A power distribution energy grid optimization division method considering source-load complementary characteristics

Abstract

The application discloses a power distribution energy grid optimization division method considering source-load uncertainty and complementary characteristics. The main content comprises in-depth analysis on operation uncertainty characteristics of distributed photovoltaic, distributed energy station, electric vehicle charging station and multi-element load of the power distribution system, establishment of probability operation models of various source-load elements, acquisition of regional net load characteristic curves based on three-point estimation method and Gram-Chalier series expansion method, superposition of multi-type source-load probability energy curves, acquisition of credible load density of power supply range of each substation, calculation of equivalent credible power supply radius and reference substation boundary distance of each substation, calculation of source-load matching degree between adjacent substations and updating of substation boundary distance, acquisition of an energy grid division scheme set by using a K-means++ clustering algorithm, and scheme optimization according to relevant indexes. By the method, the connection between each station in the energy grid is closely connected by simultaneously considering the uncertainty of multi-type source-load, the coupling mapping relationship among multi-energy loads and the influence of the complementary characteristics between the source and load on the connection between substations, and the economy and effectiveness of the grid planning system method for the multi-energy system are improved.

Description

Technical Field

[0001] This invention relates to a method for optimizing the partitioning of power distribution energy grids that considers source-load uncertainty and source-load complementarity. It is applicable to the optimization planning problem of power distribution systems and belongs to the field of power distribution networks. Background Technology

[0002] Against the backdrop of the construction of new multi-energy coupled distribution systems, the existing grid-based planning system can no longer support power grid construction. Specifically, on the source side of the new distribution system, the uncertainty and uncontrollability of massive distributed renewable energy output will lead to low efficiency of power grid assets, and the existing planning system and technology cannot solve this problem well. On the load side, the access of diverse loads such as electric vehicles and user-side multi-energy coupled systems will add new burdens to the distribution network. At the same time, the rational use of the active characteristics of load-side resources can improve the flexibility of the distribution network, but the existing planning system cannot take into account the impact of diverse loads.

[0003] To address the aforementioned issues, scholars both domestically and internationally have conducted research on source-grid-load-storage game-theoretic operational strategies for different operating entities. Some scholars have focused on power supply blocks, breaking the current situation where power supply block division is constrained by administrative divisions and fully considering electrical influences during power supply block division. Meanwhile, some researchers have focused on expanding the power supply range of feeders, proposing a power supply unit division method based on relaxed load rate to address the problem of low capacity utilization of inter-substation interconnection feeders. Furthermore, some scholars have considered the complementary characteristics between loads and proposed power supply unit division methods to improve feeder utilization.

[0004] Analysis of the above studies reveals that current research on grid-based planning systems is primarily limited to the power grid side, failing to fully consider the deep coupling relationship between heating and cooling loads, and between the heating / cooling grid and the power grid. Furthermore, current research does not adequately consider the impact of demand response resources on power grid partitioning, resulting in overly conservative conclusions that are detrimental to reducing power grid construction costs. Therefore, this invention, building upon existing research, proposes an optimized partitioning method for distribution energy grids that considers source-load uncertainty and complementary characteristics. This method fulfills the multi-energy coupling requirements of power grid construction and improves equipment utilization efficiency. Summary of the Invention

[0005] To address the shortcomings and deficiencies of existing technologies, this invention aims to propose a method for optimizing the partitioning of power distribution energy grids that considers source-load uncertainty and complementary characteristics. This method, by simultaneously considering the uncertainties of multiple types of source loads, the coupling mapping relationships between multiple energy loads, and the impact of the complementary characteristics of source loads on inter-substation connections, strengthens the connections between stations within the energy grid and improves the economy and effectiveness of grid-based planning methods for multi-energy systems.

[0006] This invention provides a method for optimizing the partitioning of power distribution energy grids considering source-load uncertainty and complementary characteristics. The specific steps include:

[0007] (1) Based on the historical output and energy consumption characteristics of power load, the cold and heat loads are converted to the power side through the cold and heat load power supply model, and a source load power side energy consumption probability model is established.

[0008] (2) Based on the power consumption probability model of the source load, the net load curve and the power output / energy consumption curve of each source load under the α confidence level are obtained in each power supply range, and the regional credible load density is calculated.

[0009] (3) Based on the credible load density, calculate the equivalent credible power supply radius of each substation and the distance between the benchmark substation boundary, calculate the source-load matching degree between adjacent substations and update the substation boundary distance.

[0010] (4) Determine the range of grid numbers for the partition scheme based on the energy grid number partitioning method, obtain the results of different numbers of energy grid partitioning schemes, and optimize the scheme according to the evaluation index.

[0011] The specific process of step (1) is as follows:

[0012] Based on historical power output / energy consumption data of various elements within the area to be planned, establish probabilistic models for power output / energy consumption of multiple types of sources and loads, including:

[0013] 1) Probabilistic model for distributed photovoltaic power output:

[0014]

[0015] Where Γ(·) is the gamma function; P PV P represents the output value of distributed photovoltaic power. PV1 This represents the theoretical maximum output of distributed photovoltaic power; α and β are both shape parameters, whose magnitudes can be approximated using historical output data of distributed photovoltaic power, as shown in the following formula:

[0016]

[0017]

[0018] Where, μ PV The average historical power output of distributed photovoltaic systems; Variance of historical output data for distributed photovoltaic power.

[0019] 2) Energy consumption probability models for heating and cooling loads, electric vehicle charging stations, and conventional multi-type electrical loads:

[0020]

[0021] Among them, SL Energy consumption data for load; μ L σ L , These represent the mean, standard deviation, and variance of historical load energy consumption data, respectively.

[0022] 3) Probabilistic model for energy supply and consumption by distributed energy stations and grids:

[0023] f(P DES )=f(α DES )g(f(S L ))

[0024]

[0025] Where, f(α) DES ) and f(S L ) are random variables α DES and S L The probability density function; g(f(S) L α is the conversion function between heating / cooling load and electrical load, and it is a linear function; DES The proportion of power supplied to distributed energy stations.

[0026] The method for obtaining the probability net load curve based on the three-point estimation method and Gram-Charlier series expansion described in step (2) is as follows:

[0027] 1) Based on the distributed photovoltaic power output and time-series energy consumption model of various new elements established in step (1), n ​​in the following steps indicates that there are n access elements in the area to be calculated;

[0028] 2) In the independent probability distribution space Y = [y1, y2, ..., y n In the above, the three-point estimation method is used to calculate the sampled value y at each point. i,k and the weight p of the corresponding point i,k , obtain S Y :

[0029] S Y =[Y 1,1 ,Y 1,2 ,Y 2,1 ,Y 2,2 ,...,Y n,1 ,Y n,2 ,Y 2n+1 ] T

[0030] Where Y i,k With Y 2n+1 The calculation formula is as follows:

[0031]

[0032] 3) Obtain the relevant sample matrix S through matrix transformation. Z :

[0033] S Z =[Z 1,1 Z 1,2 Z 2,1 Z 2,2 ,...,Z n,1 Z n,2 Z 2n+1 ] T

[0034] Z i,k With Z 2n+1 The calculation formula is as follows:

[0035]

[0036] 4) The sample matrix S of the standard normal distribution Z Transformed into a sample matrix S in the actual distribution space X .

[0037] S X =[X 1,1 ,X 1,2 ,X 2,1 ,X 2,2 ,...,X n,1 ,X n,2 ,X 2n+1 ] T

[0038] Where X i,k With X 2n+1 The calculation formula is as follows:

[0039]

[0040] 5) For S X The elements of each row in the matrix are summed to obtain the sample matrix W = [W1, W2, ..., W...]. 2n+1 ] T The mean and variance of the net load random variable can be calculated using the three-point estimation method.

[0041] 6) Based on the moments of the net load random variable W, the probability density function f(W) and probability distribution function F(W) of the net load W are obtained by using the Gram-Charlier series expansion method.

[0042] The method for obtaining the reliable load density of the region described in step (2) is as follows:

[0043] ρ f,i(W)=f i (W)A trans,i

[0044] ρ F,i (W)=F i (W)A trans,i

[0045] Where, ρ f,i (W) and ρ F,i (W) represent the probability density function and probability distribution function, respectively, of the load density within the power supply range of substation i; A trans,i The actual area within the power supply range of substation i.

[0046] The method for calculating the equivalent reliable power supply radius described in step (3) specifically includes the following steps:

[0047] 1) Calculate the corresponding probability density function and probability distribution function using the method for calculating the equivalent power supply radius probability density function and probability distribution function:

[0048] R f,i (W)=ρ f,i (W)S trans,i ·η trans

[0049] R F,i (W)=ρ F,i (W)S trans,i ·η trans

[0050] Among them, R f,i (W) and R F,i (W) represent the probability density function and probability distribution function, respectively, that the equivalent power supply radius of substation i follows; S trans The construction capacity of substation i; η trans is the maximum load rate of substation i.

[0051] 2) By using numerical transformation methods under α confidence levels, the uncertain problem is transformed into a deterministic problem:

[0052] Based on the probability density function of the equivalent power supply radius, the upper limit of the equivalent power supply radius of the substation under this confidence level can be obtained from the position of the α quantile and used in subsequent calculations. The specific expression is as follows:

[0053] Pr{R trans,i}≥α

[0054]

[0055] Among them, R trans,i Let be the reliable equivalent power supply radius of substation i, and take the lower limit value in actual calculation.

[0056] The substation boundary spacing update calculation method described in step (3) specifically includes the following steps:

[0057] 1) Calculate the reference boundary distance between each substation using the reference boundary distance calculation method:

[0058] The reference boundary distance is physically represented as the shortest distance between the equivalent circular power supply areas of two substations. It is directly affected by two factors: the reliable equivalent power supply radius of the two substations and the center location of the substations. The specific calculation formula is as follows:

[0059]

[0060] Where D(i,j) is the reference boundary distance between substation i and substation j; (x i ,y i ) and (x j ,y j ) represent the coordinates of substation i and substation j, respectively.

[0061] 2) Using the forward conversion method for reference boundary spacing, the reference boundary spacing values ​​between substations are converted to positive values:

[0062] Since the above-mentioned reference boundary spacing has three possible values, it needs to be processed into a consistently positive value while ensuring that the order of the boundary spacings remains unchanged. The specific calculation formula is as follows:

[0063]

[0064] 3) Based on the positive reference boundary spacing, the updated boundary spacing of the substation is calculated using the boundary spacing update method that considers source-load matching:

[0065] The constant boundary spacing is numerically updated based on the regional source-load matching degree that may be transferred between any adjacent substations, in order to shorten the boundary spacing of substations with better source-load matching, and ensure that substations assigned to the same energy grid take into account both interconnection characteristics and source-load matching characteristics.

[0066] Step (3) describes a boundary spacing update method that considers source-load matching, specifically including:

[0067] 1) Using the method for estimating the transferable area between substation i and adjacent substations, the transferable area is estimated:

[0068] ① Connect the endpoints of the power supply range of substation i with the endpoints of the boundary lines of all surrounding substations' power supply ranges, and calculate the sum of all straight boundary distances l and l'. m ;

[0069] ② Calculate the angle of the area that can be transferred to the power grid based on the proportion of the distance to the adjacent straight boundary. The specific calculation formula is as follows:

[0070]

[0071] Where, β i,j For the angle of the area that can be supplied; l i,j The straight-line boundary distance between substation i and substation j;

[0072] ③ Using the connecting line between any two substations as the center line, divide the substation into several areas according to the range angle β. The area within substation i and between substation j that can be transferred to power supply is denoted as A. i,j ;

[0073] ④ Repeat the above process for all substations until all transferable areas within the region have been divided.

[0074] 2) Calculate the curve matching degree of the transferable supply area using the curve matching degree calculation method: ① Put all A i,j With A j,i The system is combined to form an inter-station power transfer zone, and the net power probability curve for this zone is calculated. The specific calculation formula is shown below:

[0075] Pr{F P,t (W)}≥α

[0076]

[0077] Among them, P W,t For the time-series net power curve of this region, the lower limit value is taken during calculation; F P,t (W) is the probability distribution function of the sum of source load characteristics in this region.

[0078] ②Based on the net power curve obtained at confidence level α, calculate the source-load matching degree of this region. The specific calculation formula is as follows:

[0079]

[0080] Where, σ i,j The degree of source-load matching is expressed as standard deviation; T is the evaluation period; P t Let be the net power at time t; P be the average net power over the evaluation period.

[0081] ③ Calculate the updated boundary spacing for substations participating in clustering using the boundary spacing update method:

[0082]

[0083]

[0084] in, The boundary spacing update coefficient, with a value range of [0.5, 1]; D match (i,j) represents the update boundary spacing considering source load matching.

[0085] Step (4) of obtaining the energy grid partitioning scheme set and scheme optimization based on the K-Means++ clustering algorithm specifically includes:

[0086] 1) Calculate the range of energy grid divisions. Where m is the number of substations.

[0087] 2) For each k, use the K-Means++ cluster center initialization method to generate a cluster center group;

[0088] 3) Grid division target number k and its corresponding cluster centers, and use K-Means++ clustering algorithm to cluster substations to form k energy grids;

[0089] 4) Optimize the set of alternative energy grid schemes:

[0090] ① Calculate the minimum average boundary distance b1(i,j) from the j-th substation in the i-th energy grid to the substations in each other energy unit, and the average boundary distance b2(i,j) within the energy grid. The calculation formula is as follows:

[0091]

[0092]

[0093] Where, n i Let be the number of substations in the i-th energy grid.

[0094] ② Calculate the structural indices of the clustering results

[0095] ③ Calculate the load balancing index between grids

[0096] ④ The energy grid partitioning scheme is optimized based on the objectives of having a large BWP index and a small σ index. Attached Figure Description

[0097] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0098] Figure 1This is a flowchart illustrating a power grid optimization partitioning method that considers source-load uncertainty and complementary characteristics, as proposed in this embodiment.

[0099] Figure 2 This is a schematic diagram of the probability net load curve for the area to be planned in this embodiment;

[0100] Figure 3 This is a schematic diagram of the energy grid partitioning results obtained by the power distribution energy grid optimization partitioning method that considers source-load uncertainty and complementary characteristics in this implementation. Detailed Implementation

[0101] To make the structure and advantages of the present invention clearer, the structure of the present invention will be further described below with reference to the accompanying drawings.

[0102] like Figure 1 As shown, a method for optimizing the partitioning of a power distribution energy grid, considering source-load uncertainty and complementary characteristics, includes the following steps:

[0103] (1) Based on the historical output and energy consumption characteristics of power load, the cold and heat loads are converted to the power side through the cold and heat load power supply model, and a source load power side energy consumption probability model is established.

[0104] (2) Based on the power consumption probability model of the source load, the net load curve and the power output / energy consumption curve of each source load under the α confidence level are obtained in each power supply range, and the regional credible load density is calculated.

[0105] (3) Based on the credible load density, calculate the equivalent credible power supply radius of each substation and the distance between the benchmark substation boundary, calculate the source-load matching degree between adjacent substations and update the substation boundary distance.

[0106] (4) Determine the range of grid numbers for the partition scheme based on the energy grid number partitioning method, obtain the results of different numbers of energy grid partitioning schemes, and optimize the scheme according to the evaluation index.

[0107] I. The specific process of step (1) is as follows:

[0108] Based on historical power output / energy consumption data of various elements within the area to be planned, establish probabilistic models for power output / energy consumption of multiple types of sources and loads, including:

[0109] 1) Probabilistic model for distributed photovoltaic power output:

[0110]

[0111] Where Γ(·) is the gamma function; P PV P represents the output value of distributed photovoltaic power. PV1This represents the theoretical maximum output of distributed photovoltaic power; α and β are both shape parameters, whose magnitudes can be approximated using historical output data of distributed photovoltaic power, as shown in the following formula:

[0112]

[0113]

[0114] Where, μ PV The average historical power output of distributed photovoltaic systems; Variance of historical output data for distributed photovoltaic power.

[0115] 2) Energy consumption probability models for heating and cooling loads, electric vehicle charging stations, and conventional multi-type electrical loads:

[0116]

[0117] Among them, S L Energy consumption data for load; μ L σ L , These represent the mean, standard deviation, and variance of historical load energy consumption data, respectively.

[0118] 3) Probabilistic model for energy supply and consumption by distributed energy stations and grids:

[0119] f(P DES )=f(α DES )g(f(S L ))

[0120]

[0121] Where, f(α) DES ) and f(S L ) are random variables α DES and S L The probability density function; g(f(S) L α is the conversion function between heating / cooling load and electrical load, and it is a linear function; DES The proportion of power supplied to distributed energy stations.

[0122] Step (1) can obtain the probabilistic energy consumption model and power output model of various source load elements in the area to be planned. This model serves as an important boundary condition in step (2), which can enable the planning scheme to fully consider the uncertainty of various source loads and improve the planning effect.

[0123] II. The method for obtaining the probability net load curve based on the three-point estimation method and Gram-Charlier series expansion described in step (2) is as follows:

[0124] 1) Based on the distributed photovoltaic power output and time-series energy consumption model of various new elements established in step (1), n ​​in the following steps indicates that there are n access elements in the area to be calculated;

[0125] 2) In the independent probability distribution space Y = [y1, y2, ..., y n In the above, the three-point estimation method is used to calculate the sampled value y at each point. i,k and the weight p of the corresponding point i,k , obtain S Y :

[0126] S Y =[Y 1,1 ,Y 1,2 ,Y 2,1 ,Y 2,2 ,...,Y n,1 ,Y n,2 ,Y 2n+1 ] T

[0127] Where Y i,k With Y 2n+1 The calculation formula is as follows:

[0128]

[0129] 3) Obtain the relevant sample matrix S through matrix transformation. Z :

[0130] S Z =[Z 1,1 Z 1,2 Z 2,1 Z 2,2 ,...,Z n,1 Z n,2 Z 2n+1 ] T

[0131] Z i,k With Z 2n+1 The calculation formula is as follows:

[0132]

[0133] 4) The sample matrix S of the standard normal distribution Z Transformed into a sample matrix S in the actual distribution space X .

[0134] S X =[X 1,1 ,X 1,2 ,X 2,1 ,X 2,2 ,...,X n,1 ,Xn,2 ,X 2n+1 ] T

[0135] Where X i,k With X 2n+1 The calculation formula is as follows:

[0136]

[0137] 5) For S X The elements of each row in the matrix are summed to obtain the sample matrix W = [W1, W2, ..., W...]. 2n+1 ] T The mean and variance of the net load random variable can be calculated using the three-point estimation method.

[0138] 6) Based on the moments of the net load random variable W, the probability density function f(W) and probability distribution function F(W) of the net load W are obtained by using the Gram-Charlier series expansion method.

[0139] The method for obtaining the reliable load density of the region described in step (2) is as follows:

[0140] ρ f,i (W)=f i (W) / A trans,i

[0141] ρ F,i (W)=F i (W) / A trans,i

[0142] Where, ρ f,i (W) and ρ F,i (W) represent the probability density function and probability distribution function, respectively, of the load density within the power supply range of substation i; A trans,i The actual area within the power supply range of substation i.

[0143] Step (2) can organically unify the source load elements that follow different distributions within the area to be planned, introduce uncertainty in the load density stage, and enhance the actual value and adaptability of the planning.

[0144] III. The method for calculating the reliable power supply radius described in step (3) specifically includes:

[0145] 1) Calculation method of probability density function and probability distribution function of equivalent power supply radius

[0146] R f,i (W)=ρ f,i (W)S trans,i ·η trans

[0147] R F,i (W)=ρ F,i (W)S trans,i ·η trans

[0148] Among them, R f,i (W) and R F,i (W) represent the probability density function and probability distribution function, respectively, that the equivalent power supply radius of substation i follows; S trans The construction capacity of substation i; η trans is the maximum load rate of substation i.

[0149] 2) Numerical transformation under α confidence level

[0150] Based on the probability density function of the equivalent power supply radius, the upper limit of the equivalent power supply radius of the substation under this confidence level can be obtained from the position of the α quantile and used in subsequent calculations. The specific expression is as follows:

[0151] Pr{R trans,i}≥α

[0152]

[0153] Among them, R trans,i Let be the reliable equivalent power supply radius of substation i, and take the lower limit value in actual calculation.

[0154] The substation boundary spacing update calculation method described in step (3) specifically includes:

[0155] 1) Calculation method for reference boundary spacing:

[0156] The reference boundary distance is physically represented as the shortest distance between the equivalent circular power supply areas of two substations. It is directly affected by two factors: the reliable equivalent power supply radius of the two substations and the center location of the substations. The specific calculation formula is as follows:

[0157]

[0158] Where D(i,j) is the reference boundary distance between substation i and substation j; (x i ,y i ) and (x j ,y j ) represent the coordinates of substation i and substation j, respectively.

[0159] 2) Forward conversion method of datum boundary spacing:

[0160] Since the above-mentioned reference boundary spacing has three possible values, it needs to be processed into a consistently positive value while ensuring that the order of the boundary spacings remains unchanged. The specific calculation formula is as follows:

[0161]

[0162] 3) Boundary spacing update method considering source load matching:

[0163] The constant boundary spacing is numerically updated based on the regional source-load matching degree that may be transferred between any adjacent substations, in order to shorten the boundary spacing of substations with better source-load matching, and ensure that substations assigned to the same energy grid take into account both interconnection characteristics and source-load matching characteristics.

[0164] IV. The boundary spacing update method considering source load matching described in step (3) specifically includes:

[0165] 1) Method for estimating the area that can be supplied between substation i and adjacent substations:

[0166] ① Connect the endpoints of the power supply range of substation i with the endpoints of the boundary lines of all surrounding substations' power supply ranges, and calculate the sum of all straight boundary distances l and l'. m ;

[0167] ② Calculate the angle of the area that can be transferred to the power grid based on the proportion of the distance to the adjacent straight boundary. The specific calculation formula is as follows:

[0168]

[0169] Where, β i,j For the angle of the area that can be supplied; l i,j The straight-line boundary distance between substation i and substation j;

[0170] ③ Using the connecting line between any two substations as the center line, divide the substation into several areas according to the range angle β. The area within substation i and between substation j that can be transferred to power supply is denoted as A. i,j ;

[0171] ④ Repeat the above process for all substations until all transferable areas within the region have been divided.

[0172] 2) Calculation method for curve matching degree of transferable supply area:

[0173] ① Put all A i,j With A j,i The system is combined to form an inter-station power transfer zone, and the net power probability curve for this zone is calculated. The specific calculation formula is shown below:

[0174] Pr{F P,t (W)}≥α

[0175]

[0176] Among them, PW,t For the time-series net power curve of this region, the lower limit value is taken during calculation; F P,t (W) is the probability distribution function of the sum of source load characteristics in this region.

[0177] ②Based on the net power curve obtained at confidence level α, calculate the source-load matching degree of this region. The specific calculation formula is as follows:

[0178]

[0179] Where, σ i,j The degree of source-load matching is expressed as standard deviation; T is the evaluation period; P t Let be the net power at time t; The average net power over the assessment period;

[0180] ③ Boundary spacing update method:

[0181]

[0182]

[0183] in, The boundary spacing update coefficient, with a value range of [0.5, 1]; D match (i,j) represents the update boundary spacing considering source load matching.

[0184] By taking into account the matching degree of source and load characteristics in the adjacent areas between substations, the tightness of inter-station connections is determined, and the optimal inter-station connections are achieved, thereby improving the utilization rate of inter-substation connection feeders.

[0185] Step (4) of obtaining the energy grid partitioning scheme set and scheme optimization based on the K-Means++ clustering algorithm specifically includes:

[0186] 1) Calculate the range of energy grid divisions. Where m is the number of substations.

[0187] 2) For each k, use the K-Means++ cluster center initialization method to generate a cluster center group;

[0188] 3) Grid division target number k and its corresponding cluster centers, and use K-Means++ clustering algorithm to cluster substations to form k energy grids;

[0189] 4) Optimize the set of alternative energy grid schemes:

[0190] ① Calculate the minimum average boundary distance b1(i,j) from the j-th substation in the i-th energy grid to the substations in each other energy unit, and the average boundary distance b2(i,j) within the energy grid. The calculation formula is as follows:

[0191]

[0192]

[0193] Where, n i Let be the number of substations in the i-th energy grid.

[0194] ② Calculate the structural indices of the clustering results

[0195] ③ Calculate the load balancing index between grids

[0196] ④ The optimal energy grid partitioning scheme is selected based on the objectives of a large BWP index in the clustering results and a small σ index in the inter-grid load balance.

[0197] Combination Figure 2 , Figure 3 With specific embodiments, the beneficial effects of the power grid optimization partitioning method considering source-load uncertainty and complementary characteristics established in this invention are described in detail below:

[0198] (1) Overview of the Implementation Examples:

[0199] Taking the load distribution of a certain area to be planned as an example, the total area of ​​this area is 68.25 km². 2 The area comprises 348 residential communities (including electrical and heating loads). The electrical load types include residential, industrial, commercial, administrative, and electric vehicle charging pile loads. The maximum peak power load is 744.5MW, with a power factor of 0.95. The distributed photovoltaic (PV) system connected to the grid in this area has a total installed capacity of 350.5MW. Seven substations are planned within the area, with capacities of six 3×50MW units and one 2×50MW unit. (See attached image) Figure 2 The figure in the middle shows the probability net load curve for a typical day in a certain community within the planning area.

[0200] (2) Overall solution process of the multi-energy system source-grid-load-storage interactive game operation model based on the embodiment:

[0201] Based on steps (1) to (2) of the multi-energy complementary power distribution system source-grid-load-storage interactive game operation strategy that considers the interests of different stakeholders, the probabilistic net load curve set for the area to be planned can be obtained, as shown in the attached figure. Figure 2 As shown, the specific steps are as follows:

[0202] ① Based on historical data, it can be seen that the output of distributed photovoltaic power follows a beta distribution at any given moment; the energy demand of heating and cooling loads, traditional multi-type power loads, and electric vehicle charging stations follows a normal distribution at any given moment; and the proportion of electricity purchased by the grid above the distributed energy station to the total energy purchase at any given moment follows a beta distribution. Heating and cooling loads are then converted to the power supply side through the distributed energy station. Based on historical data and the above parameters, output and energy consumption probability models for various types of equipment can be obtained.

[0203] ② Based on the aforementioned probability model, and using the mean of the probability model and historical data as boundary conditions, by setting the confidence level α to 0.95 and 0.9 using the relevant formulas in step (2), the output and energy consumption probability curves of various elements within the region, as well as the corresponding values ​​at each time point on the curves, can be obtained. The curve set is shown in the attached figure. Figure 2 As shown;

[0204] ③According to the appendix Figure 2 Based on the data distribution, 0.95 was selected as the confidence level for the example, and the reliable load density within the power supply range of each substation was calculated using the formula in step (2).

[0205] Based on the above reliable load density, the equivalent power supply radius of each substation and the updated boundary spacing considering source-load matching are calculated using the method in step (3) and related formulas. The specific details are as follows:

[0206] ①Based on the reliable load density and combined with the power supply range of the substation, the equivalent power supply radius of each substation is obtained by using the equivalent power supply radius calculation method and calculation formula in step (3);

[0207] ②Based on the equivalent power supply radius and the actual geographical distance between substations as the boundary condition, the basic boundary distance between each substation considering the electrical connection tightness is obtained by using the basic boundary distance calculation method and calculation formula in step (3), and the aforementioned benchmark boundary distance forward conversion method and processing all boundary distances into values ​​in the same direction are used.

[0208] ③ Based on the basic boundary spacing, the net load characteristic data variance of the transferable area between substations is obtained by using the estimation method of the transferable area range between substations, the calculation method of the curve matching degree of transferable area in step (3), and the range is compressed by the normalization method to obtain the updated boundary spacing considering source load matching.

[0209] Using the updated boundary spacing as the boundary condition, and employing the energy grid partitioning method and scheme optimization method based on the K-Means++ clustering algorithm in step (4), the final energy grid partitioning scheme can be obtained. The partitioning results of the embodiment are attached. Figure 3 As shown, the specific content is as follows:

[0210] ①Based on the energy grid division interval described in step (4), the grid quantity interval in this embodiment is determined to be [2,3];

[0211] ② For each partitioning scheme, the initial cluster centers of the corresponding scheme are obtained by using the initial cluster center generation method of the K-Means++ clustering algorithm;

[0212] ③ Using the updated boundary spacing as the sample moment and avoiding the constraint that there is only one substation in the grid, the above substations are divided based on the K-Means++ clustering algorithm;

[0213] ④ Calculate the corresponding structural index BWP and inter-grid load balancing index σ based on the clustering results, and optimize the partitioning scheme set to obtain the optimal partitioning scheme.

[0214] Based on this embodiment, and combined with the solution method and process described above, the partitioning results can be obtained as shown in the appendix. Figure 3 As shown.

[0215] According to the above figures, this method can fully consider the uncertainty of power output and energy consumption in the area to be planned, realize the probability conversion mapping between the heating and cooling network and the power grid, fully consider the power source and load matching characteristics between substations, improve the connection between substations, improve equipment utilization efficiency, and provide effective support for subsequent feeder optimization layout.

[0216] The above description is merely an 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 within the protection scope of the present invention.

Claims

1. A method for optimizing the partitioning of power distribution energy grids considering source-load uncertainty and complementary characteristics, characterized in that, The specific steps include the following: Step 1: Based on the historical output and energy consumption characteristics of the power load, the cold and heat loads are converted to the power side through the cold and heat load power supply model, and a source-load power side energy consumption probability model is established. Step 2: Based on the power consumption probability model of the source and load, obtain the net load curve and the power output / energy consumption curve of each source and load under the α confidence level within each power supply range, and calculate the regional credible load density. Step 3, based on the reliable load density, calculates the equivalent reliable power supply radius of each substation and the boundary distance of the benchmark substation, calculates the source-load matching degree between adjacent substations, and updates the substation boundary distance; including the following steps: (1) Calculate the reference boundary distance between each substation using the reference boundary distance calculation method; (2) Using the positive conversion method of the reference boundary spacing, the reference boundary spacing values ​​between each substation are converted into positive values; (3) Based on the positive reference boundary spacing, the updated boundary spacing of the substation is calculated using the boundary spacing update method that considers source-load matching; the boundary spacing update method that considers source-load matching includes the following steps: (3.1) Utilizing substations i A method for estimating the area that can be transferred between adjacent substations, and for estimating the area to be transferred: 1) Substation i Calculate the distances between the endpoints of the power supply range and the boundary lines of all surrounding substations, connecting them in a straight line. l Sum of distances from the straight boundary l m ; 2) Calculate the angle of the area that can be transferred to the power grid based on the proportion of the distance to the adjacent straight boundary. The specific calculation formula is as follows: ， in, β i,j The angle of the area that can be supplied; l i,j For substation i With substation j The straight-line boundary distance between them; 3) Using the connecting line between any two substations as the center line, according to the range angle β The substation is divided into several areas, and the area within substation i and between substation j that can be transferred to power is denoted as... A i,j ; 4) Perform the above process for all substations until all transferable areas within the region have been divided. (3.2) Calculate the curve matching degree of the transferable supply area using the curve matching degree calculation method: 1) All A i,j and A j,i The system is combined to form an inter-station power transfer zone, and the net power probability curve for this zone is calculated. The specific calculation formula is shown below: ， in, The lower limit value is taken for the time-series net power curve of this region during calculation; This is the probability distribution function of the sum of source load characteristics in this region; 2) Based on the obtained α The net power curve at the confidence level is used to calculate the source-load matching degree in this region. The specific calculation formula is shown below: ， in, The degree of source-load matching is expressed as standard deviation; T For the evaluation period; P t Let be the net power at time t; The average net power over the assessment period; (3.3) Using the boundary spacing update method, calculate the updated boundary spacing of the substations participating in the clustering: ， in, The boundary spacing update coefficient has a value range of [0.5, 1]. To update the boundary spacing considering source load matching; Step 4: Determine the range of grid numbers for the partitioning scheme based on the energy grid number partitioning method, obtain the results of different numbers of energy grid partitioning schemes, and optimize the scheme according to the evaluation index.

2. The method for optimizing the partitioning of power distribution energy grids considering source-load uncertainty and complementary characteristics according to claim 1, characterized in that, The method for calculating the reliable load density of the region includes: (1) A method for obtaining the net power probability curve within the power supply range based on the three-point estimation method and the Gram-Charlier series expansion method; (2) Calculation method of reliable load density based on net power curve of power supply range: ， in, and Substations i The load density within the power supply area follows the probability density function and probability distribution function; For substation i The actual area of ​​the power supply range.

3. The method for optimizing the partitioning of power distribution energy grids considering source-load uncertainty and complementary characteristics according to claim 1, characterized in that, The method for calculating the equivalent reliable power supply radius includes the following steps: (1) Calculate the probability density function and probability distribution function of the equivalent power supply radius of the substation using the calculation method of the probability density function and probability distribution function of the equivalent power supply radius; (2) Based on the set α confidence level, the uncertainty problem is treated as a deterministic problem.

4. The method for optimizing the partitioning of power distribution energy grids considering source-load uncertainty and complementary characteristics according to claim 1, characterized in that, The method for forward conversion of the reference boundary spacing includes the following steps: (1) The reference boundary distance is physically represented as the shortest distance between the equivalent circular power supply areas of the two substations. It is directly affected by two factors: the credible equivalent power supply radius of the two substations and the center location of the substations. The specific calculation formula is as follows: ， in, For substation i With substation j The reference boundary spacing between them; Substations i With substation j The coordinates of the location; (2) Since the above-mentioned reference boundary spacing has three possible values, it needs to be processed into a constant positive value, and the size arrangement of each boundary spacing should remain unchanged. The specific calculation formula is as follows: 。 5. The method for optimizing the partitioning of a power distribution energy grid considering source-load uncertainty and complementary characteristics according to claim 1, characterized in that, The energy grid partitioning method includes the following steps: (1) Calculate the range of energy grid divisions. ,in m The number of substations; (2) For each k Use the K-Means++ cluster center initialization method to generate cluster center groups; (3) Divide the target number k and its corresponding cluster centers into grids, and use the K-Means++ clustering algorithm to cluster the substations to form a grid. k One energy grid; (4) Optimize the indicators for the partitioning scheme: 1) Calculate the first i The first energy grid in the j The minimum average boundary distance between the substation and each other substation in each energy unit. and the average boundary spacing within the energy grid The calculation formula is: ， in, Let i be the number of substations within the i-th energy grid; 2) Calculate the structural indices of the clustering results ; 3) Calculate the load balancing index between grids. ; 4) Structural indicators based on clustering results BWP Larger metric and inter-grid load balancing metric The energy grid partitioning scheme is optimized with the goal of minimizing the target value.

6. A program system for an optimized partitioning method of power distribution energy grid considering source-load uncertainty and complementary characteristics, characterized in that, For performing the method according to any one of claims 1-5, comprising: Acquisition module: Based on the historical output and energy consumption characteristics of the power load, it converts the cold and hot load to the power side through the cold and hot load power supply model to obtain the source load power side energy consumption probability model; The processing module is used to obtain the optimized division results of the power distribution energy grid based on the source-load power side energy consumption probability model. Specifically, it includes: first, obtaining the net load curve and the power output / energy consumption curve of each source-load under the α confidence level within each power supply range based on the source-load power side energy consumption probability model, and calculating the regional credible load density; second, calculating the equivalent credible power supply radius of each substation and the boundary distance of the benchmark substation based on the credible load density, calculating the source-load matching degree between adjacent substations and updating the substation boundary distance; finally, determining the range of the number of grids in the division scheme based on the energy grid number division method, obtaining the results of energy grid division schemes with different numbers of grids, and optimizing the scheme according to the evaluation index. Sending module: Used to output the preferred power grid partitioning scheme that takes into account the uncertainty of source load and complementary characteristics.

7. An apparatus for optimizing the partitioning of a power distribution energy grid, considering source-load uncertainty and complementary characteristics, characterized in that, It includes a memory and a processor, the memory storing a program that runs on the processor, and the processor executing the steps of the power grid optimization partitioning method for considering source-load uncertainty and complementary characteristics as described in any one of claims 1-5 when running the program.

8. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed, they perform the steps of the power grid optimization partitioning method that considers source-load uncertainty and complementary characteristics as described in any one of claims 1-5.

Images