County power distribution network AC / DC hybrid optimal topology planning method, system and device, and storage medium

Abstract

The invention discloses an AC-DC hybrid optimal topology planning method, system and device for a county power distribution network, and a storage medium, and relates to the technical field of power distribution network planning, and the method comprises the steps: building an undirected graph model based on the node and line data of the power distribution network, calculating a complex network index, recognizing key nodes and lines, and obtaining a topology analysis result; calculating load density and the like by utilizing a topological structure and historical operation data, positioning high-loss and low-voltage regions, and obtaining an electrical characteristic analysis result; combining the load and the distributed energy data, calculating permeability and energy storage configuration, and analyzing influence to obtain a distributed energy index; dividing an alternating current region and a direct current region by using a spectral clustering algorithm according to the result; based on a division result, establishing a multi-objective planning model and performing optimization solution, generating an optimal solution set, and completing optimal topology planning; according to the method, key nodes and lines can be accurately identified, and a new thought and method are provided for fine management and optimization of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of power distribution network planning technology, and in particular to a method, system, equipment and storage medium for optimal AC / DC hybrid topology planning of a county-level power distribution network. Background Technology

[0002] With the widespread integration of distributed energy resources and power electronic equipment into county-level distribution networks, traditional AC distribution networks face significant challenges in terms of flexibility, economy, and reliability. DC distribution technology, with its advantages of high-efficiency transmission and ease of integration with distributed energy resources, is gradually gaining application in distribution networks. However, current research on hybrid AC / DC distribution networks mainly focuses on system control and stability analysis, with insufficient research on topology planning.

[0003] The existing AC / DC hybrid distribution network planning methods have the following problems: First, there is a lack of systematic methods for dividing AC and DC zones, and they rely heavily on empirical judgments; second, the impact of new energy penetration and energy storage facility access on network topology is not fully considered; third, the configuration optimization of tie switches and DC circuit breakers is insufficient, making it difficult to balance system reliability and economy; and fourth, there is a lack of long-term planning methods that take into account the development of time, making it difficult to adapt to the dynamic evolution needs of the distribution network. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by this invention are: first, how to scientifically analyze the topology of a county-level distribution network based on complex network theory to determine the optimal division between AC and DC zones; second, how to construct an AC / DC hybrid distribution network planning optimization model that considers multiple objectives such as investment cost, operating cost, reliability, and flexibility; third, how to design an optimization algorithm suitable for AC / DC hybrid distribution network planning to solve complex multi-objective optimization problems; and fourth, how to construct a distribution network evolution planning method oriented towards time-series development to achieve dynamic optimization of the distribution network.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for optimal AC / DC hybrid topology planning of a county-level power distribution network, comprising:

[0008] An undirected graph model is constructed based on distribution network node and line data. Complex network indicators are calculated, key nodes and lines are identified, and topology analysis results are obtained.

[0009] By using the distribution network topology and historical operating data, load density, voltage distribution and line loss are calculated, high-loss and low-voltage areas are located, and electrical characteristic analysis results are obtained.

[0010] By combining data on distribution network load and distributed energy resources, we calculate the penetration rate and energy storage configuration and analyze their impact on the power grid to obtain distributed energy resource indicators.

[0011] Based on the results of topology analysis, electrical characteristic analysis, and distributed energy indicators, a spectral clustering algorithm is used to divide the AC and DC regions.

[0012] Based on the AC / DC region division results, a multi-objective programming model is established and optimized to generate the optimal solution set, thus completing the optimal topology planning.

[0013] As a preferred scheme of the AC / DC hybrid optimal topology planning method for county-level power distribution networks, the following is provided:

[0014] The process of constructing an undirected graph model based on distribution network node and line data, calculating complex network indices, identifying key nodes and lines, and obtaining topology analysis results includes:

[0015] The distribution network topology is abstracted as an undirected graph. Based on the constructed undirected graph model, the topological characteristics of the distribution network are analyzed using complex network theory. Indicators that can reflect the topological characteristics of the distribution network are extracted and calculated. Based on the calculated indicators, key nodes and lines in the distribution network are identified, and the topological analysis results of the distribution network are output.

[0016] The beneficial effects of this preferred technical solution are as follows: by transforming the distribution network topology into an undirected graph model and using complex network theory for analysis, the topological characteristics of the distribution network can be comprehensively and quantitatively evaluated, and key nodes and lines can be accurately identified, providing new ideas and methods for the refined management and optimization of the distribution network.

[0017] As a preferred scheme of the AC / DC hybrid optimal topology planning method for county-level power distribution networks, the following is provided:

[0018] The method of calculating load density, voltage distribution, and line loss using the distribution network topology and historical operating data, locating high-loss and low-voltage areas, and obtaining electrical characteristic analysis results includes:

[0019] Based on the distribution network topology and historical operating data, a load density distribution map is generated to show the load density distribution at different locations in the distribution network; the voltage distribution characteristics of the distribution network are analyzed using power flow calculation to clarify the voltage status of each node; and the line loss distribution characteristics of the distribution network are calculated based on the distribution network topology and historical operating data to determine the loss status of each line.

[0020] Based on the analysis results of load density, voltage distribution, and line loss distribution, high-loss areas and low-voltage areas in the distribution network are located, and the electrical characteristic analysis results of the distribution network are output.

[0021] The beneficial effects of this preferred technical solution are as follows: by comprehensively utilizing the topology and historical operation data of the distribution network, the load density, voltage distribution and line loss distribution characteristics of the distribution network can be analyzed in a comprehensive and systematic manner, and high loss and low voltage areas can be accurately located. Compared with traditional methods, it can more effectively discover problems in the operation of the distribution network, and provide more targeted decision support for the refined management and optimization of the distribution network.

[0022] As a preferred scheme of the AC / DC hybrid optimal topology planning method for county-level power distribution networks, the following is provided:

[0023] The method of dividing the AC and DC regions using a spectral clustering algorithm, which combines topology analysis results, electrical characteristic analysis results, and distributed energy indicators, includes:

[0024] Combining the topology analysis results, electrical characteristic analysis results, and distributed energy indicators of the distribution network, and comprehensively considering topology characteristics, electrical characteristics, distributed energy, energy storage, and cost, a division evaluation index system is constructed. Based on the constructed evaluation index system, a decision model for AC / DC regional division is established to measure the degree to which each region is suitable for DC power distribution. By setting thresholds and based on the results of the decision model, the power distribution method of each region, i.e., AC power distribution or DC power distribution, is determined.

[0025] As a preferred scheme of the AC / DC hybrid optimal topology planning method for county-level power distribution networks, the following is provided:

[0026] The spectral clustering algorithm includes:

[0027] A similarity matrix is ​​constructed, considering the distance between nodes and a comprehensive evaluation function. Based on the similarity matrix, the Laplace matrix is ​​calculated, and the eigenvalues ​​and eigenvectors of the Laplace matrix are calculated. The eigenvectors corresponding to the smallest non-zero eigenvalues ​​are selected to form a matrix. The row vectors of the matrix are clustered to obtain the preliminary results of the region division. Based on the preliminary results of the division, the region boundary is optimized to complete the division of the AC and DC regions.

[0028] The beneficial effects of this preferred technical solution are as follows: It integrates multiple indicators to construct an evaluation system and decision-making model, and employs an improved spectral clustering algorithm for regional division, thereby enhancing the scientific rigor and rationality of AC / DC regional division. Simultaneously, it considers the dynamic evolution requirements of the distribution network to design a time-series development planning method, making the division scheme more forward-looking and adaptable, and better able to cope with future developments and changes in the distribution network.

[0029] As a preferred scheme of the AC / DC hybrid optimal topology planning method for county-level power distribution networks, the following is provided:

[0030] The process of establishing a multi-objective programming model based on the AC / DC region division results, optimizing and solving it to generate the optimal solution set, and completing the optimal topology planning includes:

[0031] Based on the AC / DC regional division results, a multi-objective programming model is constructed for the AC / DC hybrid mode of county-level distribution networks, taking into account objectives and corresponding constraints, including economy, reliability, flexibility and environmental protection.

[0032] As a preferred scheme of the AC / DC hybrid optimal topology planning method for county-level power distribution networks, the following is provided:

[0033] The process of establishing a multi-objective programming model based on the AC / DC region division results, optimizing and solving it to generate the optimal solution set, and completing the optimal topology planning also includes:

[0034] Decision variables are represented using a hybrid encoding method, including binary encoding, integer encoding, and real number encoding;

[0035] Heuristic methods are employed, combining complex network theory, load and energy distribution characteristics, and reliability analysis to determine the initial population. Individual fitness is calculated based on multi-objective function values, and non-dominated sorting is performed. Crowding distance is calculated to select individuals. A binary tournament selection strategy based on rank and crowding is adopted. Multi-point crossover operations are designed for mixed coding. An adaptive mutation strategy is designed, employing different mutation methods for different coding parts. A repair operator is designed to ensure offspring satisfy constraints. A local search mechanism is introduced, using a variable neighborhood search strategy to update the non-dominated solution set. An elite retention strategy is adopted to ensure that superior individuals are not lost. After multiple generations of evolution, a set of non-dominated solutions, i.e., the optimal solution set, is obtained.

[0036] Secondly, the present invention provides a hybrid AC / DC optimal topology planning system for a county-level power distribution network, comprising:

[0037] The distribution network topology feature mining module is used to construct an undirected graph model based on distribution network node and line data, calculate complex network indicators, identify key nodes and lines, and obtain topology analysis results.

[0038] The distribution network electrical performance diagnostic module is used to calculate load density, voltage distribution and line loss using the distribution network topology and historical operating data, locate high loss and low voltage areas, and obtain electrical characteristic analysis results.

[0039] The distributed energy impact assessment module is used to combine distribution network load and distributed energy data to calculate penetration rate and energy storage configuration and analyze the impact on the power grid, thereby obtaining distributed energy indicators.

[0040] The AC / DC zone intelligent partitioning module is used to divide the AC zone and DC zone by combining topology analysis results, electrical characteristic analysis results and distributed energy indicators, and using a spectral clustering algorithm.

[0041] The distribution network topology planning and optimization module is used to establish a multi-objective programming model based on the AC / DC region division results, perform optimization and solution, generate the optimal solution set, and complete the optimal topology planning.

[0042] Thirdly, the present invention provides an electronic device, comprising:

[0043] Memory and processor;

[0044] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the one or more programs are executed by the one or more processors, the one or more processors implement the optimal AC / DC hybrid topology planning method for county-level power distribution networks as described in this invention.

[0045] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the aforementioned method for optimal AC / DC hybrid topology planning of county-level power distribution networks.

[0046] The beneficial effects of this invention are as follows: Through complex network theory analysis, multi-objective optimization modeling, and an improved NSGA-II algorithm, this invention achieves optimal topology planning for county-level distribution networks under AC / DC hybrid mode, improving the economy of county-level distribution networks and reducing long-term operating costs; enhancing the reliability and flexibility of the distribution network, improving its ability to withstand faults and adapt to load changes; optimizing the division between AC and DC zones, achieving efficient resource utilization; and improving the environmental friendliness of the distribution network, promoting the consumption of new energy sources and reducing carbon emissions. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the 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.

[0048] Figure 1 This is an overall flowchart of the optimal AC / DC hybrid topology planning method for county-level power distribution networks provided by this invention; Detailed Implementation

[0049] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0050] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for optimal AC / DC hybrid topology planning of a county-level power distribution network, comprising:

[0051] S1: Construct an undirected graph model based on distribution network node and line data, calculate complex network indicators, identify key nodes and lines, and obtain topology analysis results;

[0052] S2: Calculate load density, voltage distribution and line loss using the distribution network topology and historical operating data, locate high-loss and low-voltage areas, and obtain electrical characteristic analysis results;

[0053] S3: Combining distribution network load and distributed energy data, calculate the penetration rate and energy storage configuration and analyze the impact on the power grid to obtain distributed energy indicators;

[0054] S4: Combining the topology analysis results, electrical characteristic analysis results, and distributed energy indicators, the spectral clustering algorithm is used to divide the AC region into the DC region;

[0055] S5: Based on the AC / DC region division results, establish a multi-objective programming model and optimize it to generate the optimal solution set and complete the optimal topology planning.

[0056] It should be noted that through steps S1-S5, using complex network theory analysis, multi-objective optimization modeling, and an improved NSGA-II algorithm, a reasonable division between the AC and DC zones, as well as the optimal configuration of tie switches and DC circuit breakers, were achieved. This fully considers the characteristics of county-level distribution networks and the needs of new energy access, demonstrating strong practicality and innovation, and is of great significance in promoting the development of distribution networks towards greater efficiency, reliability, and environmental friendliness.

[0057] Example 2, refer to Figure 1 As one embodiment of the present invention, based on the previous embodiment, a method for optimal AC / DC hybrid topology planning of a county-level power distribution network is provided, comprising:

[0058] In this embodiment, the steps in S1 above, including constructing an undirected graph model based on distribution network node and line data, calculating complex network indices, identifying key nodes and lines, and obtaining topology analysis results, include:

[0059] The distribution network topology is represented by an undirected graph G:

[0060] G = (V, E, W)

[0061] Where V is the set of nodes, representing substations, switching stations, load points, etc. in the distribution network; E is the set of edges, representing lines in the distribution network; and W is the weight matrix, representing the electrical parameters, length, capacity, and other attributes of the lines.

[0062] Based on the distribution network topology, the topological characteristics of the distribution network are analyzed using complex network theory, and key indicators are extracted and expressed as follows:

[0063]

[0064] Among them, C G N represents the clustering coefficient of the graph. Δ Let N be the number of triangles in the network. conn This represents the possible number of connections.

[0065] Average path length L G The calculation is as follows:

[0066]

[0067] Where n is the number of nodes, d ij Let be the shortest path length between nodes i and j.

[0068] Betweenness centrality of nodes B v Defined as:

[0069]

[0070] Where, σ st Let σ be the number of shortest paths between nodes s and t. st (v) represents the number of shortest paths passing through node v.

[0071] Betweenness centrality of edges B e Defined as:

[0072]

[0073] Where, σ st (e) represents the number of shortest paths containing edge e.

[0074] In this embodiment, step S2 above uses the distribution network topology and historical operating data to calculate load density, voltage distribution, and line loss, locate high-loss and low-voltage areas, and obtain electrical characteristic analysis results including:

[0075] The load density distribution map is represented as follows:

[0076]

[0077] Among them, D load (x,y) is a load density distribution map, describing the load density distribution at different locations (x,y) in the distribution network. P is the load power of node i, and K is the kernel function. i ,y i ) represents the coordinates of node i, and h represents the bandwidth parameter.

[0078] Voltage distribution characteristics are obtained through power flow calculations and are expressed as follows:

[0079] V i =f pf (P,Q,Z)

[0080] Among them, V i Let f be the voltage at node i. pf Let P be the power flow calculation function, Q be the active power vector, and Z be the reactive power vector.

[0081] The line loss distribution characteristics are calculated as follows:

[0082]

[0083] Among them, P loss,ij R represents the loss of line ij. ij P is the line resistance. ij Q represents the active power on the line. ij This refers to the reactive power on the line.

[0084] In this embodiment, step S3 above combines distribution network load and distributed energy data to calculate penetration rate and energy storage configuration, and analyzes the impact on the power grid, resulting in distributed energy indicators including:

[0085] Spatial distribution of distributed energy penetration (DG) ratio (x,y) can be represented as:

[0086]

[0087] in, It is a collection of distributed energy resources. For the load set, P DG,i For the capacity of distributed energy i, P load,j Let be the power of load j.

[0088] The distribution characteristics of energy storage facilities are represented as follows:

[0089]

[0090] in, For a collection of energy storage facilities, C ES,i Let i be the capacity of the energy storage facility.

[0091] In this embodiment, step S4 above, which combines topology analysis results, electrical characteristic analysis results, and distributed energy indicators, uses a spectral clustering algorithm to divide the AC and DC regions, including:

[0092] Construct a categorized evaluation index system, represented as follows:

[0093] S i ={S i,topo ,S i,elec ,S i,dg ,S i,es ,S i,cost}

[0094] Among them, S i,topo S is a topology characteristic index. i,elec S is an electrical characteristic indicator. i,dg S is an indicator for distributed energy resources. i,es S is an energy storage indicator. i,cost For cost indicators.

[0095] The formula for calculating topology characteristic indices is:

[0096] S i,topo =α1B i +α2C i +α3L i

[0097] Among them, B i For the average betweenness centrality of region i, C i L is the clustering coefficient. i α1, α2, and α3 are the average path lengths and the weighting coefficients.

[0098] The formula for calculating electrical characteristic parameters is:

[0099]

[0100] in, Let i be the average voltage quality index for region i. The average line loss rate, β1, β2, and β3 are the average current load factor and weighting coefficients, respectively.

[0101] The formula for calculating distributed energy indicators is:

[0102] S i,dg =γ1DG ratio,i +γ2DG var,i +γ3DG dist,i

[0103] Among them, DG ratio,i For distributed energy penetration rate in region i, DG var,i To address power output fluctuations, DG dist,i To ensure uniformity of distribution, γ1, γ2, and γ3 are weighting coefficients.

[0104] The formula for calculating energy storage indicators is:

[0105] S i,es=δ1ES ratio,i +δ2ES dist,i +δ3ES flex,i

[0106] Among them, ES ratio,i S represents the proportion of energy storage capacity in region i. dist,i For uniform distribution, ES flex,i To ensure flexible adjustment capabilities, δ1, δ2, and δ3 are weighting coefficients.

[0107] The formula for calculating cost indicators is:

[0108] S i,cost =η1C inv,i +η1C op,i +η3C maint,i

[0109] Among them, C inv,i C represents the investment cost index for region i. op,i As an operating cost indicator, C maint,i For the maintenance cost indicator, η1, η2, and η3 are weighting coefficients.

[0110] Based on the above indicators, a decision model for AC / DC zone division is constructed, expressed as:

[0111]

[0112] in, w represents the degree to which region i is suitable for DC power distribution. j The weights of each indicator.

[0113] By setting a threshold θ DC Determine the DC region:

[0114]

[0115] Among them, Region i This indicates the power distribution method selection result for region i, where DC represents direct current power distribution and AC represents alternating current power distribution.

[0116] To optimize region partitioning, an improved spectral clustering algorithm is adopted:

[0117] Construct a similarity matrix W, where w ij Indicate the similarity between nodes i and j:

[0118]

[0119] Where, d ij σ is the node distance, σ is the kernel parameter, and f comp(i,j) is a comprehensive evaluation function that considers factors such as electrical characteristics and distributed energy characteristics.

[0120] Calculate the Laplace matrix:

[0121] L=DW

[0122] Where D is the degree matrix, and the diagonal elements are...

[0123] Calculate the eigenvalues ​​and eigenvectors of L, and select the eigenvectors corresponding to the k smallest non-zero eigenvalues ​​to form a matrix.

[0124] K-means clustering is performed on the row vectors of V to obtain the region division results.

[0125] Optimize regional boundaries to ensure the rationality of AC / DC zone division.

[0126] In another possible implementation, considering the dynamic evolution requirements of the distribution network, a time-series development planning method is designed. The stage transition matrix is ​​defined as follows:

[0127] T t→t+1 =[p ij ] m×m

[0128] Where, p ij This represents the probability of a region transitioning from state i to state j, where m is the number of states. States include: all AC, partially DC, hybrid, all DC, etc.

[0129] The phase transition probability was determined through a multi-factor assessment:

[0130] p ij =f trans (E i E j C i→j B i→j )

[0131] Among them, E i and E j The evaluation metrics for states i and j are C and j, respectively. i→j For transfer costs, B i→j To transfer profits, f trans This is the transfer evaluation function.

[0132] In this embodiment, step S5 above, based on the AC / DC region division results, establishes a multi-objective programming model and performs optimization to generate the optimal solution set, thus completing the optimal topology planning, including:

[0133] Based on the AC / DC regional division results, a multi-objective programming model considering various objectives and constraints is constructed for the AC / DC hybrid mode of county-level power distribution networks.

[0134] The decision variables X in the planning model include:

[0135]

[0136] in, and These indicate whether an AC or DC line is constructed between nodes i and j, respectively. Indicates whether node i is equipped with an AC / DC conversion device; and These indicate whether an AC interconnection switch or a DC circuit breaker is installed between nodes i and j, respectively. and These represent the distributed energy and energy storage capacity of node i, respectively.

[0137] The objective function includes:

[0138] Economic objective: Minimize total cost, including investment cost, operating cost, and maintenance cost.

[0139]

[0140] Among them, C inv (X) represents the investment cost, and C... op,t (X) represents the operating cost in year t, and C represents the operating cost in year t. maint,t (X) represents the maintenance cost in year t, r represents the discount rate, and T represents the planning period.

[0141] The investment cost is calculated as follows:

[0142]

[0143] in, and These are the unit investment costs for AC lines and DC lines, respectively. The investment cost of AC / DC conversion equipment, and The investment costs are for AC tie switches and DC circuit breakers, respectively. and These are the unit investment costs for distributed energy and energy storage, respectively.

[0144] Operating costs are calculated as follows:

[0145] C op,t (X)=C loss,t (X)+C energy,t (X)+C penalty,t (X)

[0146] Among them, C loss,t (X) represents network loss cost, and C energy,t (X) represents the energy purchase cost, and C represents the energy purchase cost. penalty,t (X) represents the reliability penalty cost.

[0147] Maintenance costs are calculated as follows:

[0148]

[0149] in, These are the unit maintenance costs for each piece of equipment.

[0150] Reliability objective: Maximize system reliability, using a weighted combination of the System Average Outage Frequency Index (SAIFI) and the System Average Outage Duration Index (SAIDI):

[0151] f2(X)=w SAIFI ·SAIFI(X)+w SAIDI ·SAIDI(X)

[0152] Among them, w SAIFI and w SAIDI These are the weighting coefficients.

[0153] The formulas for calculating SAIFI and SAIDI are as follows:

[0154]

[0155] Where, N i Let λ be the number of users at node i. i (X) and U i (X) represents the average number of power outages and the duration of power outages per year for node i.

[0156] Flexibility objective: Maximize system flexibility, measuring the distribution network's ability to adapt to load changes and distributed energy fluctuations.

[0157] f3(X) = w1·F topo (X)+w2·F ctrl (X)+w3·F res (X)

[0158] Among them, F topo (X) is the topology flexibility index, F ctrl (X) is the control flexibility index, F res (X) is the resource flexibility index, and w1, w2, and w3 are weighting coefficients.

[0159] The topology flexibility index is calculated as follows:

[0160]

[0161] Where, N conn,i (X) represents the number of connectable paths for node i under configuration scheme X, N max,i This represents the theoretical maximum number of connections.

[0162] The control flexibility index is calculated as follows:

[0163]

[0164] in, These are the accessibility indices for the AC tie switch, DC circuit breaker, and AC / DC converter at node i, respectively. a w d w c These are the weighting coefficients.

[0165] The resource flexibility index is calculated as follows:

[0166]

[0167] Among them, F i dg and F i es P represents the flexibility coefficient for distributed energy resources and energy storage, respectively. load,i Let be the load power of node i.

[0168] Environmental goal: Minimize carbon emissions.

[0169]

[0170] Among them, E carbon,t (X) represents the carbon emissions in year t.

[0171] Carbon emissions are calculated as follows:

[0172]

[0173] Among them, P grid,i,t For the electricity purchased by node i from the grid in year t, e grid P is the carbon emission factor of the power grid. dg,j,t Let e ​​be the amount of electricity generated by distributed energy source j in year t. dg,j This refers to the corresponding carbon emission factor.

[0174] Set constraints, including:

[0175] Power balance constraints:

[0176]

[0177] Among them, P ij and Q ijP represents the active and reactive power on line ij, respectively. dg,i and Q dg,i P represents the active and reactive power outputs of the distributed energy source at node i, respectively. dg,i and Q es,i These represent the active and reactive power of the energy storage device, P. load,i and Q load,i These represent the active and reactive loads of node i, respectively.

[0178] Voltage constraint:

[0179] V min ≤V i ≤V max

[0180] Among them, V i Let V be the voltage at node i. min and V max These are the minimum and maximum allowable voltages, respectively.

[0181] Line capacity constraints:

[0182]

[0183] Among them, S ij S represents the apparent power of line ij. ij,max This represents the maximum capacity of the line.

[0184] AC / DC converter constraints:

[0185]

[0186] Among them, P cv,i and Q cv,i These represent the active and reactive power of the AC / DC converter, P. cv,i,max and Q cv,i,max These are the maximum active and reactive capacities, respectively.

[0187] Constraints of Distributed Energy Resources and Energy Storage:

[0188]

[0189] SOC min ·E es,i ≤E es,i,t ≤SOC max ·E es,i

[0190] Among them, P dg,i,max and Q dg,i,max These represent the maximum active and reactive power capacities of distributed energy resources, P. es,i,max and Q es,i,max These represent the maximum active and reactive power of energy storage, E.es,i For energy storage capacity, E es,i,t Let SOC be the energy state of the stored energy at time t. min and SOC max These represent the minimum and maximum permissible states of charge, respectively.

[0191] Connection switch and circuit breaker constraints:

[0192]

[0193] The constraint on tie switches and circuit breakers indicates that the corresponding tie switches or circuit breakers can only be installed if the corresponding type of line is constructed.

[0194] Network radiation state constraints:

[0195]

[0196] Where n is the number of nodes.

[0197] Network radiation state constraints ensure that the network maintains its radiation state during normal operation, while taking into account the configuration of tie switches and circuit breakers.

[0198] AC / DC region connection constraints:

[0199]

[0200] Among them, V AC and V DC These are the node sets for the AC and DC regions, respectively. This refers to the minimum number of connections required between AC and DC regions.

[0201] Combining the above objective function and constraints, a multi-objective optimization model is constructed, expressed as:

[0202] minF(X)=[f1(X),f2(X),-f3(X),f4(X)]

[0203] To address the complexity of AC / DC hybrid distribution network planning, an improved non-dominated sorting genetic algorithm (NSGA-II) is designed to solve the multi-objective optimization problem.

[0204] Specifically, a hybrid encoding method is used to represent the decision variable Chrom:

[0205] Chrom = [Chrom] ac Chrom dc Chrom cv Chrom as Chrom ds Chrom dg Chrom es ]

[0206] Each part corresponds to the configuration of AC lines, DC lines, AC / DC conversion equipment, AC tie switches, DC circuit breakers, distributed energy resources, and energy storage.

[0207] For the coding of lines and switches, binary coding is used:

[0208]

[0209] For AC / DC conversion equipment, integer encoding is used:

[0210]

[0211] in, is an integer representing the type and capacity level of the AC / DC conversion equipment installed at node i.

[0212] For distributed energy resources and energy storage, real number encoding is used:

[0213]

[0214] in, and Let be a real number, representing the distributed energy and energy storage capacity of node i.

[0215] The initial population is generated using a heuristic method:

[0216] Based on the analysis results of complex network theory, AC / DC conversion equipment should be prioritized at key nodes;

[0217] The line type and capacity are determined based on load density and the distribution characteristics of distributed energy resources;

[0218] The configuration of tie switches and circuit breakers was initially determined using reliability analysis methods.

[0219] Based on load characteristics and distributed energy output characteristics, the energy storage capacity is initially configured.

[0220] The pseudocode for the initial population generation algorithm is as follows:

[0221] Input: Network topology G, load data, distributed energy data, constraints;

[0222] Output: Initial population Pop;

[0223] Based on the network analysis results, the division between the AC and DC zones was determined.

[0224] Calculate the importance index of each node;

[0225] For each individual i=1 to PopSize, based on node importance, configure AC / DC conversion equipment, configure AC and DC lines to ensure network connectivity, configure tie switches and circuit breakers to improve reliability, configure distributed energy and energy storage to optimize economy and environmental protection, check constraints, make necessary corrections, and add the individual to the population Pop;

[0226] Return to the initial population Pop;

[0227] Based on the multi-objective function value, the fitness of each individual is calculated and a non-dominated ranking is performed:

[0228] Calculate the objective function values ​​[f1(X), f2(X), -f3(X), f4(X)] for each individual;

[0229] Based on Pareto dominance, the population is divided into different non-dominant levels;

[0230] Within each non-dominated level, the crowding distance is calculated, and individuals with higher crowding are prioritized.

[0231] The formula for calculating congestion distance is:

[0232]

[0233] Where m is the number of objective functions, X i+1 and X i-1 These are the adjacent individuals of individual i after sorting on target j. and These are the maximum and minimum values ​​of target j, respectively.

[0234] A binary tournament selection strategy based on ranking and congestion is adopted:

[0235] Two individuals are randomly selected;

[0236] Compare their non-dominant levels and select the individuals with the lower levels;

[0237] If the levels are the same, choose the individual with the higher crowding level.

[0238] Design a multi-point crossover operation for hybrid encoding:

[0239] For the binary coded portion (circuit and switch), uniform crossover is used:

[0240]

[0241] Where r is a random number between 0 and 1, p c This represents the crossover probability.

[0242] For the integer encoding section (AC / DC conversion equipment), arithmetic crossover is used:

[0243] Child1 i =α·Parent1 i +(1-α)·Parent2 i

[0244] Child2 i = (1-α)·Parent1 i +α·Parent2 i

[0245] Where α is a random number between 0 and 1.

[0246] For the real-number encoded portion (distributed energy resources and energy storage), SBX cross-linking is used:

[0247] Child1 i =0.5·[(1+β)·Parent1 i +(1-β)·Parent2 i ]

[0248] Child2 i =0.5·[(1-β)·Parent1 i +(1+β)·Parent2 i ]

[0249] Where β is the distribution index η c Generated random numbers.

[0250] Design an adaptive mutation strategy:

[0251] For the binary encoded portion, a bit-flipping mutation is used:

[0252]

[0253] Where, p m (i) represents the adaptive mutation probability, which is related to the individual's fitness.

[0254] For the integer encoding part, neighborhood mutation is used:

[0255] Child i =Parent i +rand(-δ,δ)

[0256] Where δ is the variable length, and rand(-δ,δ) represents a random integer between -δ and δ.

[0257] For the real number encoding part, polynomial mutation is used:

[0258] Child i =Parent i +σ·(ub i -lb i )

[0259] Where σ is the distribution index η m Generated random numbers, ub i and lb i These are the upper and lower bounds of the variable, respectively.

[0260] Design a repair operator to ensure that offspring satisfy the constraints:

[0261] Repair network connectivity constraints;

[0262] Repairing radiation state constraints;

[0263] Repair the compatibility between AC / DC conversion equipment and line type;

[0264] Repair the matching relationship between the interconnecting switches and circuit breakers and the lines.

[0265] In another possible implementation, a local search mechanism can be introduced to improve the search efficiency of the algorithm:

[0266] Select the non-dominated solutions in the current population for local search;

[0267] A variable neighborhood search strategy is adopted to define multiple neighborhood structures: adjustment of tie switches and circuit breaker configurations; adjustment of the location and capacity of AC / DC conversion equipment; and adjustment of distributed energy and energy storage capacity.

[0268] The non-dominated solution set is updated by searching each neighborhood with a certain probability.

[0269] An elite preservation strategy is employed to ensure that superior individuals are not lost during the evolutionary process:

[0270] Merge the parent and offspring populations;

[0271] Perform non-dominated sorting and crowding calculation;

[0272] Select individuals of lower rank to form a new generation of the population;

[0273] For individuals of the same rank, those with higher crowding levels are given priority.

[0274] After multiple generations of evolution, a set of non-dominated solutions, i.e., the optimal solution set, is obtained.

[0275] In another possible implementation, after generating the optimal solution set, various visualization tools are designed to support decision-makers in understanding and analyzing the characteristics of each option:

[0276] Radar chart: Visually displays the performance of the solution across various indicators;

[0277] Parallel coordinate graphs: Compare the differences between different schemes on various indicators;

[0278] Scatter matrix: used to analyze the correlation between indicators;

[0279] Sensitivity analysis: Evaluate the impact of weight changes on the ranking of alternatives.

[0280] Recommendation and decision-making

[0281] Based on the comprehensive evaluation results and sensitivity analysis, the optimal solution is recommended to the decision-maker:

[0282] The solution with the highest overall score will be the recommended solution.

[0283] Several distinctive alternatives are provided, such as the most economical option and the most reliable option.

[0284] Analyze the advantages, disadvantages, and applicable conditions of each scheme;

[0285] It supports decision-makers in flexibly adjusting indicator weights according to actual needs to obtain the optimal solution that meets specific requirements.

[0286] In another possible implementation, the evaluation results and sensitivity analysis specifically include:

[0287] Design a decision support system based on the Analytic Hierarchy Process (AHP).

[0288] Construct an evaluation index system that includes four dimensions: economy, reliability, flexibility, and environmental protection.

[0289] I = I econ ,I reli ,I flex ,I env

[0290] Among them, I econ As an economic indicator, I reli As a reliability indicator, I flex As a flexibility indicator, I env These are environmental indicators.

[0291] Multiple secondary indicators are set under each dimension:

[0292] I econ =I econ,1 ,I econ,2 ,...,I econ,m1

[0293] I reli =I reli,1 ,Ireli,2 ,...,I reli,m2

[0294] I flex =I flex,1 ,I flex,2 ,...,I flex,m3

[0295] I env =I env,1 ,I env,2 ,...,I env,m4

[0296] Where m1, m2, m3, and m4 represent the number of secondary indicators under each dimension.

[0297] The weights of the indicators are determined using the analytic hierarchy process (AHP).

[0298] Construct the judgment matrix:

[0299]

[0300] Among them, a ij This indicates the degree of importance of indicator i relative to indicator j.

[0301] Calculate the weight vector:

[0302]

[0303] Among them, w i Let i be the weight of indicator i.

[0304] Consistency check:

[0305]

[0306] Where, λ max To determine the largest eigenvalue of a matrix, CI is the consistency index, RI is the random consistency index, and CR is the consistency ratio.

[0307] Based on the weights obtained from the analytic hierarchy process (AHP), a comprehensive evaluation is performed on the various solutions in the non-dominated solution set:

[0308]

[0309] Among them, S j For the overall score of scheme j, w i f is the weight of indicator i. i (X j ) represents the normalized value of scheme j on index i.

[0310] Example 3: The above is an illustrative scheme of the AC / DC hybrid optimal topology planning method for a county-level distribution network according to this embodiment. It should be noted that the technical solution of the AC / DC hybrid optimal topology planning system for a county-level distribution network and the technical solution of the AC / DC hybrid optimal topology planning method for a county-level distribution network belong to the same concept. Details not described in detail in the technical solution of the AC / DC hybrid optimal topology planning system for a county-level distribution network in this embodiment can be found in the description of the technical solution of the AC / DC hybrid optimal topology planning method for a county-level distribution network described above.

[0311] This embodiment also provides a hybrid AC / DC optimal topology planning system for a county-level power distribution network, including:

[0312] The distribution network topology feature mining module is used to construct an undirected graph model based on distribution network node and line data, calculate complex network indicators, identify key nodes and lines, and obtain topology analysis results.

[0313] The distribution network electrical performance diagnostic module is used to calculate load density, voltage distribution and line loss using the distribution network topology and historical operating data, locate high loss and low voltage areas, and obtain electrical characteristic analysis results.

[0314] The distributed energy impact assessment module is used to combine distribution network load and distributed energy data to calculate penetration rate and energy storage configuration and analyze the impact on the power grid, thereby obtaining distributed energy indicators.

[0315] The AC / DC zone intelligent partitioning module is used to divide the AC zone and DC zone by combining topology analysis results, electrical characteristic analysis results and distributed energy indicators, and using a spectral clustering algorithm.

[0316] The distribution network topology planning and optimization module is used to establish a multi-objective programming model based on the AC / DC region division results, perform optimization and solution, generate the optimal solution set, and complete the optimal topology planning.

[0317] This embodiment also provides an electronic device applicable to the AC / DC hybrid optimal topology planning method for county-level power distribution networks, including:

[0318] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the optimal AC / DC hybrid topology planning method for county-level power distribution networks as proposed in the above embodiments.

[0319] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the optimal AC / DC hybrid topology planning method for county-level power distribution networks as proposed in the above embodiments.

[0320] The storage medium proposed in this embodiment and the optimal AC / DC hybrid topology planning method for county power distribution networks proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0321] Example 4, referring to Tables 1-3, is an embodiment of the present invention, which provides a hybrid AC / DC optimal topology planning method for a county-level power distribution network. To verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0322] Experiments were conducted in MATLAB using actual data from a county-level power distribution network. The experimental data included a 10kV power distribution network in the area supplied by a 110kV substation, comprising 86 nodes and 92 existing lines, with a distributed energy penetration rate of 28%. The experimental results are shown in Tables 1-3.

[0323] Table 1 Comparative Analysis of Key Indicators for Different Planning Schemes

[0324]

[0325]

[0326] Table 2: Economic Analysis of Different Planning Schemes

[0327]

[0328]

[0329] Table 3: Reliability and Flexibility Analysis of Different Planning Schemes

[0330]

[0331] Experimental results show that this method effectively solves the topology planning problem in the AC / DC hybrid mode of county-level power distribution networks: compared with the all-AC scheme, the optimized hybrid scheme increases the total cost by 11.6%, but improves power supply reliability by 14.3% and flexibility by 15.4%, while increasing carbon emission reduction by 21.9%; compared with the all-DC scheme, the optimized hybrid scheme reduces the total cost by 12.2%, while maintaining the advantages of reliability and flexibility, with only a small loss of carbon emission reduction benefits; the optimized hybrid scheme achieves a reasonable division between AC and DC zones, with the AC zone mainly distributed in areas with low load density, and the DC zone mainly distributed in areas with concentrated distributed energy and high load density; the optimized configuration of tie switches and DC circuit breakers improves system reliability and flexibility, while controlling investment costs.

[0332] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A hybrid AC / DC optimal topology planning method for a county-level power distribution network, characterized in that, include: An undirected graph model is constructed based on distribution network node and line data. Complex network indicators are calculated, key nodes and lines are identified, and topology analysis results are obtained. By using the distribution network topology and historical operating data, load density, voltage distribution and line loss are calculated, high-loss and low-voltage areas are located, and electrical characteristic analysis results are obtained. By combining data on distribution network load and distributed energy resources, we calculate the penetration rate and energy storage configuration and analyze their impact on the power grid to obtain distributed energy resource indicators. Based on the results of topology analysis, electrical characteristic analysis, and distributed energy indicators, a spectral clustering algorithm is used to divide the AC and DC regions. Based on the AC / DC region division results, a multi-objective programming model is established and optimized to generate the optimal solution set, thus completing the optimal topology planning.

2. The optimal AC / DC hybrid topology planning method for a county-level power distribution network as described in claim 1, characterized in that, The process of constructing an undirected graph model based on distribution network node and line data, calculating complex network indices, identifying key nodes and lines, and obtaining topology analysis results includes: The distribution network topology is abstracted as an undirected graph. Based on the constructed undirected graph model, the topological characteristics of the distribution network are analyzed using complex network theory. Indicators that can reflect the topological characteristics of the distribution network are extracted and calculated. Based on the calculated indicators, key nodes and lines in the distribution network are identified, and the topological analysis results of the distribution network are output.

3. The optimal AC / DC hybrid topology planning method for a county-level power distribution network as described in claim 2, characterized in that, The method of calculating load density, voltage distribution, and line loss using the distribution network topology and historical operating data, locating high-loss and low-voltage areas, and obtaining electrical characteristic analysis results includes: Based on the distribution network topology and historical operating data, a load density distribution map is generated to show the load density distribution at different locations in the distribution network; the voltage distribution characteristics of the distribution network are analyzed using power flow calculation to clarify the voltage status of each node; and the line loss distribution characteristics of the distribution network are calculated based on the distribution network topology and historical operating data to determine the loss status of each line. Based on the analysis results of load density, voltage distribution, and line loss distribution, high-loss areas and low-voltage areas in the distribution network are located, and the electrical characteristic analysis results of the distribution network are output.

4. The optimal AC / DC hybrid topology planning method for a county-level power distribution network as described in claim 3, characterized in that, The method of dividing the AC and DC regions using a spectral clustering algorithm, which combines topology analysis results, electrical characteristic analysis results, and distributed energy indicators, includes: Combining the topology analysis results, electrical characteristic analysis results, and distributed energy indicators of the distribution network, and comprehensively considering topology characteristics, electrical characteristics, distributed energy, energy storage, and cost, a division evaluation index system is constructed. Based on the constructed evaluation index system, a decision model for AC / DC regional division is established to measure the degree to which each region is suitable for DC power distribution. By setting thresholds and based on the results of the decision model, the power distribution method of each region, i.e., AC power distribution or DC power distribution, is determined.

5. The optimal topology planning method for a county-level power distribution network with AC / DC hybrid configuration as described in claim 4, characterized in that, The spectral clustering algorithm includes: A similarity matrix is ​​constructed, considering the distance between nodes and a comprehensive evaluation function. Based on the similarity matrix, the Laplace matrix is ​​calculated, and the eigenvalues ​​and eigenvectors of the Laplace matrix are calculated. The eigenvectors corresponding to the smallest non-zero eigenvalues ​​are selected to form a matrix. The row vectors of the matrix are clustered to obtain the preliminary results of the region division. Based on the preliminary results of the division, the region boundary is optimized to complete the division of the AC and DC regions.

6. The optimal AC / DC hybrid topology planning method for a county-level power distribution network as described in claim 5, characterized in that, The process of establishing a multi-objective programming model based on the AC / DC region division results, optimizing and solving it to generate the optimal solution set, and completing the optimal topology planning includes: Based on the AC / DC regional division results, a multi-objective programming model is constructed for the AC / DC hybrid mode of county-level distribution networks, taking into account objectives and corresponding constraints, including economy, reliability, flexibility and environmental protection.

7. The optimal AC / DC hybrid topology planning method for a county-level power distribution network as described in claim 6, characterized in that, The process of establishing a multi-objective programming model based on the AC / DC region division results, optimizing and solving it to generate the optimal solution set, and completing the optimal topology planning also includes: Decision variables are represented using a hybrid encoding method, including binary encoding, integer encoding, and real number encoding; Heuristic methods are employed, combining complex network theory, load and energy distribution characteristics, and reliability analysis to determine the initial population. Individual fitness is calculated based on multi-objective function values, and non-dominated sorting is performed. Crowding distance is calculated to select individuals. A binary tournament selection strategy based on rank and crowding is adopted. Multi-point crossover operations are designed for mixed coding. An adaptive mutation strategy is designed, employing different mutation methods for different coding parts. A repair operator is designed to ensure offspring satisfy constraints. A local search mechanism is introduced, using a variable neighborhood search strategy to update the non-dominated solution set. An elite retention strategy is adopted to ensure that superior individuals are not lost. After multiple generations of evolution, a set of non-dominated solutions, i.e., the optimal solution set, is obtained.

8. A hybrid AC / DC optimal topology planning system for a county-level power distribution network, using the method described in any one of claims 1 to 7, characterized in that, include: The distribution network topology feature mining module is used to construct an undirected graph model based on distribution network node and line data, calculate complex network indicators, identify key nodes and lines, and obtain topology analysis results. The distribution network electrical performance diagnostic module is used to calculate load density, voltage distribution and line loss using the distribution network topology and historical operating data, locate high loss and low voltage areas, and obtain electrical characteristic analysis results. The distributed energy impact assessment module is used to combine distribution network load and distributed energy data to calculate penetration rate and energy storage configuration and analyze the impact on the power grid, thereby obtaining distributed energy indicators. The AC / DC zone intelligent partitioning module is used to divide the AC zone and DC zone by combining topology analysis results, electrical characteristic analysis results and distributed energy indicators, and using a spectral clustering algorithm. The distribution network topology planning and optimization module is used to establish a multi-objective programming model based on the AC / DC region division results, perform optimization and solution, generate the optimal solution set, and complete the optimal topology planning.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images