Method for Identifying Schedulable Flexible Resources Based on Grey Relational Theory, Storage Medium

Abstract

A method and storage medium for identifying schedulable flexible resources based on the grey correlation theory according to the present invention include establishing a power grid topological network structure according to the scale of the distribution network in the analysis area, selecting various schedulable flexible resources in the distribution network, and establishing a mathematical model; setting a system objective function and constraint functions, and generating an optimal unit output combination for realizing the coordinated optimization of robustness and economy; randomly generating configuration data of the schedulable flexible resources, substituting them into the objective function to calculate various costs and forming multiple groups of data; calculating the correlation degrees of multi-scenario data by using the grey correlation theory based on the entropy weight method to obtain the weighted comprehensive correlation degrees of each schedulable flexible resource; and the sorting result of the magnitudes of the weighted comprehensive correlation degrees is the influence degree of the schedulable flexible resources on the power grid planning. The present invention overcomes the problem of identifying the influence of schedulable flexible resources in the current power system planning, and improves the accuracy and efficiency of power system planning decisions.

Description

Technical Field

[0001] The present invention relates to power system planning technology, and in particular to a method for identifying schedulable flexible resources based on grey correlation theory. Background Art

[0002] After entering the industrial society, the importance of power development in social development has been increasing day by day, but there are still power accidents caused by power shortages. At the same time, with the proposal of the "dual carbon" goal, the proportion of renewable new energy in the power grid is increasing, and the uncertainty of its output will have a series of impacts on the power system during the grid connection process, mainly manifested in transient voltage fluctuations and flicker, the introduction and distortion of network voltage harmonics, and increased line losses. At the same time, in order to make the schedulable new energy approach "conventional power sources", energy storage devices are usually connected. Because energy storage devices can achieve fast, stable, and accurate charge and discharge power regulation characteristics at the millisecond level, they can improve the instantaneous, short-term, and time-section balance capabilities of the power system, and improve the safety and stability level of the power system, the clean energy consumption capacity, the energy and power supply guarantee, and the efficiency and benefit of the energy and power system. In addition to schedulable new energy, the spatial mobility and temporal flexibility of electric vehicles enable them to better participate in grid interaction and relieve the power consumption pressure of the grid. In recent years, they have received strong support from national policies, resulting in a sharp increase in the number of electric vehicles and the number of supporting electric charging stations. However, this also puts new requirements on the load structure and characteristics of the distribution network, and high-power fast charging also increases the difficulty of optimizing and controlling the grid operation.

[0003] In recent years, the academic community has focused on flexible loads such as schedulable new energy, energy storage, and electric vehicles, which can actively participate in grid operation control, interact with the grid in terms of energy, and have flexible characteristics, and defined them as flexible loads. Different from traditional power systems with only rigid loads, today's power systems have a large number of flexible loads on the demand side, and the energy flow of flexible loads is not unidirectional from the grid to the load, which gives flexible loads the ability to actively participate in grid scheduling. Therefore, in recent years, schedulable flexible resources have been incorporated into the power system planning model.

[0004] Most of the current power system planning only focuses on the planning of thermal power units and some schedulable flexible resources, and fails to fully consider the current variety of flexible loads. Even in some literatures, although the coordinated optimization of multiple flexible loads is considered, only an optimization modeling method for output is provided. Different flexible loads have different mechanism characteristics, and their influencing ways and degrees on power system planning are different. Therefore, based on the power system coordinated optimization technology, a method is proposed in this paper to quantitatively analyze the influence of schedulable flexible resources in power system planning.

[0005] Quantitative analysis of the impact degree on flexible loads is very important, which can deeply explore the core factors of power grid planning and layout, highly targeted improve the efficiency of power grid planning, provide data support for the collaborative optimization of schedulable flexible resources in power grid planning and layout, and improve the economy and reliability of power grid operation.

[0006] The main quantitative analysis methods for power load influencing factors at home and abroad are as follows: direct curve comparison method, regression analysis method and correlation analysis method in mathematical statistics, as well as grey relational analysis, data envelopment analysis, neural network and wavelet analysis in artificial intelligence methods. Compared with other methods, the grey relational degree analysis method has little requirement for the quantity and distribution of sample data, and there is no obvious strong direct correlation between the two. The result of grey relational degree calculation is also relatively accurate and in good agreement with the qualitative analysis result. Summary of the Invention

[0007] To overcome the deficiencies of the above-mentioned existing technologies, based on the existing power system planning theory, this paper fully considers the existing complex types of schedulable flexible resources, and innovatively applies the grey relational theory to the identification of the influence of schedulable flexible resources on power grid planning, and quantitatively analyzes the degree of their influence on power grid planning.

[0008] To achieve the purpose of the above invention, the present invention adopts the following technical solutions:

[0009] S1, establish a power grid topological network structure according to the scale of the distribution network in the analysis area;

[0010] S2, screen the schedulable flexible resources according to the output situation, scale, etc. of the schedulable flexible resources in the distribution network of the analysis area, and establish a mathematical model;

[0011] S3, set the objective function and constraint function of the system according to the mathematical model, and generate the optimal unit output combination for realizing the coordinated optimization of robustness and economy;

[0012] S4, on the premise of ensuring the stable operation of the power grid, randomly generate multiple groups of configuration data of each schedulable flexible resource, substitute them into the objective function in step S3 for calculation, and generate the economic parameters of power grid planning under multiple scenarios;

[0013] S5, use the grey relational theory based on the entropy weight method to analyze the data in step S5, and judge the influence degree of each flexible resource on power grid planning through the size of the final weighted comprehensive correlation degree.

[0014] Further, in step S2, the key schedulable flexible resources are defined as five categories: load that can be curtailed, energy storage, microgrid, schedulable new energy, and electric vehicle.

[0015] Furthermore, in step S2, the modeling of schedulable flexible resources is divided into three parts, namely:

[0016] 1) Wind turbine uncertain output model:

[0017]

[0018] In the formula: is the actual value of wind speed; is the predicted value of wind speed; Δv w is the prediction error of wind speed; and are the upper and lower limits of wind speed prediction;

[0019] The relationship between the output power of the wind turbine and the wind speed is shown in the following formula:

[0020]

[0021] In the formula: P w,t is the power output of the wind turbine at time t; v w,t is the actual wind speed of the wind turbine at time t; and are the cut-in and cut-out wind speeds of the wind turbine; is the rated wind speed of the wind turbine; is the rated power generation of the wind turbine;

[0022] 2) Distributed energy storage model:

[0023]

[0024] In the formula: SOC represents the state of charge of the energy storage device; E now and E N are the current battery capacity and the rated battery capacity of the energy storage device; SOC(t) and SOC(t + 1) respectively represent the state of charge of the energy storage device before and after charging; P c and P d are the charging and discharging powers of the energy storage device respectively; η c and η d are the charging and discharging efficiencies of the energy storage device respectively; Δt is the time for the energy storage device to participate in power grid scheduling;

[0025] 3) Curtailable load model:

[0026]

[0027]

[0028] In the formula: Q p,t and Q c,tThe penalty load and compensation load in the t-th hour period, respectively; Q true,t , Q req,t , Q fore,t The actual load of the user, the reduced load, and the grid predicted load in the t-th hour period, respectively.

[0029] Furthermore, in step S3, the modeling objective functions for dispatchable flexible resources are divided into six, which are respectively:

[0030] Objective function minf1:

[0031]

[0032] In the formula: N wi represents the number of wind turbines in this area; P waste,t represents the wind curtailment volume in the t-th hour; m waste,t represents the unit price of wind curtailment penalty in the t-th hour;

[0033] Objective function minf2:

[0034] M p,t = m p,t ·Q p,t

[0035] M c,t = m c,t ·Q c,t

[0036] f2 = M c,t - M p,t

[0037] In the formula: M p,t and M c,t are the penalty amount and compensation amount of the user in the t-th hour period respectively; m p,t and m c,t are the unit prices of the penalty and compensation amounts agreed in the contract in the t-th hour respectively;

[0038] Objective function minf3:

[0039]

[0040] In the formula: m loss is the unit cost of network loss; E line is the set of all branches in this area; I ij,t is the current of branch ij in the t-th hour period; R ij,t is the resistance of branch ij in the t-th hour period;

[0041] Objective function minf4:

[0042] f4 = m buy,t ·Q buy,t

[0043] Where: m buy,t is the upward power purchase unit price at the t-th hour; Q buy,t is the upward power purchase load at the t-th hour;

[0044] Objective function minf5:

[0045] f5 = m sell,t ·Q sell,t

[0046] Where: m sell,t is the power selling unit price at the t-th hour; Q sell,t is the upward power purchase load at the t-th hour;

[0047] Objective function minf6:

[0048]

[0049] Where: p is the unit loss of order cancellation; D i is the average load of load node i; U i is the average annual power outage time.

[0050] Furthermore, in step S4, the configuration data of dispatchable flexible resources are randomly generated, the capacity of the energy storage device, the output power of the wind turbine generator, and the reduction duration parameter of the load that can be curtailed are changed, the output of the grid unit under multiple scenarios is simulated, and the economic parameters of the grid planning under multiple scenarios are generated.

[0051] Furthermore, in step S5, an index sequence (Y1, Y2,..., Y a ) composed of a dispatchable flexible resources, and an index sequence X = (X1, X2,..., X b ) composed of the operating planning costs of b power system plans are formed, and an analysis matrix is formed:

[0052]

[0053] And the elements of the analysis matrix are dimensionless processed to obtain an initial value matrix:

[0054]

[0055] Calculate the difference Δ ab between the a-th index and the b-th economic characteristic index in the n-th group, and obtain a difference matrix as follows:

[0056] Δ ab = |y′ a (n) - x′ b(n)| (27)

[0057]

[0058] Calculate the maximum value Δ of the two-stage range by the following two formulas max and the minimum value Δ min :

[0059] Δ max = max(maxΔ)

[0060] Δ min = min(minΔ)

[0061] Finally, calculate the grey correlation coefficient

[0062] Calculate the value coefficient λ ab (n) of the a-th index and the b-th economic characteristic index in the n-th group by the following formula:

[0063]

[0064] In the formula: ρ is the resolution coefficient, which is a number between 0 and 1;

[0065] Take the mean of each column in the correlation coefficient matrix λ ab (n) to obtain the grey correlation degree r ab :

[0066]

[0067] And denote the above formula as:

[0068] G = [ξ1, ξ2, …, ξ b .

[0069] Furthermore, in step S5, the index sequence (Y1, Y2, …, Y a ) composed of a schedulable flexible resources, and the index sequence X = (X1, X2, …, X b ) composed of the operating planning costs of b power system plans are formed, and an analysis matrix is formed:

[0070]

[0071] And perform dimensionless processing on the elements of the analysis matrix to obtain the initial value matrix:

[0072]

[0073] Calculate the difference Δ ab, and obtain the difference matrix as follows:

[0074] Δ ab = |y′ a (n) - x′ b (n)|

[0075]

[0076] Calculate the maximum value Δ of the two - level range and the minimum value Δ by the following two formulas: max and the minimum value Δ min :

[0077] Δ max = max(maxΔ)

[0078] Δ min = min(minΔ)

[0079] Finally, calculate the grey correlation coefficient

[0080] Calculate the value coefficient λ ab (n) of the a - th index and the b - th economic characteristic index in the n - th group by the following formula:

[0081]

[0082] In the formula: ρ is the resolution coefficient, which is a number between 0 and 1;

[0083] Take the mean of each column in the correlation coefficient matrix λ ab (n), and the grey correlation degree r between the a - th characteristic index and the b - th influencing factor index can be obtained: ab :

[0084]

[0085] And denote the above formula as:

[0086] G = [ξ1, ξ2, …, ξ b .

[0087] Furthermore, in step S5, first form a matrix with the operation planning cost index sequence of the power system planning:

[0088]

[0089] In the formula: x ij represents the element in matrix D; n represents the number of groups of economic cost data generated after the adjustable flexible resources change the configuration data, and b represents the number of economic characteristic indexes of the power system planning;

[0090] Subsequently, perform dimensionless processing on the elements in matrix D, as shown in the formula:

[0091]

[0092] Where: x max represents the maximum value of the j-th index, and x min represents the minimum value of the j-th index;

[0093] Finally, calculate the entropy value and entropy weight coefficient according to the following formula:

[0094] Let the dimensionless evaluation data matrix be D′ = (x′ ij ), n×b , x′ ij is an element of the evaluation matrix D′, then the entropy value calculation formula for the j-th economic cost index is as follows:

[0095]

[0096] The entropy weight coefficient calculation formula for the j-th economic cost index is as follows:

[0097]

[0098] And the above formula satisfies:

[0099]

[0100] Then the entropy weight coefficient matrix of each index is:

[0101] H = [ω1, ω2, …, ω b T .

[0102] Furthermore, in step S5, considering that the importance degrees of each index are different, the weighted average of the correlation coefficients is used as the correlation degree. Therefore, the final weighted comprehensive correlation degree is as follows:

[0103] β i = [ξ1, ξ2, …, ξ b · [ω1, ω2, …, ω b T

[0104] The sorting of the results of the above formula is the sorting of the correlation degree between the a-th index and the economic characteristic indexes of the b power system planning.

[0105] On the other hand, the present invention also discloses a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the processor is caused to execute the steps of the above method.

[0106] The advantages of the present invention are as follows:

[0107] ​​(1) When constructing the model, fully consider the current situation of large-scale dispatchable flexible resources connected to the grid, supplement and improve the mathematical model of dispatchable flexible resources in power system planning, and set the objective function and constraint function based on the mathematical model;

[0108] (2) Randomly generate multiple groups of configuration data for each dispatchable flexible resource, innovatively use the grey correlation theory to conduct multi-scenario analysis on the generated multiple groups of economic parameters. At the same time, considering the uncertainty of photovoltaic and wind power generation output, use the entropy weight method to conduct weighted analysis on the grey correlation theory to reduce the impact of discrete data on the final weighted comprehensive correlation degree;

[0109] (3) Quantify the impact degree of dispatchable flexible resources on grid planning, accurately identify the dispatchable flexible resources that have a greater impact on the economy and reliability of grid planning, which is highly applicable after promotion, and provide accurate data support for the decision-making of grid planning. Brief Description of the Drawings

[0110] Figure 1 It is the flowchart of the method of the present invention;

[0111] Figure 2 It is the block diagram of the joint planning model of dispatchable flexible resources based on robust optimization. Detailed Embodiment

[0112] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0113] As Figure 1 shown, a method for identifying dispatchable flexible resources based on grey correlation theory includes the following steps:

[0114] S1. According to the actual situation of the evaluation area, select a suitable power grid structure of the IEEE node system, such as the power grid structure of the IEEE 57-node and IEEE 118-node systems.

[0115] S2. According to the current quantity and development situation of social dispatchable flexible resources, define the key dispatchable flexible resources as five categories: load shedding, energy storage, microgrid, dispatchable new energy, and electric vehicle. Among them, although load shedding cannot achieve two-way energy flow with the power grid, it can reduce the electricity consumption of users through measures such as policy support and contract agreements, and has the ability to actively participate in power grid dispatching.

[0116] S3. The specific steps include establishing a schedulable flexible resource model, setting up the objective function and constraint function, and introducing robustness into the economic objective of the optimization model, achieving the joint optimization of robustness and economy during the optimization process, and obtaining a joint planning unit configuration plan for schedulable flexible resources with better robustness and economy. To highlight the impact of schedulable new energy on power grid planning, the thermal power units in the present invention do not participate in the setting of the objective function and constraint function of the optimization model.

[0117] S301. In the optimization model of power grid planning, the uncertain parameter is the output of distributed wind power generation. The main reason is that there is an error between the model established during the planning process and the actual value, and the prediction error of wind speed is regarded as a normal distribution random variable with a mean of 0 and a standard deviation of σ. The actual value of wind speed should be the sum of the predicted value and the prediction error. Therefore, the expression of wind speed is shown in Equation (1):

[0118]

[0119] In the formula: is the actual value of wind speed; is the predicted value of wind speed; Δv w is the prediction error of wind speed; and are the upper and lower limits of wind speed prediction.

[0120] Furthermore, the relationship between the output power of the wind turbine generator and the wind speed is shown in Equation (2):

[0121]

[0122] In the formula: P w,t is the power output of the wind turbine at time t; v w,t is the actual wind speed of the wind turbine at time t; and are the cut-in and cut-out wind speeds of the wind turbine; is the rated wind speed of the wind turbine; is the rated power generation of the wind turbine.

[0123] To a certain extent, an electric vehicle in a parked state can be regarded as a distributed energy storage in the power distribution network and participate in the grid dispatching in the form of V2G (vehicle-to-grid). The current charging methods of electric vehicles are mainly fast charging, slow charging, and battery replacement. According to investigations, only a very small number of electric vehicles use fast charging and battery replacement for power replenishment. Therefore, the electric vehicle model in this paper only targets electric vehicles using slow charging. At the same time, considering that the time for electric vehicles to participate in grid dispatching is relatively short and the self-discharge effect of the battery is relatively small, the self-discharge loss of electric vehicles is ignored in the model. That is, the model of the electric vehicle is the same as that of the energy storage device, as shown in Equation (3):

[0124]

[0125] In the formula: SOC represents the state of charge of the energy storage device; E now and E N are the current battery capacity and the rated battery capacity of the energy storage device; SOC(t) and SOC(t + 1) respectively represent the state of charge of the energy storage device before and after charging; P c and P d are the charging and discharging powers of the energy storage device respectively; η c and η d are the charging and discharging efficiencies of the energy storage device respectively; Δt is the time for the energy storage device to participate in grid dispatching.

[0126] The load that can be curtailed generally uses a contract to clarify data such as the user's load consumption, load curtailment volume, and compensation unit price. Usually, this type of load is informed in advance when the grid reliability is affected so that it can respond and curtail the load in a timely manner. Therefore, when establishing the model, it should be considered that when the sum of the user's actual load and the curtailed load is less than the predicted load, the user should be compensated; when the sum of the user's actual load and the curtailed load is greater than the predicted load, the user should be punished. Further, the penalty load volume and compensation load volume of the user in the t-hour time period are shown in Equation (4) and Equation (5):

[0127]

[0128]

[0129] In the formula: Q p,t and Q c,t are the penalty load volume and compensation load volume in the t-hour time period respectively; Q true,t 、Q req,t 、Q fore,t are the user's actual load volume, curtailed load volume, and grid predicted load volume in the t-hour time period respectively.

[0130] A microgrid mainly refers to a small-scale power grid composed of distributed power sources, energy storage devices, loads, energy conversion devices, etc. For the convenience of research, this paper only considers a microgrid system with wind power, and simplifies it to a small-scale power generation and distribution system containing only a wind power generation model, a distributed energy storage model, and a load when constructing the model.

[0131] S302. The optimization objective of this paper is to jointly plan dispatchable flexible resources such as dispatchable flexible resources and distributed energy storage devices, so as to minimize the power grid loss, operation cost and reliability economic indicators of the distribution network, and have a strong ability to absorb high-penetration wind power. Therefore, the objective function of this paper is the curtailment penalty cost, load shedding cost, power grid loss cost, upward power purchase cost, power sales revenue and power outage loss cost. The objective function is shown in Equation (6), and each part of the function is shown in Equations (7) to (14).

[0132] F = min(f1 + f2 + f3 + f4 + f5 + f6) (6)

[0133] 1) Curtailment penalty cost

[0134]

[0135] In the formula: N wi represents the number of wind turbines in this area; P waste,t represents the curtailment volume at the t-th hour; m waste,t represents the curtailment penalty unit price at the t-th hour.

[0136] 2) Load shedding cost

[0137] The load shedding cost mainly includes the compensating cost of the shed load and the penalty income, as shown in Equations and respectively:

[0138] M p,t = m p,t ·Q p,t (8)

[0139] M c,t = m c,t ·Q c,t (9)

[0140] f2 = M c,t - M p,t (10)

[0141] In the formula: M p,t and M c,t are the penalty amount and compensation amount of users in the t-th hour period respectively; m p,t and m c,t are the contractually agreed penalty and compensation amount unit prices in the t-th hour respectively.

[0142] 3) Power grid loss cost

[0143]

[0144] Where: m loss is the unit cost of network loss; E line is the set of all branches in this area; I ij,t is the current of branch ij during the t-th hour time period; R ij,t is the resistance of branch ij during the t-th hour time period.

[0145] 4) Upward power purchase cost

[0146] f4 = m buy,t ·Q buy,t (12)

[0147] Where: m buy,t is the upward power purchase unit price at the t-th hour; Q buy,t is the upward power purchase load at the t-th hour.

[0148] 5) Electricity sales revenue

[0149] f5 = m sell,t ·Q sell,t (13)

[0150] Where: m sell,t is the electricity sales unit price at the t-th hour; Q sell,t is the upward power purchase load at the t-th hour.

[0151] 6) Outage loss cost

[0152]

[0153] Where: p is the unit outage loss; D i is the average load of load node i; U i is the annual average outage time.

[0154] S303, the constraint conditions of the joint robust optimization scheduling model considering various types of dispatchable flexible resources are shown in Equations (15) to (24):

[0155] 1) Power balance constraint

[0156] P L = P w + P SOC - P req + P micro (15)

[0157] Where: P L is the load power; P w is the wind turbine power; P SOC is the power of the energy storage device; Preq Power for load shedding; P micro Power transmitted by the microgrid connected to the distribution network.

[0158] 2) Output constraint of wind turbines

[0159]

[0160] In the formula: and respectively represent the upper and lower limits of the output of wind turbines.

[0161] 3) Energy storage device constraint

[0162] The constraints of the energy storage device are mainly the state-of-charge constraint shown in the formula and the charge-discharge constraint shown in the formula.

[0163] SOC min ≤SOC(t + 1) ≤ SOC max (17)

[0164]

[0165] In the formula: SOC max and SOC min respectively represent the upper and lower limits of the state of charge of the energy storage device; SOC(t + 1) represents the power state after the charge and discharge of the energy storage device; and represent the upper and lower limits of the charging power of the energy storage device; and represent the upper and lower limits of the discharging power of the energy storage device; P t EVC and P t EVD represent the charge and discharge power of the energy storage device in the t-th hour time period.

[0166] 4) Power flow constraint

[0167]

[0168]

[0169]

[0170]

[0171] In the formula: and represent the active and reactive net power of node i at time t; and represent the active and reactive power supplied by the main grid at node i at time t; and denote the active and reactive power supplied by the wind turbine at node i at time t; and denote the active and reactive power supplied by the microgrid at node i at time t; and denote the active and reactive power of the load at node i at time t; U i,t is the voltage amplitude of node i at time t; G ij and B ij are the conductance and susceptance of line ij, B ij,s is the shunt susceptance of line ij to the ground; θ ij,t is the voltage phase angle difference between nodes ij at time t.

[0172] (5) Interruptible load constraint

[0173] According to the actual situation, it is necessary to constrain the reduction capacity and reduction frequency of the interruptible load to achieve the economic operation of the distribution network and the microgrid. The constraint on the load reduction capacity is shown in the formula, and the constraint on the load reduction frequency is shown in the formula.

[0174]

[0175]

[0176] In the formula: represents the maximum load reduction capacity within the t-th hour time period; μ m,t is a 0-1 variable representing the status variable of user m's response to scheduling within the t-th hour time period. When μ m,t is 1, it means that the user responds to the scheduling. When μ m,t is 0, it means that the user does not respond to the scheduling; represents the maximum load reduction frequency that user m accepts within the t-th hour time period.

[0177] S304, The calculation steps of the joint planning model of dispatchable flexible resources based on robust optimization are as Figure 2 shown, mainly including:

[0178] Module 101 is used to construct the uncertain output interval of the wind turbine generator, that is, within the short-term scheduling, by predicting the wind speed, an uncertain output interval of the wind turbine generator is established considering errors, and at the same time, the actual wind speed is calculated At the same time, parameters such as the compensation and penalty unit prices for load reduction and the candidate interval of the load nodes of each flexible resource are set;

[0179] Module 102 constructs a grid scheduling model that realizes the coordinated optimization of robustness and economy based on parameters such as the constructed wind power uncertain output interval and the candidate interval parameters of load nodes, etc., and minimizes the sum of curtailment penalty cost, load shedding cost, network loss cost, upward power purchase cost, power sales revenue, and power outage loss cost on the premise of considering the constraint conditions;

[0180] Module 103 obtains the optimal output combination of load shedding, energy storage, microgrid, dispatchable new energy, electric vehicles, and external grid with the minimum total operating cost of the power grid in this area by solving the objective function.

[0181] S4. On the premise of ensuring the reliable operation of the power grid, randomly generate parameters such as the capacity of the energy storage device, the load nodes of the wind turbine, and the curtailment duration of the load shedding, to obtain a multi-scenario model of power grid planning including the optimal scheduling model in step S3. Substitute the generated power grid planning model into the objective function of the S3 model for calculation to obtain multiple groups of economic parameters of power grid planning including curtailment penalty cost, load shedding cost, network loss cost, upward power purchase cost, power sales revenue, and power outage loss cost.

[0182] S5. This paper intends to use the grey correlation theory based on the entropy weight method to analyze the correlation degree of the economic parameters of the power system planning generated in step S4. The finally generated comprehensive weighted correlation degree is a measure of the correlation size between two systems or two elements, mainly including:

[0183] S501. To calculate the correlation degree, first determine the analysis sequence, secondly, perform dimensionless processing on the original sequence, then calculate the correlation coefficient, thereby calculating the correlation degree, and arranging the correlation order. Considering the different importance degrees of each index, the weighted average value of the correlation coefficients is often used as the correlation degree.

[0184] 1) Form an analysis matrix

[0185] The index sequence (Y1, Y2, …, Y a ) composed of a dispatchable flexible resources, and the index sequence X = (X1, X2, …, X b ) composed of b operating planning costs of power system planning. The sequence matrix is:

[0186]

[0187] In the formula: n is an integer, representing the number of groups of economic cost data generated after changing the configuration data of dispatchable flexible resources.

[0188] 2) Generate an initial value matrix

[0189] Therefore, before quantitative analysis, it is necessary to perform dimensionless processing on the original data. The commonly used methods are initial value method and mean value method, and the initial value matrix is obtained.

[0190]

[0191] 3) Generate the difference matrix

[0192] Perform a difference operation on the elements in the initial value matrix according to Equation (27) to calculate the difference Δ between the a-th index and the b-th economic characteristic index in the n-th group ab , and thus obtain the difference matrix Δ

[0193] Δ ab = |y′ a (n) - x′ b (n)| (27)

[0194]

[0195] Select the maximum value Δ max and the minimum value Δ min of the two-stage range from the formula and the formula:

[0196] Δ max = max(maxΔ) (29)

[0197] Δ min = min(minΔ) (30)

[0198] 4) Calculation of the grey correlation coefficient

[0199] Calculate the correlation coefficient λ ab (n) of the a-th index and the b-th economic characteristic index in the n-th group according to Equation (31) as follows:

[0200]

[0201] In the formula: ρ is the resolution coefficient, which is a number between 0 and 1, and generally takes 0.5

[0202] Take the mean value of each column in the correlation coefficient matrix λ ab (n) to obtain the grey correlation degree r ab of the a-th index and the b-th economic characteristic index:

[0203]

[0204] And record the formula as:

[0205] G = [ξ1, ξ2, …, ξ b (33)

[0206] S502. Since the importance of each evaluation index for the evaluation object is different, and the amount of information reflected by each evaluation index about the evaluation object is also different, we need to assign weight coefficients to each index according to its characteristics. Generally, the methods of weight assignment are mainly divided into two categories: objective weight assignment method and subjective weight assignment method. In this paper, the entropy weight method in the objective weight assignment method is adopted. This method can use the information reflected by sufficient objective data for weight assignment, and can highlight the importance of different types of schedulable flexible resources for power grid planning.

[0207] To calculate the index weights by the entropy weight method, first, a matrix is formed with the influencing factor sequence composed of schedulable flexible resources, the indexes are unified, and the entropy value and entropy weight coefficient of the dimensionless data are calculated.

[0208] 1) The operation planning cost index sequence of the power system planning forms a matrix as shown in Equation (34):

[0209]

[0210] In the formula: x ij represents the element in matrix D, where; n represents the number of groups of economic cost data generated after the configuration data of schedulable flexible resources is changed, and b represents the number of economic characteristic indexes of the power system planning.

[0211] 2) Matrix preprocessing

[0212] The economic characteristic indexes of the power system planning in this paper are all minimum-type indexes, and the smaller the index, the better. Therefore, it is not necessary to unify the indexes. At the same time, considering that the units and orders of magnitude of each economic characteristic index are different, the situation of "big numbers swallowing small numbers" may occur. Therefore, we need to make the dimensionless processing of these index data. Here, the range method is used for dimensionless processing, as shown in the formula

[0213]

[0214] In the formula: x max represents the maximum value of the j-th index, and x min represents the minimum value of the j-th index.

[0215] 3) Calculate the entropy value and entropy weight coefficient

[0216] Let the dimensionless evaluation data matrix be D′ = (x′ ij ), n×b x′ ij is the element of the evaluation matrix D′, then the entropy value calculation formula of its j-th economic cost index is as follows:

[0217]

[0218] The calculation formula for the entropy weight coefficient of the j-th economic cost index is as follows:

[0219]

[0220] And the formula satisfies:

[0221]

[0222] Then the entropy weight coefficient matrix of each index is:

[0223] H = [ω1, ω2, …, ω b T (39)

[0224] S503. Since the information reflected by the correlation coefficient is not clear enough, the average value of the correlation coefficient, that is, the correlation degree, is used as a measure scale to reflect the correlation degree between the comparison sequence and the reference sequence. Considering the different importance degrees of each index, the weighted average value of the correlation coefficient is often used as the correlation degree. Therefore, the final weighted comprehensive correlation degree is the product of formula and formula, as follows:

[0225] β i = [ξ1, ξ2, …, ξ b ·[ω1, ω2, …, ω b T (40)

[0226] The magnitude sorting of formula is the correlation degree sorting of the a-th index and the economic characteristic indexes of the b power system planning.

[0227] In summary, a method for identifying key dispatchable flexible resources affecting power grid planning based on grey correlation theory in an embodiment of the present invention relates to the fields of power system planning and operation and grey correlation analysis theory. The method includes the following steps: establishing a power grid topological network structure according to the scale of the distribution network in the analysis area, and selecting various dispatchable flexible resources in the distribution network to establish a mathematical model; setting a system objective function and constraint functions according to the mathematical model, and generating an optimal unit output combination that realizes the coordinated optimization of robustness and economy; randomly generating configuration data such as the capacity and unit output of dispatchable flexible resources on the premise of meeting the power grid reliability, substituting them into the objective function to calculate various costs and forming multiple groups of data; calculating the correlation degree of multi-scenario data by using grey correlation theory based on the entropy weight method to obtain the weighted comprehensive correlation degree of each dispatchable flexible resource; the magnitude sorting result of each weighted comprehensive correlation degree is the influence degree of the dispatchable flexible resource on power grid planning. The identification method proposed by the present invention overcomes the problem of identifying the influence of dispatchable flexible resources in current power system planning, provides a method for quantifying the influence degree of dispatchable flexible resources on power grid planning, and improves the accuracy and efficiency of power system planning decisions. ​​

[0228] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which when executed by a processor causes the processor to execute the steps of the above method.

[0229] In yet another aspect, the present invention also discloses a computer device including a memory and a processor, the memory storing a computer program, which when executed by the processor causes the processor to execute the steps of the above method.

[0230] In another embodiment provided by the present application, there is also provided a computer program product containing instructions, which when run on a computer causes the computer to execute any one of the schedulable flexible resource identification methods based on the grey correlation theory in the above embodiments.

[0231] It can be understood that the system provided by the embodiments of the present invention corresponds to the method provided by the embodiments of the present invention. For the explanations, examples and beneficial effects of related content, reference can be made to the corresponding parts in the above method.

[0232] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it may include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database or other medium used in the various embodiments provided by the present application may include non-volatile and / or volatile memories. Non-volatile memories may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memories may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0233] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0234] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A method for identifying schedulable flexible resources based on grey relational theory, characterized in that, It includes the following steps: S1. Establish a power grid topological network structure according to the scale of the distribution network in the analysis area; S2. Screen the dispatchable flexible resources according to the output and scale of the dispatchable flexible resources in the distribution network of the analysis area, and establish a mathematical model; S3. Set the objective function and constraint function of the system according to the mathematical model, and generate the optimal unit output combination for realizing the coordinated optimization of robustness and economy; S4. On the premise of ensuring the stable operation of the power grid, randomly generate multiple groups of configuration data of each dispatchable flexible resource, substitute them into the objective function in step S3 for calculation, and generate the economic parameters of the power grid planning under multiple scenarios; S5. Analyze the economic parameters of the power grid planning in step S4 by using the grey correlation theory based on the entropy weight method, and judge the influence degree of each flexible resource on the power grid planning through the size of the final weighted comprehensive correlation degree; In step S2, the modeling of the dispatchable flexible resources is divided into three parts, namely: 1) Uncertain output model of wind turbines: （1） Where: is the actual value of the wind speed; is the predicted value of the wind speed; is the prediction error of the wind speed; and are the upper and lower limits of wind speed prediction; The relationship between the output power of the wind turbine generator set and the wind speed is shown in the following formula: （2） Where: is the power output of the wind turbine at time t; is the actual wind speed of the wind turbine at time t; and are the cut-in and cut-out wind speeds of the wind turbine; is the rated wind speed of the wind turbine; is the rated power generation of the wind turbine; 2) Distributed energy storage model: （3） Wherein: represents the state of charge of the energy storage device; and are the current battery capacity and the rated battery capacity of the energy storage device; and respectively represent the state of charge of the energy storage device before and after charging; and are the charge and discharge powers of the energy storage device respectively; and are the charge and discharge efficiencies of the energy storage device respectively; is the time for the energy storage device to participate in power grid dispatching; 3) Curtailable load model: （4） （5） Wherein: and are respectively the penalty load amount and the compensation load amount in the t-th hour time period; , , are respectively the actual load amount of the user, the reduced load amount, and the power grid predicted load amount in the t-th hour time period; In step S3, the objective functions for the modeling of the dispatchable flexible resources are divided into six, namely: Objective function : （7） Wherein: represents the number of wind turbines in the slice area; represents the curtailment volume at the t-th hour; represents the curtailment penalty unit price at the t-th hour; Objective function : （8） （9） （10） Where: and are the penalty amount and compensation amount of the user in the t-th hour period respectively; and are the unit prices of the penalty and compensation amounts stipulated in the contract in the t-th hour respectively; Objective function :[[]]END]] （11） Wherein: is the unit cost of network loss; is the set of all branches in this area; is the current of branch during the t-th hour time period; is the resistance of branch during the t-th hour time period; Objective function : （12） where: is the unit price of upward power purchase at the t-th hour; is the upward power purchase load at the t-th hour; Objective function : （13） Where: is the electricity selling unit price at the t-th hour; is the upward power purchase load at the t-th hour; Objective function : （14） Wherein: is the unit outage loss; is the load node 's average load; is the annual average power outage time; The constraint functions in step S3 include 1) Power balance constraint （15） Wherein: is the load power; is the power of the wind turbine; is the power of the energy storage device; is the power of load shedding; is the power transmitted by the microgrid connected to the distribution network; 2) Output constraint of wind turbines (16) Wherein: and respectively represent the upper and lower limits of the output of the wind turbine generator set; 3) Energy storage device constraint The constraints of the energy storage device are mainly the state of charge constraint shown in formula (17) and the charge and discharge constraint shown in formula (18); (17) (18) In the formula: and respectively represent the upper and lower limits of the state of charge of the energy storage device; represents the power state after the charge and discharge of the energy storage device are completed; and represent the upper and lower limits of the charging power of the energy storage device; and represent the upper and lower limits of the discharging power of the energy storage device; and represent the charge and discharge power of the energy storage device within the t-th hour time period; 4) Power flow constraint (19) (20) (21) (22) Where: and represent the active and reactive net power of node at time t; and represent the active and reactive power supplied by the main grid to node at time t; and represent the active and reactive power supplied by the wind turbine to node at time t; and represent the active and reactive power supplied by the microgrid to node at time t; and represent the active and reactive power of the load at node at time t; is the voltage amplitude of node at time t; and are the conductance and susceptance of line ; is the shunt susceptance to ground of line ; is the voltage phase angle difference of node at time t; 5) Interruptible load constraint According to the actual situation, it is necessary to constrain the reduction capacity and reduction frequency of the interruptible load to realize the economic operation of the distribution network and the microgrid. Among them, the constraint on the load reduction capacity is shown in formula (23), and the constraint on the load reduction frequency is shown in formula (24); (23) (24) Wherein: represents the maximum load curtailment capacity during the t-th hour time period; is a 0-1 variable, representing the status variable of user m's response to scheduling during the t-th hour time period. When is 1, it means the user responds to the scheduling. When is 0, it means the user does not respond to the scheduling; represents the maximum load curtailment frequency received by user m during the t-th hour time period.

2. The method for identifying schedulable flexible resources based on the grey relational theory according to claim 1, characterized in that, In step S2, the key dispatchable flexible resources are defined as five categories: curtailable load, energy storage, microgrid, dispatchable new energy, and electric vehicle.

3. The method for identifying schedulable flexible resources based on the grey relational theory according to claim 1, characterized in that In step S4, randomly generate the configuration data of the dispatchable flexible resources, change the capacity of the energy storage device, the output power of the wind turbine generator set, and the reduction duration parameter of the curtailable load, simulate the output situation of the power grid units under multiple scenarios, and generate the economic parameters of the power grid planning under multiple scenarios.

4. A method for identifying schedulable flexible resources based on the grey relational theory according to claim 1, characterized in that, In step S5, the index sequence composed of schedulable flexible resources , and the index sequence composed of operating planning costs of the power system planning , and an analysis matrix is formed: （25） And dimensionless process the elements of the analysis matrix to obtain the initial value matrix: （26） Calculate the difference between the a-th index and the b-th economic characteristic index in the n-th group , and obtain the difference matrix as follows: （27） （28） Calculate the maximum and minimum values of the two-stage range using the following two formulas and minimum value : Finally, calculate the grey correlation coefficient The value coefficient of the a-th index and the b-th economic characteristic index in the n-th group is calculated by the following formula , as follows: Wherein: is a number with a resolution coefficient between 0 and 1; The correlation coefficient matrix is averaged for each column to obtain the grey correlation degree between the a-th characteristic index and the b-th influencing factor index as follows: And denote the above formula as: 。 5. The method for identifying schedulable flexible resources based on the grey relational theory according to claim 4, wherein In step S5, first form a matrix with the operation planning cost index sequence of the power system planning: In the formula: represents the element in matrix D; n represents the number of groups of economic cost data generated after the schedulable flexible resources change the configuration data, and b represents the number of economic characteristic indexes of the power system planning. Subsequently, dimensionless process the elements in matrix D as shown in the formula: Wherein: represents the maximum value of the j-th index, represents the minimum value of the j-th index; Finally, calculate the entropy value and entropy weight coefficient according to the following formula: Let the dimensionless evaluation data matrix be , be the evaluation matrix . The entropy value calculation formula for the j-th economic cost index is as follows: The calculation formula of the entropy weight coefficient of the j-th economic cost index is as follows: And the above formula satisfies: Then the entropy weight coefficient matrix of each index is: 。 6. The method for identifying schedulable flexible resources based on the grey relational theory according to claim 5, wherein In step S5, considering that the importance degrees of each index are different, take the weighted average value of the correlation coefficients as the correlation degree. Therefore, the final weighted comprehensive correlation degree is as follows: The sorting of the results of the above formula is the correlation degree sorting of the a-th index and the economic characteristic indexes of the b power system planning.

7. A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.

Images