Power distribution network demand side resource cluster optimization control method, system and device based on load directrix, and storage medium

Abstract

The invention discloses a load directrix-based power distribution network demand side resource cluster optimization control method, system and device, and a storage medium, and the method comprises the steps: constructing a load directrix calculation model, analyzing the uncertainty of new energy output during the construction of the load directrix calculation model through a Wasserstein distance-based distribution robust optimization method, and determining the load directrix calculation model. Obtaining a new energy output uncertainty fuzzy set; in combination with the new energy output uncertainty fuzzy set, reconstructing the load directrix calculation model to obtain a load calculation model considering the new energy output uncertainty; and accurately adjusting the non-adjustable load curve based on the adjustable power domain of the demand side resource cluster. According to the invention, the problem of how to participate in quasi-linear demand response by a large-scale demand side resource cluster can be solved.

Description

Technical Field

[0001] The present invention relates to a method for cluster optimization control of demand-side resources in a distribution network, and in particular to a method, system, device and storage medium for cluster optimization control of demand-side resources in a distribution network based on a load criterion. Background Art

[0002] With the large-scale access of new energy to the distribution network, the adjustable capacity of demand-side resource clusters has become the key to improving the flexibility of the distribution network and the ability to absorb new energy. Demand response is a key means of adjusting the flexibility of the distribution network and optimizing the dispatch of demand-side resources. However, the existing demand response mechanism is difficult to track the fluctuations of new energy in real time and is not suitable for distribution systems with a high proportion of new energy. For this reason, some scholars have proposed load guidelines and quasi-linear demand response ideas, but the current model does not take into account the prediction error of new energy. When calculating the load guidelines, the uncertainty of new energy output must be fully considered to ensure the accuracy and effectiveness of demand response.

[0003] In the dispatching of distribution networks, demand-side resources report their power adjustable range to the distribution network, and the distribution network operation control center optimizes the dispatching according to the reported adjustment capacity. Therefore, accurately quantifying the power adjustment capacity of demand-side resources becomes the key to their participation in the dispatching of distribution networks. Due to the characteristics of demand-side resources such as a wide variety, large number, small single capacity and geographical dispersion, the operating status and control methods of different adjustable resources are different, so it is necessary to consider the different operating characteristics of different types of equipment. However, if each adjustable resource is modeled and quantitatively analyzed separately, a large number of variables and constraints will be brought about, and as the types, quantities and dispatching periods of demand-side resources continue to increase, it will bring a huge burden on calculation and communication. Therefore, it is necessary to aggregate the adjustable capacity of demand-side resources and build a general model that can reflect the adjustable capacity of aggregated resources, which will help manage and utilize demand-side resources and maximize the value of demand-side resources. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a method, system, device and storage medium for optimizing and controlling a distribution network demand-side resource cluster based on a load quasi-line, which can solve the problem of how large-scale demand-side resource clusters participate in quasi-linear demand response.

[0005] Technical solution: A method for optimizing and controlling a resource cluster on the demand side of a distribution network based on a load criterion line of the present invention comprises:

[0006] The load criterion calculation model is constructed. Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of renewable energy output is analyzed when the load criterion calculation model is constructed, and the uncertainty fuzzy set of renewable energy output is obtained.

[0007] Combined with the uncertainty fuzzy set of renewable energy output, the load criterion calculation model is reconstructed to obtain the load calculation model considering the uncertainty of renewable energy output.

[0008] According to the calculation results of the load quasi-line calculation model considering the uncertainty of renewable energy output, a load quasi-line demand-side resource cluster optimization control strategy considering the uncertainty of renewable energy output is proposed. The power grid dispatching center accurately adjusts the non-adjustable load curve according to the adjustable power domain information of the demand-side resource cluster.

[0009] Furthermore, a load criterion calculation model is constructed, including:

[0010] First, define the load curve shape. Let l(t) be a load curve whose total power in time period T is α D ; with α D The load curve l(t) is normalized as a reference, that is:

[0011]

[0012] l*(t) represents the normalized load value;

[0013] Reconstruct the load line calculation model, including:

[0014] With the goal of minimizing operating costs, the objective function of the load criterion calculation model is established:

[0015]

[0016] Where N G Indicates the number of controllable generator sets; N R Indicates the number of new energy generating units; P G，i (t) represents the active output of the i-th controllable generator set in period t; P R，j (t) represents the output of the jth renewable energy generator set in period t; represents the maximum output of the jth renewable energy generator set in period t; C R represents the energy abandonment cost of the j-th renewable energy generator set; f represents the total operating cost; a i 、b i 、c i Represents the operating cost coefficient of the controllable generator set.

[0017] Set the following constraints:

[0018] Power balance constraints:

[0019] The non-adjustable resources including new energy and rigid loads form a fluctuating curve, and the fluctuation is balanced by the controllable generators and the adjustable loads on the demand side, as follows:

[0020]

[0021] Where P D (t) represents the power consumption of the adjustable load participating in demand response in the system; P C (t) represents the power consumption of rigid loads in the system that do not participate in demand response; Indicates P D (t) is the normalized value; A D It represents the total power consumption of the load participating in the demand response during the T period;

[0022] Directrix Constraints:

[0023]

[0024]

[0025] Output constraints of controllable conventional units:

[0026]

[0027] In the formula, It represents the upper and lower bounds of the active output of the ith controllable generator set in period t;

[0028] Unit climbing constraints:

[0029] -R D,i (t)≤P G,i (t)-P G,i (t-1)≤R U,i (t),t=1,2,…,T

[0030] In the formula, R D,i (t), R U,i (t) represents the maximum downward and upward climbing rate of the i-th controllable generator set in time period t;

[0031] New energy output constraints:

[0032]

[0033] In the formula, It represents the maximum output of the j-th renewable energy generator set in period t.

[0034] Furthermore, based on the distributed robust optimization method of Wasserstein distance, the uncertainty of renewable energy output when constructing the load criterion calculation model is analyzed, and the uncertainty fuzzy set of renewable energy output is obtained, including:

[0035] Define the uncertain variable of new energy output prediction as l, then:

[0036]

[0037] In the formula, Indicates the actual maximum output and predicted maximum output of the j-th renewable energy unit at time t;

[0038] M historical data {l1,l2,…,l M} is uniformly distributed as P e , P e is the empirical distribution, and the Wasserstein distance is used to measure the true probability distribution P and P e The distance between them is used to construct the distributed robust fuzzy set:

[0039]

[0040] In the formula, l and σ represent the distribution P e , P random variables; Π represents the joint distribution of l and σ; ||·|| represents the 1-norm; inf represents the infimum; dl, dσ represent differentials; l1, l2, ... l M M groups of sample data representing uncertain variables;

[0041] W(P e ,P) represents the distance between the true probability distribution and the empirical distribution, namely the Wasserstein distance;

[0042] New energy output uncertainty set P N It is expressed as:

[0043] P N = {P∈P(Ξ):W(P e ,P)≤ε(M)}

[0044] Where Ξ represents the support set of random variables; P(Ξ) represents the set of all probability distributions with Ξ as the support set of random variables; ε(M) represents the radius of the probability distribution.

[0045] Furthermore, combined with the uncertainty fuzzy set of new energy output, the load criterion calculation model is reconstructed to obtain the load calculation model considering the uncertainty of new energy output, including:

[0046] Reconstruct the objective function of the load alignment calculation model:

[0047] After considering the new energy prediction error, the objective function of the load criterion calculation model is adjusted to:

[0048]

[0049] Where, f represents the total operating cost; a i 、bi 、c i Represents the operating cost coefficient of the controllable generator set; N G Indicates the number of controllable generator sets; P G,i (t) represents the active output of the i-th controllable generator set in period t; P R,j (t) represents the output of the jth renewable energy generator set in period t; represents the actual maximum output of the j-th renewable energy unit at time t; C R N represents the energy abandonment cost of the j-th renewable energy generator set; R Indicates the number of renewable energy generating units.

[0050] The above objective function is simplified as follows:

[0051]

[0052] Where x represents each decision variable; X represents the set of decision variables x; f(x) is the deterministic part of the objective function and has nothing to do with the uncertainty variable l; E represents the expected cost under the worst case uncertainty. P (·) represents the expected probability; sup represents the supremum function; P represents the probability distribution of uncertain variables;

[0053] Combined with the uncertainty of new energy output fuzzy set P N , based on the strong duality theory Reconstruct it into the following formula to reduce the difficulty of model calculation:

[0054]

[0055] Where ε(M) represents the radius of the probability distribution; l represents the random variable that obeys the distribution P; Ξ represents the support set of the random variable; M represents the historical data {l1,l2,…,l M}; l k represents the kth group of samples; λ represents the dual variable.

[0056] Introduce auxiliary variable s k , the above formula can be further rewritten as:

[0057]

[0058]

[0059] And because h(x,l)-λ||ll k || is a convex function on the interval of the random variable k, so its supremum is obtained at the interval boundary:

[0060]

[0061] sth(x,l)+λ( l -l k )≤s k

[0062]

[0063] h(x,k)≤s k

[0064] In the formula, l , represents the upper and lower bounds of uncertain variables; h(x,l) represents the loss function.

[0065] Therefore, the final model is transformed into:

[0066]

[0067] stAx≤b

[0068] h(x, l )+λ( l -l k )≤s k

[0069]

[0070] h(x,l)≤s k

[0071] Among them, Ax≤b represents the deterministic constraint condition, A and b are the coefficient matrix and parameter vector of the constraint condition respectively;

[0072] Opportunity Constraint Construction:

[0073] The opportunity constraint method is used to model the inequality constraints containing uncertain variables of renewable energy output, so that the inequality constraints containing the probability distribution P of uncertain variables are modeled in the fuzzy set P of renewable energy output uncertainty. N The expected probability E P Greater than the minimum allowed confidence level, that is, 1-γ, where γ is the risk factor for the inequality constraint violation, as follows:

[0074]

[0075] Furthermore, based on the adjustable power domain of the demand-side resource cluster, the non-adjustable load curve is accurately adjusted, including:

[0076] On the day-ahead, the forecast data of new energy, the adjustable power domain of demand-side resources and the published load quasi-line are comprehensively considered, and the overall benefit is maximized to build a day-ahead optimization model for the demand-side resource cluster to participate in quasi-linear demand response.

[0077] During the day, after receiving the cluster control curve and expected benefits, users build a daily decomposition optimization model for decomposing cluster electricity consumption into various adjustable loads based on their own load control costs, benefit distribution, and load operation requirements, with the goal of optimizing the economic efficiency of adjustable load i.

[0078] Furthermore, the construction process of the day-ahead optimization model is as follows:

[0079] Determine the objective function of the day-ahead optimization model:

[0080]

[0081] Among them, F buy represents the cost of purchasing energy; F cost represents the adjustment cost of the demand-side resource cluster; F sell Indicates the benefit of electricity sales; F reward It represents the incentive compensation obtained by the demand-side resource cluster participating in the quasi-linear demand response; the cost of purchasing energy includes the cost of purchasing electricity from the power grid company, wind power and photovoltaic power, and its expression is as follows:

[0082]

[0083] In the formula, λ t represents the price of electricity purchased from the power grid company; WT represents the purchase price of wind power; PV P represents the purchase price of photovoltaic power generation; t buy represents the amount of electricity purchased by demand-side resources during period t; P t WT , P t PV It represents the wind power and photovoltaic power consumption in period t; Δt represents the length of the time interval.

[0084] The revenue from electricity sales includes:

[0085]

[0086] In the formula, F sell Represents the revenue from electricity sales; S t P represents the electricity price of the demand-side resource cluster participating in the response; t E represents the power consumption of demand-side resources; r t represents other benefits at time t;

[0087] Set day-ahead constraints, including:

[0088] Power balance constraints:

[0089] P t buy +P t PV +P t WT =P t E +r t

[0090] Wind power and photovoltaic power generation power constraints:

[0091] The following formula indicates that the wind and photovoltaic power generation should be less than the wind and photovoltaic predicted value on the day before;

[0092] 0≤P P tV ≤P t PV0

[0093] 0≤P t WT ≤P t WT0

[0094] Where P t WT0 , P t PV0 Indicates the day-ahead forecast value of wind power and photovoltaic power generation;

[0095] Demand side adjustable load constraints:

[0096] P t E +r t ∈Ω agg，t

[0097] In the formula, Ω agg，t Indicates the adjustable power domain of the demand-side resource cluster.

[0098] Furthermore, the construction process of the intraday decomposition optimization model is as follows:

[0099] Considering that there are n adjustable loads in the demand-side resource cluster, the decomposition strategy needs to meet the following conditions:

[0100] Load operation requirements:

[0101] The power consumption of each adjustable load i and the adjustment amount r of the adjustable load i participating in the response i，t The sum is within its adjustable power domain to ensure the feasibility of regulation. The expression is as follows:

[0102]

[0103]

[0104] In the formula, represents the power consumption of adjustable load i; r i，t Indicates the adjustment amount of the adjustable load i participating in the response; Ω i，t represents the adjustable power domain of adjustable load i.

[0105] Control costs:

[0106]

[0107] In the formula, f i，cost represents the regulation cost of adjustable load i; β i represents the control cost coefficient of adjustable load i;

[0108] Profit distribution:

[0109] The improved Shapley method is used to distribute benefits within the adjustable load cluster. The benefit distribution is optimized according to the importance of different types of adjustable loads. The greater the importance of adjustable load i, the more benefits can be allocated, as follows:

[0110]

[0111] Δ i (S) = I (S) - I (S - {i})

[0112] I i =∑W(|S|)Δ i (S)

[0113] Where S represents any alliance with adjustable resource i, and Δ i (S) represents the marginal benefit contribution of adjustable load i to alliance S; W(|S|) represents the weighting factor; I i represents the profit distribution function of adjustable load i in alliance N; |S| represents the number of stakeholder alliances; (|S|-1)! represents the ranking number of a member when participating in alliance S; n represents the total number of members in the alliance; n! represents the factorial of n;

[0114] The benefit R of the adjustable load i i for

[0115]

[0116] In the formula, represents the importance factor of adjustable load i. The larger the importance factor, the greater the benefit allocated. reward It represents the incentive compensation obtained by the demand-side resource cluster participating in the quasi-linear demand response;

[0117] Therefore, with the purpose of optimizing the economic efficiency of adjustable load i, the intra-day decomposition optimization model is as follows:

[0118] maxF i =R i -f i,cost .

[0119] Based on the same inventive concept, a distribution network demand side resource cluster optimization control system based on a load criterion line of the present invention comprises:

[0120] The fuzzy set acquisition module is used to construct the load criterion calculation model. Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of new energy output when constructing the load criterion calculation model is analyzed to obtain the uncertainty fuzzy set of new energy output;

[0121] The load calculation model construction module is used to reconstruct the load criterion calculation model by combining the uncertainty fuzzy set of new energy output, and obtain the load calculation model considering the uncertainty of new energy output;

[0122] The load curve adjustment module is used to accurately adjust the non-adjustable load curve based on the adjustable power domain of the demand-side resource cluster.

[0123] Based on the same inventive concept, a distribution network demand side resource cluster optimization control device based on load standard line of the present invention includes a processor and a memory, the memory stores computer instructions, the processor is used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the electronic device implements the steps of the distribution network demand side resource cluster optimization control method based on load standard line as mentioned above.

[0124] Based on the same inventive concept, a computer-readable storage medium of the present invention stores a computer program thereon, and when the program is executed by a processor, the steps of the above-mentioned method for optimizing and controlling resource clusters on the demand side of a distribution network based on a load standard line are implemented.

[0125] Beneficial effects: Compared with the prior art, the significant technical effects of the present invention are:

[0126] (1) The present invention adopts a distributed robust optimization method based on Wasserstein distance to analyze the uncertainty of renewable energy output in the load alignment calculation model, thereby improving the accuracy and reliability of the load alignment calculation model.

[0127] (2) The present invention reconstructs the traditional load criterion calculation model to form a load calculation model that takes into account the uncertainty of renewable energy output, thereby enhancing the adaptability and robustness of the model in complex environments and ensuring the stability of power grid operation.

[0128] (3) The present invention utilizes the adjustable power domain of the demand-side resource cluster to accurately adjust the non-adjustable load curve, thereby achieving flexible load regulation, significantly improving the accuracy and efficiency of power grid load regulation, reducing unnecessary power waste, and optimizing the overall operation effect of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0129] Figure 1 A schematic flow chart of a method for optimizing and controlling a resource cluster on the demand side of a distribution network based on a load criterion disclosed in an embodiment of the present invention;

[0130] Figure 2 A curve diagram of a non-adjustable load and a predicted output value of new energy sources disclosed in an embodiment of the present invention;

[0131] Figure 3 A time-of-use electricity price diagram disclosed in an embodiment of the present invention;

[0132] Figure 4 The optimal power consumption trajectory diagram of Case 1 disclosed in the embodiment of the present invention;

[0133] Figure 5 The optimal power consumption trajectory diagram of Case 2 disclosed in the embodiment of the present invention;

[0134] Figure 6 The optimal power consumption trajectory diagram of Case 3 disclosed in the embodiment of the present invention;

[0135] Figure 7 A comparison chart of new energy consumption of Case 1, Case 2, and Case 3 disclosed in the embodiment of the present invention;

[0136] Figure 8 An air conditioner and electric vehicle response capacity diagram disclosed in an embodiment of the present invention;

[0137] Fig. 9 A schematic diagram of the structure of a distribution network demand-side resource cluster optimization control system based on a load criterion disclosed in an embodiment of the present invention;

[0138] Fig.10 A schematic diagram of the structure of a distribution network demand-side resource cluster optimization control device based on a load criterion disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0139] The technical solution of the present invention is described in detail below in conjunction with specific implementation methods and the accompanying drawings of the specification.

[0140] Example 1

[0141] like Figure 1 As shown, a method for optimizing and controlling a resource cluster on the demand side of a distribution network based on a load criterion line of the present invention comprises the following steps:

[0142] S1. Construct a load criterion calculation model, analyze the uncertainty of new energy output when constructing the load criterion calculation model based on the distributed robust optimization method of Wasserstein distance, and obtain the uncertainty fuzzy set of new energy output.

[0143] S1.1. Construct a load criterion calculation model, including:

[0144] The core of the demand response mechanism based on the load standard is to optimize the original load curve shape rather than simply increase or decrease the power consumption. In this process, the total amount of electricity consumption of users is kept unchanged to ensure that their energy consumption benefits are not affected. By fine-tuning the distribution and changes of the load curve, the optimal allocation of power resources is achieved, and the efficiency and stability of the distribution network operation are improved.

[0145] First, define the load curve shape. Let l(t) be a load curve. The total power of this curve in time period T is α D In order to eliminate the difference in the magnitude of the load curve, α D The load curve l(t) is normalized as a reference, that is:

[0146]

[0147] l * (t) represents the normalized load value;

[0148] Normalization is to divide each point on the load curve by the total power to obtain a standardized load curve. After processing, the numerical characteristics of the load curve are eliminated, and only the shape characteristics are retained.

[0149] Reconstruct the load criterion calculation model, including the following:

[0150] (1) Objective function

[0151] With the goal of minimizing operating costs and absorbing new energy as much as possible, the objective function of the load standard line calculation model is established:

[0152]

[0153] In the above formula, the first item is the total operating cost of the generator set, and the second item is the cost of curtailing wind and solar power.

[0154] Where, f represents the total operating cost; a i , b i 、c i Represents the operating cost coefficient of the controllable generator set; N G Indicates the number of controllable generator sets; N RIndicates the number of new energy generating units; P G,i (t) represents the active output of the i-th controllable generator set in period t; P R,j (t) represents the output of the jth renewable energy generator set in period t; represents the maximum output of the j-th renewable energy generator set in period t, which is predicted on the day before; C R It represents the energy abandonment cost of the j-th renewable energy generator set. To absorb more renewable energy, this coefficient can be set larger.

[0155] (2) Constraints:

[0156] (a) Power balance constraints

[0157] Based on the idea of ​​balancing "unadjustable" with "adjustable", unadjustable resources such as new energy and rigid loads form a fluctuating curve. These fluctuations are balanced by controllable generators and demand-side adjustable loads, as follows:

[0158]

[0159] Where P D (t) represents the power consumption of the adjustable load participating in demand response in the system; P C (t) represents the power consumption of rigid loads in the system that do not participate in demand response; According to formula (1), P D (t) is the normalized value; A D It represents the total power consumption of the load participating in the demand response during the T period.

[0160] (b) Directrix Constraint

[0161]

[0162]

[0163] (c) Output constraints of controllable conventional units

[0164]

[0165] In the formula, It represents the upper and lower bounds of the active output of the i-th controllable generator set in time period t.

[0166] (d) Unit climbing constraints

[0167] -R D,i (t)≤P G,i (t)-P G,i (t-1)≤R U,i (t),t=1,2,…,T(6)

[0168] In the formula, R D,i (t), R U,i (t) represents the maximum downward and upward climbing rate of the i-th controllable generator set in time period t.

[0169] (e) Constraints on new energy output:

[0170]

[0171] In the formula, It represents the maximum output of the j-th renewable energy generator set in period t.

[0172] In this model, P G,i (t), P R,j (t) is a variable, and all others are constants. R,j (t) is regarded as a controllable variable because the phenomenon of wind power curtailment definitely exists, but what the model needs to consider is to promote the consumption of new energy as much as possible, that is, to minimize the cost of wind power curtailment.

[0173] However, due to the existence of new energy prediction errors, the uncertainty of new energy output must be fully considered when calculating the load standard. The present invention analyzes the uncertainty of new energy based on the distributed robust optimization method of Wasserstein distance to ensure the accuracy and effectiveness of demand response.

[0174] S1.2. Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of new energy output when constructing the load standard line calculation model is analyzed, and the uncertainty fuzzy set of new energy output is obtained, including:

[0175] Define the uncertain variable of new energy output prediction as l, then:

[0176]

[0177] In the formula, It represents the actual maximum output and predicted maximum output of the j-th renewable energy unit at time t.

[0178] M historical data {l1,l2,…,l M} is uniformly distributed as P e , P e is the empirical distribution, and the Wasserstein distance is used to measure the true probability distribution P and P e The distance between them is used to construct the distributed robust fuzzy set:

[0179]

[0180] In the formula, l and σ represent the distribution P e , P random variables; Π represents the joint distribution of l and σ; · represents the 1-norm; inf represents the infimum, which means finding the one with the smallest Wasserstein distance among all possible joint distributions; dl, dσ represent differentials; l1, l2, ... l M represents M groups of sample data of uncertain variables; E(P e ,P) represents the distance between the true probability distribution and the empirical distribution, namely the Wasserstein distance;

[0181] New energy output uncertainty set P N It can be expressed as:

[0182] P N = {P∈P(Ξ):W(P e ,P)≤ε(M)} (10)

[0183] Where Ξ represents the support set of random variables; P(Ξ) represents the set of all probability distributions with Ξ as the support set of random variables; ε(M) represents the radius of the probability distribution.

[0184] Among them, the uncertain set of new energy output P N is a circle with radius ε(M) and center P e The selection of parameter M has an important influence on the distributed blue bar optimization model. In the model, the size of the Wasserstein sphere radius is related to the uncertainty set P of the new energy output. N The larger the radius of the Wasserstein sphere, the larger the uncertainty set of new energy output P. N The types of uncertain distributions covered will increase, and the reliability of the system will also be improved.

[0185] S2. Combined with the uncertainty fuzzy set of new energy output, the load criterion calculation model is reconstructed to obtain the load calculation model considering the uncertainty of new energy output. The details are as follows:

[0186] When adjusting the fuzzy set sphere radius in the distributed robust optimization model, it is necessary to balance the economy and reliability of the distribution network system as much as possible.

[0187] S2.1. Objective function of load alignment calculation model reconstruction:

[0188] After considering the new energy prediction error, the objective function of the load criterion calculation model is adjusted, that is, the objective function formula (2) should be adjusted as shown in formula (11), which consists of a deterministic cost part and an uncertain cost part.

[0189]

[0190] Where, f represents the total operating cost; a i 、b i 、c i Represents the operating cost coefficient of the controllable generator set; N G Indicates the number of controllable generator sets; P G,i (t) represents the active output of the i-th controllable generator set in period t; P R,j (t) represents the output of the jth renewable energy generator set in period t; represents the actual maximum output of the j-th renewable energy unit at time t; C R N represents the energy abandonment cost of the j-th renewable energy generator set; R Indicates the number of renewable energy generating units.

[0191] The above objective function can be simplified as follows:

[0192]

[0193] Where x represents each decision variable; X represents the set of decision variables x; f(x) is the deterministic part of the objective function and has nothing to do with the uncertainty variable l; E represents the expected cost under the worst case uncertainty. P (·) represents the expected probability; sup represents the supremum function; P represents the probability distribution of uncertain variables;

[0194] Combined with the uncertainty of new energy output fuzzy set P N , based on the strong duality theory, Reconstruct it into the following formula to reduce the difficulty of model calculation.

[0195]

[0196] Where ε(M) represents the radius of the probability distribution; l represents the random variable that obeys the distribution P; Ξ represents the support set of the random variable; M represents the historical data {l1,l2,…,l M}; l k represents the kth group of samples; λ represents the dual variable.

[0197] Introduce auxiliary variable s k , the above formula can be further rewritten as:

[0198]

[0199] And because h(x,l)-λ||ll k || is a convex function on the interval of the random variable l, so its supremum is obtained at the interval boundary:

[0200]

[0201] In the formula, l , represents the upper and lower bounds of uncertain variables; h(x,l) represents the loss function.

[0202] Therefore, the final model is transformed into:

[0203]

[0204] Among them, Ax≤b represents the deterministic constraint condition, and A and b are the coefficient matrix and parameter vector of the constraint condition respectively.

[0205] S2.2、Construction of opportunity constraints:

[0206] In addition to the uncertain variables in the objective function, the constraints of renewable energy output also contain uncertain variables. The opportunity constraint method is used to model the inequality constraints containing uncertain variables in renewable energy output, so that the inequality constraints containing the probability distribution P of uncertain variables are included in the fuzzy set P of renewable energy output uncertainty. N The expected probability E P is greater than the minimum allowed confidence level (i.e., 1-γ, where γ is the risk factor for exceeding the inequality constraint), that is,

[0207]

[0208] S3. Based on the calculation results of the load quasi-line calculation model considering the uncertainty of new energy output, a load quasi-line demand-side resource cluster optimization control strategy considering the uncertainty of new energy output is proposed. The power grid dispatching center accurately adjusts the non-adjustable load curve based on the adjustable power domain information of the demand-side resource cluster.

[0209] The distribution network operation control center accurately adjusts the non-adjustable load curve according to the adjustable capacity of the demand-side resources. Step S3 specifically includes the following steps:

[0210] S3.1. On the day-ahead, comprehensively consider the forecast data of new energy, the adjustable power domain of demand-side resources and the published load criterion, with the goal of maximizing overall benefits. Construct a day-ahead optimization model for the demand-side resource cluster to participate in quasi-linear demand response.

[0211] In this embodiment, the process of constructing the day-ahead optimization model is as follows:

[0212] (1) Determine the objective function of the day-ahead optimization model:

[0213]

[0214] Among them, F buyrepresents the cost of purchasing energy; F cost represents the adjustment cost of the demand-side resource cluster; F sell Indicates the benefit of electricity sales; F reward It represents the incentive compensation obtained by the demand-side resource cluster participating in the quasi-linear demand response; the cost of purchasing energy includes the cost of purchasing electricity from the power grid company, wind power and photovoltaic power, and its expression is as follows:

[0215]

[0216] In the formula, λ t represents the price of electricity purchased from the power grid company; WT represents the purchase price of wind power; PV P represents the purchase price of photovoltaic power generation; t buy represents the amount of electricity purchased by demand-side resources during period t; P t WT , P t PV It represents the wind power and photovoltaic power consumption in period t; Δt represents the length of the time interval.

[0217] The revenue from electricity sales includes:

[0218]

[0219] In the formula, F sell Represents the revenue from electricity sales; S t P represents the electricity price of the demand-side resource cluster participating in the response; t E represents the power consumption of demand-side resources; r t Represents other benefits at time t.

[0220] (2) Set day-ahead constraints, including:

[0221] ① Power balance constraints:

[0222] P t buy +P t PV +P t WT =P t E +r t (twenty one)

[0223] ② Wind power and photovoltaic power generation power constraints:

[0224] The following formula indicates that the wind and photovoltaic power generation should be less than the wind and photovoltaic predicted value on the day before;

[0225] 0≤P tPV ≤P t PV0 (twenty two)

[0226] 0≤P t WT ≤P t WT0 (twenty three)

[0227] Where P t WT0 , P t PV0 Indicates the day-ahead forecast value of wind power and photovoltaic power generation;

[0228] ③ Adjustable load constraints on the demand side:

[0229] P t E +r t ∈Ω agg,t (twenty four)

[0230] In the formula, Ω agg,t Indicates the adjustable power domain of the demand-side resource cluster.

[0231] S3.2. In the day, the cluster power consumption needs to be decomposed into various adjustable loads. After receiving the cluster control curve and expected benefits, the user builds a daily decomposition optimization model for decomposing the cluster power consumption into various adjustable loads based on its own load control cost, benefit distribution, load operation requirements, etc., with the purpose of optimizing the economic efficiency of adjustable load i.

[0232] In this embodiment, the construction process of the intraday decomposition optimization model is as follows:

[0233] Considering that there are n adjustable loads in the demand-side resource cluster, the decomposition strategy needs to meet the following conditions:

[0234] (1) Load operation requirements

[0235] The power consumption of each adjustable load i and the adjustment amount r of the adjustable load i participating in the response i,t The sum must be within its adjustable power domain to ensure the feasibility of regulation. , the expression is as follows:

[0236]

[0237] In the formula, represents the power consumption of adjustable load i; r i,t Indicates the adjustment amount of the adjustable load i participating in the response; Ω i,t represents the adjustable power domain of adjustable load i.

[0238] (2) Control costs

[0239]

[0240] In the formula, f i,cost represents the regulation cost of adjustable load i; β i represents the control cost coefficient of adjustable load i;

[0241] (3) Profit distribution

[0242] The improved Shapley method is used to distribute benefits within the adjustable load cluster, and the benefit distribution is optimized according to the importance of different types of adjustable loads; the greater the importance of adjustable load i, the more benefits can be allocated, as follows:

[0243]

[0244] Δ i (S) = I (S) - I (S - {i}) (29)

[0245] I i =∑W(|S|)Δ i (S) (30)

[0246] Where S represents any alliance with adjustable resource i, and Δ i (S) represents the marginal benefit contribution of adjustable load i to alliance S; W(|S|) represents the weighting factor, which indicates the probability of alliance S appearing; I i represents the benefit distribution function of adjustable load i in alliance N; |S| represents the number of stakeholder alliances; (|S|-1)! represents the number of rankings of a member when participating in alliance S; n represents the total number of members in the alliance; n! represents the factorial of n, which represents the total number of permutations and combinations of all elements.

[0247] The benefit R of the adjustable load i i for

[0248]

[0249] In the formula, represents the importance factor of adjustable load i. The larger the importance factor, the greater the benefit allocated. reward Represents the incentive compensation obtained by the demand-side resource cluster participating in quasi-linear demand response.

[0250] Therefore, with the purpose of optimizing the economic efficiency of adjustable load i, the intra-day decomposition optimization model is as follows:

[0251] maxF i =R i -f i,cost (32)

[0252] The effectiveness of the load criterion-based demand-side resource cluster optimization control method proposed in the present invention is verified by a specific embodiment below.

[0253] The verification environment is set as follows: In a distribution network with a high proportion of new energy, a large number of users are considered to participate in demand response, and the rationality and effectiveness of the proposed strategy are analyzed. In this distribution network, the adjustable loads consider 200 electric vehicles and 200 air conditioners. The load parameters are shown in Table 1. Their energy consumption accounts for 10% and 16% of the total load respectively, and the similarity threshold is 70%. The non-adjustable load and the predicted value of new energy output are as follows Figure 2 As shown in Figure 2, the penetration rate of renewable energy electricity is 65.4%, and the time-of-use electricity price is as follows: Figure 3 shown.

[0254] The load parameter settings are shown in Table 1:

[0255] Table 1: Equipment parameters

[0256]

[0257]

[0258] Firstly, the uncertainty of renewable energy output when constructing the load standard line calculation model is analyzed based on the distributed robust optimization method of Wasserstein distance, and the fuzzy set of renewable energy output uncertainty is obtained. The load standard line calculation model is reconstructed to obtain the load calculation model considering the uncertainty of renewable energy output. The distribution network operation control center uses the adjustable capacity information of demand-side resources received to make refined load adjustments and analyze the optimal power consumption trajectory on the day before.

[0259] It should be noted that: in the day-ahead stage, the adjustable capacity of demand-side resources and load guidelines are comprehensively considered to obtain the electricity consumption trajectory of demand-side resources for the next day with the best overall economic efficiency; in the intraday stage, based on factors such as load control costs, profit distribution, and load operation requirements, a refined decomposition and control strategy is formulated to decompose the electricity consumption trajectory obtained in the day-ahead stage into each type of load in real time.

[0260] The dispatch cycle is divided into 96 time periods, each of which is 15 minutes. In order to verify the implementation effect of the demand-side adjustable resources participating in the quasi-linear demand response mechanism, the following comparative experiments are set up:

[0261] Case 1: Adjustable load participates in quasi-linear demand response and sets P m = 0.25 yuan / (kWh), adjustable load tracking load line to obtain response incentives, thereby reducing electricity costs;

[0262] Case 2: The adjustable load does not participate in any incentive demand response, that is, P m =0 yuan / (kWh);

[0263] Case 3: Adjustable load participates in time-of-use electricity price demand response.

[0264] In the high-proportion new energy distribution network set up in this embodiment, the load standard line is calculated through controllable unit parameters, non-adjustable load data, new energy data considering uncertainty and total load data.

[0265] Figure 4 is the optimal electricity consumption trajectory diagram for Case 1, Figure 4 It can be seen that the load criterion has peak values ​​near 3:00, 13:30, and 22:30, which correspond to Figure 4 During the three output peaks of the new energy curve, the power grid hopes that the adjustable resources on the user side can absorb the new energy as much as possible; in the two time periods of 5:30-11:00 and 16:00-21:00, the electricity price is higher and the output of new energy is lower, so the distribution network hopes that the adjustable resources can reduce electricity consumption, so the load standard line is reduced to 0.

[0266] exist Figure 4 In the process, the adjustable load cluster obtains response incentives by tracking the load criterion, which makes the shape of the load curve change significantly. In the period of 1:00-5:00, under the guidance of the low electricity price and the load criterion, the adjustable load cluster actively participates in the response and accurately tracks the load criterion; around 13:30, the electricity price is high but the output of new energy is large. At this time, the adjustable load cluster should use more electricity to increase the consumption rate of new energy; in the two periods of 6:00-10:30 and 17:00-20:30, the output of new energy is low and the electricity price is at its peak. At this time, the adjustable load cluster does not use a lot of electricity, but instead reduces electricity consumption by discharging.

[0267] Figure 5 is the optimal electricity consumption trajectory diagram for Case 2, Figure 5 It can be seen that the adjustable load cluster is affected by the electricity price, discharging electricity at a high price and consuming electricity at a low price. That is, when the electricity price is at a peak, the adjustable load cluster reduces electricity consumption, and the net electricity consumption is 0; when the electricity price is at a low point, the adjustable load cluster increases electricity consumption. However, due to P m =0 yuan / (kWh), there is no response incentive, resulting in the adjustable load cluster having no motivation to track the load benchmark, so Case 2 is not as effective as Case 1 in promoting the consumption of new energy.

[0268] Figure 6 is the optimal power consumption trajectory diagram for Case 3, Figure 6It can be seen that the demand response mechanism based on time-of-use electricity prices is prone to "peak-valley inversion", that is, the electricity consumption during the valley electricity price period is higher than the electricity consumption during the peak electricity price period. This is not suitable for large-scale applications and has a poor effect on promoting the consumption of new energy.

[0269] Secondly, after obtaining the load calculation model that takes into account the uncertainty of renewable energy output, the distribution network dispatching center uses the adjustable capacity information of demand-side resources to make refined load adjustments and analyze the effect of promoting renewable energy consumption.

[0270] It should be noted that the quasi-linear demand response is aimed at improving the absorption capacity of new energy, thereby improving energy utilization efficiency. In order to verify the effectiveness of the demand-side resource cluster optimization control strategy based on the load quasi-line, the effects of promoting the absorption of new energy in three cases are analyzed.

[0271] Figure 7 The following is a comparison chart of new energy consumption in three cases. We can clearly see that in the absence of incentives (i.e. Case 3), there is an obvious phenomenon of new energy abandonment in the early morning, afternoon and late night. However, in Case 1 and Case 2, the consumption rate of new energy has been significantly improved by encouraging users to actively participate in demand response. This not only reduces energy waste, but also helps to delay investment in distribution networks. And it can be seen from the figure that the demand response mechanism based on the load criterion has the best effect in absorbing new energy. This mechanism achieves efficient utilization of new energy by accurately adjusting the load curve so that the load demand of the distribution network is more matched with the power supply capacity.

[0272] Finally, the distribution network operation and control center receives the adjustable capacity information of demand-side resources and uses it to make refined load adjustments and analyze the intra-day decomposition strategy.

[0273] It should be noted that: during the intraday stage, the load control cost, profit distribution, load operation requirements, etc. are taken into consideration, and the decomposition control strategy is formulated with the goal of maximizing the adjustable load profit.

[0274] The optimal load curve obtained by the day-ahead optimization model is completed by the joint response of the air conditioner and electric vehicle clusters. Figure 8 The decomposition results show that during the evening period of 20:00-4:00, the electric vehicle cluster plays the main response role and completes most of the optimal adjustment tasks through charging operations; during the daytime period of 6:00-18:00, as electric vehicles travel, the air conditioning cluster shoulders the main responsibility of demand response, adjusting power consumption to stabilize the power supply demand of the distribution network. This decomposition result not only fits the power consumption characteristics and habits of various load clusters, but also maximizes the overall economic benefits through refined response adjustment strategies.

[0275] Example 2

[0276] like Fig. 9 As shown, a distribution network demand side resource cluster optimization control system based on load criterion of the present invention comprises:

[0277] The fuzzy set acquisition module is used to construct the load criterion calculation model. Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of new energy output when constructing the load criterion calculation model is analyzed to obtain the uncertainty fuzzy set of new energy output;

[0278] The load calculation model construction module is used to reconstruct the load criterion calculation model by combining the uncertainty fuzzy set of new energy output, and obtain the load calculation model considering the uncertainty of new energy output;

[0279] The load curve adjustment module is used to accurately adjust the non-adjustable load curve based on the adjustable power domain of the demand-side resource cluster.

[0280] In an optional embodiment, a method for optimizing and controlling a resource cluster on the demand side of a distribution network based on a load criterion includes: a) constructing a load criterion calculation model, and analyzing the uncertainty of new energy sources based on a distributed robust optimization method based on Wasserstein distance; b) reconstructing a model containing fuzzy sets of uncertain variables to obtain a load calculation model that takes into account the uncertainty of new energy output; c) accurately adjusting the non-adjustable load curve based on the adjustable power domain of the demand side resource cluster.

[0281] Example 3

[0282] like Fig.10 As shown, the load standard line-based distribution network demand side resource cluster optimization control device of the present invention can be applied to power systems, such as for distribution network demand side resource cluster optimization, etc., and the embodiments of the present invention are not limited.

[0283] like Fig.10 As shown, the device may include a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the electronic device implements the steps of the method described in the above embodiment and can achieve the technical effect consistent with the above method.

[0284] The memory may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The device may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the memory may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as a "hard drive"). A program / utility having a set (at least one) of program modules may be stored in, for example, the memory, such program modules including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include the implementation of a network environment. The program modules typically perform the functions and / or methods in the embodiments described herein.

[0285] The processor executes various functional applications and data processing by running the program stored in the memory, such as implementing the method provided in the first embodiment of the present invention.

[0286] Example 4

[0287] Embodiment 4 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the method described in the above embodiment are implemented and the technical effect consistent with the above method can be achieved.

[0288] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, a device or a device or used in combination with it.

[0289] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0290] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0291] Computer program code for performing the operation of the present invention may be written in one or more programming languages ​​or combinations thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0292] Of course, the storage medium containing computer executable instructions provided by an embodiment of the present invention, whose computer executable instructions are not limited to the above method operations, can also execute related operations in the method provided by any embodiment of the present invention.

[0293] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for optimizing and controlling resource clusters on the demand side of a distribution network based on a load criterion, characterized in that: include: The load criterion calculation model is constructed. Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of renewable energy output is analyzed when the load criterion calculation model is constructed, and the uncertainty fuzzy set of renewable energy output is obtained. Combined with the uncertainty fuzzy set of renewable energy output, the load criterion calculation model is reconstructed to obtain the load criterion calculation model considering the uncertainty of renewable energy output. According to the calculation results of the load quasi-line calculation model considering the uncertainty of renewable energy output, a load quasi-line demand-side resource cluster optimization control strategy considering the uncertainty of renewable energy output is proposed. The power grid dispatching center accurately adjusts the non-adjustable load curve according to the adjustable power domain information of the demand-side resource cluster.

2. The method for optimizing and controlling the resource cluster on the demand side of the distribution network based on the load criterion according to claim 1 is characterized in that: Construct a load alignment calculation model, including: First, define the load curve shape. Let l(t) be a load curve whose total power in time period T is α D ; with α D The load curve l(t) is normalized as a reference, that is: l * (t) represents the normalized load value; Reconstruct the load line calculation model, including: With the goal of minimizing operating costs, the objective function of the load criterion calculation model is established: Where N G Indicates the number of controllable generator sets; N R Indicates the number of new energy generating units; P G,i (t) represents the active output of the i-th controllable generator set in period t; P R,j (t) represents the output of the jth renewable energy generator set in period t; represents the maximum output of the jth renewable energy generator set in period t; C R represents the energy abandonment cost of the j-th renewable energy generator set; f represents the total operating cost; a i , b i 、c i represents the operating cost coefficient of the controllable generating set; Set the following constraints: Power balance constraints: The non-adjustable resources including new energy and rigid loads form a fluctuating curve, and the fluctuation is balanced by the controllable generators and the adjustable loads on the demand side, as follows: Where P D (t) represents the power consumption of the adjustable load participating in demand response in the system; P C (t) represents the power consumption of rigid loads in the system that do not participate in demand response; Indicates P D (t) is the normalized value; A D It represents the total power consumption of the load participating in the demand response during the T period; Directrix Constraints: Output constraints of controllable conventional units: In the formula, It represents the upper and lower bounds of the active output of the ith controllable generator set in period t; Unit climbing constraints: -R D,i (t)≤P G,i (t)-P G,i (t-1)≤R U,i (t),t=1,2,…,T In the formula, R D,i (t), R U,i (t) represents the maximum downward and upward climbing rate of the i-th controllable generator set in time period t; New energy output constraints: In the formula, It represents the maximum output of the j-th renewable energy generator set in period t.

3. The method for optimizing and controlling the resource cluster on the demand side of the distribution network based on the load criterion according to claim 1 is characterized in that: Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of renewable energy output when constructing the load standard line calculation model is analyzed, and the uncertainty fuzzy set of renewable energy output is obtained, including: Define the uncertain variable of new energy output prediction as l, then: In the formula, Indicates the actual maximum output and predicted maximum output of the j-th renewable energy unit at time t; M historical data {l1,l2,…,l M } is uniformly distributed as P e , P e is the empirical distribution, and the Wasserstein distance is used to measure the true probability distribution P and P e The distance between them is used to construct the distributed robust fuzzy set: In the formula, l and σ represent the distribution P e , P random variables; Π represents the joint distribution of l and σ; ||·|| represents the distance; inf represents the infimum; dl, dσ represent the differential; l1, l2, ... l M M groups of sample data representing uncertain variables; W(P e ,P) represents the distance between the true probability distribution and the empirical distribution, namely the Wasserstein distance; New energy output uncertainty set P N It is expressed as: P N ={P∈P(Ξ):W(P e ,P)≤ε(M)} Where Ξ represents the support set of random variables; P(Ξ) represents the set of all probability distributions with Ξ as the support set of random variables; ε(M) represents the radius of the probability distribution.

4. The method for optimizing and controlling the resource cluster on the demand side of the distribution network based on the load criterion according to claim 1, characterized in that: Combined with the uncertainty fuzzy set of renewable energy output, the load criterion calculation model is reconstructed to obtain the load calculation model considering the uncertainty of renewable energy output, including: Reconstruct the objective function of the load alignment calculation model: After considering the new energy prediction error, the objective function of the load criterion calculation model is adjusted to: Where, f represents the total operating cost; a i , b i 、c i Represents the operating cost coefficient of the controllable generator set; N G Indicates the number of controllable generator sets; P G,i (t) represents the active output of the i-th controllable generator set in period t; P R,j (t) represents the output of the jth renewable energy generator set in period t; represents the actual maximum output of the j-th renewable energy unit at time t; C R N represents the energy abandonment cost of the j-th renewable energy generator set; R Indicates the number of new energy generating units; The above objective function is simplified as follows: Where x represents each decision variable; X represents the set of decision variables x; f(x) is the deterministic part of the objective function and has nothing to do with the uncertainty variable l; E represents the expected cost under the worst case uncertainty. P (·) represents the expected probability; sup represents the supremum function; P represents the probability distribution of uncertain variables; Combined with the uncertainty of new energy output fuzzy set P N , based on the strong duality theory Reconstructed as follows: Where ε(M) represents the radius of the probability distribution; l represents the random variable that obeys the distribution P; Ξ represents the support set of the random variable; M represents the historical data {l1, l2, …, l M }; l k represents the kth group of samples; λ represents the dual variable; Introduce auxiliary variable s k , the above formula can be further rewritten as: And because h(x,l)-λ||ll k || is a convex function on the interval of the random variable l, so its supremum is obtained at the interval boundary: s.t.h(x, l )+λ( l -l k )≤s k h(x,l)+λ(l-l k )≤s k h(x,l)≤s k In the formula, l , represents the upper and lower bounds of uncertain variables; h(x,l) represents the loss function; Therefore, the final model is transformed into: stAx≤b h(x, l )+λ( l -l k )≤s k h(x,l)≤s k Among them, Ax≤b represents the deterministic constraint condition, A and b are the coefficient matrix and parameter vector of the constraint condition respectively; Opportunity Constraint Construction: The opportunity constraint method is used to model the inequality constraints containing uncertain variables of renewable energy output, so that the inequality constraints containing the probability distribution P of uncertain variables are modeled in the fuzzy set P of renewable energy output uncertainty. N The expected probability E P Greater than the minimum allowed confidence level, that is, 1-γ, where γ is the risk factor for the inequality constraint violation, as follows:

5. The method for optimizing and controlling the resource cluster on the demand side of the distribution network based on the load criterion according to claim 1, characterized in that: Based on the adjustable power domain of the demand-side resource cluster, the non-adjustable load curve is accurately adjusted, including: On the day-ahead, the forecast data of new energy, the adjustable power domain of demand-side resources and the published load quasi-line are comprehensively considered, and the overall benefit is maximized to build a day-ahead optimization model for the demand-side resource cluster to participate in quasi-linear demand response. During the day, after receiving the cluster control curve and expected benefits, users build a daily decomposition optimization model for decomposing cluster electricity consumption into various adjustable loads based on their own load control costs, benefit distribution, and load operation requirements, with the goal of optimizing the economic efficiency of adjustable load i.

6. The method for optimizing and controlling the resource cluster on the demand side of the distribution network based on the load criterion according to claim 5 is characterized in that: The construction process of the day-ahead optimization model is as follows: Determine the objective function of the day-ahead optimization model: Among them, F buy represents the cost of purchasing energy; F cost represents the adjustment cost of the demand-side resource cluster; F sell Indicates the benefit of electricity sales; F reward It represents the incentive compensation obtained by the demand-side resource cluster participating in the quasi-linear demand response; The cost of purchasing energy includes the cost of purchasing electricity from the power grid company, wind power and photovoltaic power, and the expression is as follows: In the formula, λ t represents the price of electricity purchased from the power grid company; WT represents the purchase price of wind power; PV P represents the purchase price of photovoltaic power generation; t buy represents the amount of electricity purchased by demand-side resources during period t; P t WT , P t PV It represents the wind power and photovoltaic power generation consumption power in period t; Δt represents the length of the time interval; The revenue from electricity sales includes: In the formula, F sell Represents the revenue from electricity sales; S t P represents the electricity price of the demand-side resource cluster participating in the response; t E represents the power consumption of demand-side resources; r t represents other benefits at time t; Set day-ahead constraints, including: Power balance constraints: P t buy +P t PV +P t WT =P t E +r t Wind power and photovoltaic power generation power constraints: The following formula indicates that the wind and photovoltaic power generation should be less than the wind and photovoltaic predicted value on the day before; 0≤P t PV ≤P t PV0 0≤P t WT ≤P t WT0 Where P t WT0 , P t PV0 Indicates the day-ahead forecast value of wind power and photovoltaic power generation; Demand side adjustable load constraints: P t E +r t ∈Ω agg,t In the formula, Ω agg,t Indicates the adjustable power domain of the demand-side resource cluster.

7. The method for optimizing and controlling the resource cluster on the demand side of the distribution network based on the load criterion according to claim 5, characterized in that: The construction process of the intraday decomposition optimization model is as follows: Considering that there are n adjustable loads in the demand-side resource cluster, the decomposition strategy needs to meet the following conditions: Load operation requirements: The power consumption of each adjustable load i and the adjustment amount r of the adjustable load i participating in the response i,t The sum is within its adjustable power domain to ensure the feasibility of regulation. The expression is as follows: In the formula, represents the power consumption of adjustable load i; r i,t Indicates the adjustment amount of the adjustable load i participating in the response; Ω i,t represents the adjustable power domain of adjustable load i; Control costs: In the formula, f i,cost represents the regulation cost of adjustable load i; β i represents the control cost coefficient of adjustable load i; Profit distribution: The improved Shapley method is used to distribute benefits within the adjustable load cluster. The benefit distribution is optimized according to the importance of different types of adjustable loads. The greater the importance of adjustable load i, the more benefits can be allocated, as follows: Δ i (S)=I(S)-I(S-{i}) I i =∑W(|S|)Δ i (S) Where S represents any alliance with adjustable resource i, and Δ i (S) represents the marginal benefit contribution of adjustable load i to alliance S; W(|S|) represents the weighting factor; I i represents the profit distribution function of adjustable load i in alliance N; |S| represents the number of stakeholder alliances; (|S|-1)! represents the ranking number of a member when participating in alliance S; n represents the total number of members in the alliance; n! represents the factorial of n; The benefit R of the adjustable load i i for In the formula, represents the importance factor of adjustable load i. The larger the importance factor, the greater the benefit allocated. reward It represents the incentive compensation obtained by the demand-side resource cluster participating in the quasi-linear demand response; Therefore, with the purpose of optimizing the economic efficiency of adjustable load i, the intra-day decomposition optimization model is as follows: maxF i =R i -f i,cost 。 8. A distribution network demand side resource cluster optimization control system based on load criterion, characterized in that: include: The fuzzy set acquisition module is used to construct the load criterion calculation model. Based on the distributed robust optimization method of Wasserstein distance, the uncertainty of new energy output when constructing the load criterion calculation model is analyzed to obtain the uncertainty fuzzy set of new energy output; The load calculation model construction module is used to reconstruct the load criterion calculation model by combining the uncertainty fuzzy set of new energy output, and obtain the load calculation model considering the uncertainty of new energy output; The load curve adjustment module is used to propose a load quasi-linear demand-side resource cluster optimization control strategy that takes into account the uncertainty of renewable energy output based on the calculation results of the load quasi-linear calculation model that takes into account the uncertainty of renewable energy output. The power grid dispatching center accurately adjusts the non-adjustable load curve based on the adjustable power domain information of the demand-side resource cluster.

9. A distribution network demand side resource cluster optimization control device based on load criterion, characterized in that: The electronic device comprises a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the electronic device implements the steps of the distribution network demand side resource cluster optimization control method based on the load standard line as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for optimizing and controlling a distribution network demand-side resource cluster based on a load standard line as claimed in any one of claims 1 to 7.

Images