower and upper limits of the indoor temperature comfort zone respectively.

[0103] Among them, the third-order discrete state space model can be transformed into constraints as follows:

[0104] (10);

[0105] Where:

[0106] (11);

[0107] (12);

[0108] (13).

[0109] The solution method of the two-layer optimization model can use the KKT complementary relaxation condition to convert the user layer model into the constraints of the power sales layer model, and then form a new single-layer mixed integer linear programming model of the power sales layer model.

[0110] Among them, the dual variable of formula (7) is , ; The dual variable of formula (8) is , ; The dual variable of formula (10) is recorded as , , , the KKT conditions are as follows:

[0111] (14);

[0112] (15);

[0113] (16);

[0114] (17);

[0115] (18);

[0116] (19);

[0117] (20);

[0118] (twenty one);

[0119] (twenty two);

[0120] In the above formula: L t,min Indicates time period t The minimum flexible load of electricity users, L t,max Indicates time period t The maximum flexible load of an electricity user.

[0121] The objective function and constraints of the new electricity sales model are as follows:

[0122] (twenty three);

[0123] (twenty four);

[0124] In the above formula: and Respectively indicate time periods t The additional cost and revenue loss per MW·h of electricity generated by the power sales company when there is a power deviation is expressed as follows:

[0125] (25);

[0126] (26).

[0127] Combined with the KKT condition constraints, the objective function of the single-layer mixed integer linear programming model of the new power sales model is:

[0128] (27);

[0129] In the above formula: , , They are T R,t , T F,t , T W,t The initial values ​​of the dual variables of the three state variables, T R,0 is the indoor temperature before the period starts, T F,0 is the floor temperature before the period starts, T W,0 is the temperature in the water tank directly connected to the heat pump before the start of the period, L F,0 It is the flexible load of the power user before the start of the period.

[0130] In summary, the two-level optimization model of demand-side resource collaborative control is transformed into a single-level mixed integer linear programming model. The objective function is equation (27), and the constraints are equations (14) to (22) and (24) to (26).

[0131] The present invention will be further described below based on specific embodiments.

[0132] In this embodiment, after clustering the power user samples using the improved hierarchical clustering algorithm, the power users can be divided into three groups, each group contains a certain number of temperature-controlled loads, interruptible loads, electric vehicles and other flexible loads, and each group has similar behaviors in terms of heating preferences, charging habits, etc.

[0133] Among them, the flexible load ratios of Group 1, Group 2, and Group 3 were 30%, 40%, and 30%, respectively.

[0134] Taking the temperature control load as an example, the specific comfortable temperature zone of each group of users is shown in Table 1 (in °C).

[0135]

[0136] Table 1.

[0137] The parameters of the third-order discrete state space model in this embodiment are:

[0138] ;

[0139] ;

[0140] .

[0141] In this embodiment, the user's minimum flexible load L t,min 0kW·h, the user's maximum flexible load L t,max is 0.33kW·h; spot electricity price, deviation electricity price, rigid load, ambient temperature, etc. Figure 3 shown.

[0142] The remaining parameters in this embodiment are set as follows: Optimization cycle T The lowest electricity price of the electricity sales company for 24 hours The maximum electricity price of the power sales company is 0.34 yuan / kW·h. The daily average dynamic electricity price is 1.2 yuan / kW·h. It is 0.76 yuan / kW·h.

[0143] By solving the single-layer mixed integer linear programming model, the optimal market strategy of the power sales company and the corresponding strategy of the power user in this embodiment can be obtained as follows: Figure 4 and Figure 5 shown.

[0144] from Figure 4 and Figure 5 It can be seen that the power sales company provides lower electricity prices to power users in the 1-8h period and the 21-24h period. For their own cost considerations, users prefer to purchase flexible loads during the lower price period. In addition, the indoor temperature in the 1-8h period and the 21-24h period is relatively low. In order to reduce the deviation from the comfort zone, sufficient load is required to control the temperature in these two time periods. Comparing the power consumption strategies of the three groups, it can be seen that the users in Groups 2 and 3 have lower flexibility. In the 11-17h period when the electricity price is higher, they still choose to purchase flexible loads out of necessity.

[0145] In addition, the total amount of electricity sold by the power sales company to customers has a similar trend to the amount of electricity purchased by the power sales company in the spot market, but because the settlement price of positive deviation electricity is higher than the clearing price of negative deviation electricity, the amount of electricity purchased in the spot market has increased to a certain extent. In order to reduce the cost of purchasing electricity in the spot market, the amount of electricity purchased in the spot market reaches its peak in the 1-6 period.

[0146] When the total number of power users in the three groups remains unchanged, the proportion of each group of users is changed to study the impact of different user flexibility on the optimization results. The user proportions of the first, second and third groups are configured at 30%, 40% and 40%, respectively, and it is considered that the flexibility of the system power users is moderate; when the user proportions of the first, second and third groups are set to 60%, 30% and 10% respectively, the flexibility of the system power users is the highest; when the user proportions of the first, second and third groups are set to 10%, 30% and 60% respectively, the flexibility of the system users is the lowest.

[0147] Under three different user flexibility levels of high flexibility, medium flexibility and low flexibility, the benefits of power sales companies and the utility distribution of users are shown in Table 2 (in ten thousand yuan).

[0148]

[0149] Table 2.

[0150] It can be seen from Table 2 that under the two-layer optimization model of demand-side resource collaborative control, the higher the user's flexibility, the lower the total revenue of the power sales company.

[0151] For power sales companies, the increase in user flexibility will reduce the company's electricity purchase costs in the spot market and real-time market settlement costs. However, due to the increase in user flexibility, the company's electricity sales revenue will also be significantly reduced. Moreover, since the reduction in electricity sales revenue is greater than the reduction in spot market and real-time market costs, the company's total benefits are still decreasing.

[0152] For electricity users, the increase in flexibility leads to a significant reduction in the electricity purchase cost of flexible loads and a slight increase in the cost of rigid loads, resulting in a reduction in the total electricity purchase cost. Therefore, appropriately increasing the flexibility of electricity demand can reduce the electricity purchase cost of consumers themselves.

[0153] In summary, the following conclusions can be drawn from this embodiment:

[0154] Under the two-level optimization model of demand-side resource collaborative control, electricity users will change their electricity consumption characteristics according to the price signals given by the electricity sales company, and prefer to purchase flexible loads when the electricity sales price is low;

[0155] Power sales companies can improve their overall efficiency by increasing power sales revenue, reducing spot market power purchase costs and real-time market settlement costs, but this may also increase power users' power purchase costs to a certain extent.

[0156] The high flexibility of user combinations is not only conducive to peak load shifting, but also reduces the electricity purchase cost of users' flexible loads.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A flexible load collaborative control method considering electricity sales revenue and user utility, characterized by: The following steps are involved: S10: Obtain the electricity consumption history data of the flexible load of the power users, and take each household as a set of load sample data sets; S20: Use an improved hierarchical clustering algorithm to classify loads, analyze users’ electricity usage habits and preferences, and divide the classified loads into flexible load clusters with different flexibility; S30: Taking into account the maximization of the interests of the power sales company and the maximization of the utility of the power users, a two-layer optimization model for the coordinated control of demand-side resources is established based on the coupling relationship between the power sales price and the power consumption; The two-layer optimization model for collaborative control of demand-side resources includes a power sales layer model and a user layer model; The electricity sales model takes into account electricity sales revenue, spot market electricity purchase costs, positive deviation electricity purchase costs and negative deviation electricity purchase costs. The objective function f 1( x , y ) is the expected benefit of maximizing the electricity sales company, and its expression is as follows: ； In the above formula: x is the decision variable set of the electricity sales company, y is the set of decision variables of electricity users, T is the total time period, i.e. the optimization period; , Indicates time period t The price of electricity sold by the electricity sales company to customers. Indicates the time period of the power sales company t The electricity purchase price in the spot market, Indicates the time period of the electricity sales company t The amount of electricity purchased on the spot market; Indicates the time period of electricity users t The positive deviation of electricity consumption, that is, the part of electricity consumption that exceeds the electricity purchased in the spot market, Indicates the time period of electricity users t The negative deviation of electricity is the remaining amount after the electricity purchased in the spot market meets the actual demand of electricity users; , L F,t It is the electricity user period t The flexible load, L R,t It is the electricity user period t The rigid load capacity; It is the time period t The settlement price of positive deviation electricity in the real-time market, It is the time period t The settlement price of negative deviation electricity in the real-time market; The constraint condition of the power sales model is the transaction price constraint, which is expressed as follows: ； ； In the above formula: The minimum electricity selling price for the electricity selling company; It is the maximum electricity selling price of the electricity selling company; The daily average dynamic electricity price; S40: The two-layer optimization model of demand-side resource collaborative control is transformed into a single-layer mixed integer linear programming model, a solution method for the two-layer optimization model is formed, and a specific control strategy for the flexible load cluster to participate in peak regulation is determined.

2. According to claim 1, a flexible load collaborative control method considering electricity sales revenue and user utility is characterized by: The input part of the improved hierarchical clustering algorithm includes N Group load sample data set, target number of clusters K , the initial minimum similarity index of the mergeable clusters d The output part is K The specific steps of improving the hierarchical clustering algorithm are as follows: 1) Initialization: Treat each load sample data set as a separate cluster X , then there are N A load sample data set; 2) Calculate similarity: Calculate the similarity between clusters through Euclidean distance D , and the minimum similarity index between mergeable clusters d Compare; 3) Merge clusters: If D > d , then merge the two most similar clusters; Otherwise, go to the next step and update the minimum similarity index d , forming an information feedback mechanism; 4) Iteration: Update the new cluster center and repeat steps 2) and 3) until the preset target number of clusters is reached. K .

3. According to claim 1, a flexible load collaborative control method considering electricity sales revenue and user utility is characterized by: The user layer model considers both the electricity purchase cost and the degree of deviation from the comfort zone. The optimization goal is to maximize the user utility. Its objective function is f 2( x , y ) is as follows: ； In the above formula: x is the decision variable set of the electricity sales company, y is the set of decision variables of electricity users, T is the total time period, i.e. the optimization period, , Indicates time period t The price of electricity sold by the electricity sales company to customers. , L F,t It is the electricity user period t The flexible load, T COM Indicates the degree to which the indoor temperature deviates from the comfortable temperature zone; It represents the penalty coefficient for deviation from the indoor temperature comfort zone; The constraints of the user layer model include user power constraints and temperature deviation constraints, where the expression of user power constraints is as follows: ； In the above formula: L t,min Indicates time period t The minimum flexible load of electricity users, L t,max Indicates time period t The maximum flexible load of the electricity user; The expression of temperature deviation constraint is as follows: ； in ; In the above formula: T R,low and T R,high They represent the lower and upper limits of the indoor temperature comfort zone, T R,t Indicates time period t indoor temperature.

4. A flexible load collaborative control method considering electricity sales revenue and user utility according to claim 3, characterized in that: The user layer model satisfies the third-order discrete state space model, and the expression is as follows: ； In the above formula: T R,t , T F,t , T W,t are three state variables, among which T R,t Indicates time period t The indoor temperature T F,t Indicates time period t The floor temperature, T W,t Indicates time period t The temperature in the water tank directly connected to the heat pump; T R,t-1 , T F,t-1 , T W,t-1 Respectively in the time period t -1 Indoor temperature, floor temperature and the temperature in the water tank directly connected to the heat pump in the third-order discrete state space model; T A,t-1 is the outdoor temperature, which is a random disturbance; α , β and γ is the coefficient matrix, all of which are constants; The third-order discrete state space model is converted into constraints, which are expressed as follows: ； but: ； ； 。 5. A flexible load collaborative control method considering electricity sales revenue and user utility according to claim 4, characterized in that: The KKT complementary relaxation condition is used to convert the user layer model into the constraints of the power sales layer model, and then a single-layer mixed integer linear programming model of the new power sales layer model is formed. Solving the single-layer mixed integer linear programming model realizes the solution of the double-layer optimization model.

6. A flexible load collaborative control method considering electricity sales revenue and user utility according to claim 5, characterized in that: The objective function of the single-level mixed integer linear programming model is: ； In the above formula: , , They are T R,t , T F,t , T W,t The initial values ​​of the dual variables of the three state variables, T R,0 is the indoor temperature before the period starts, T F,0 is the floor temperature before the period starts, T W,0 is the temperature in the water tank directly connected to the heat pump before the start of the period, L F,0 is the flexible load of the power user before the start of the period, , , is the dual variable that is transformed into the constraint condition for the third-order discrete state space model. , is the dual variable of the power consumption constraint, , The dual variable of the temperature deviation constraint, L min represents the minimum flexible load of the power user, L max represents the maximum flexible load of the power user, and They represent the additional cost and revenue loss per MW·h of electricity generated by the power sales company when there is a power deviation.

7. A flexible load collaborative control method considering electricity sales revenue and user utility according to claim 6, characterized in that: The constraints of the single-level mixed integer linear programming model are: ； ； ； ； ； ； ； ； ； ； ； 。 8. A flexible load collaborative control method considering electricity sales revenue and user utility according to any one of claims 1 to 7, characterized in that: The flexible loads include temperature-controlled loads, interruptible loads and electric vehicles.

Images