A method for selecting and optimizing a honeycomb active power distribution network topology based on graph theory

Abstract

A kind of topology selection and optimization method of honeycomb active distribution network based on graph theory, with line impedance and rated apparent power as constraints, constructs multi-objective weight factor, adopts improved spectral clustering algorithm cluster partition distribution network node, transition microgrid group after cluster partition into honeycomb active distribution network;Each microgrid is abstracted as a single node of honeycomb distribution network, the connection relationship between HSPH and microgrid is constructed Mathematical representation, realize the sparse matrix method modeling of honeycomb active distribution network;With the maximum HSPH investment return rate index and system stability index as objective function, the site selection strategy of HSPH is optimized;Based on the time series data of microgrid group net load, the energy storage capacity of HSPH is optimized, and the optimal topology structure of honeycomb active distribution network is finally generated. Through honeycomb topology and HSPH site selection and capacity optimization, the accommodation capacity of distribution network to new energy is improved, and the economy, reliability and flexibility of distribution network are increased.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of intelligent power distribution network planning and operation control, and particularly relates to a method for selecting and optimizing the topology structure of a honeycomb active power distribution network based on graph theory. BACKGROUND

[0002] The power distribution network is a key component of the power system, mainly responsible for transmitting power from the substation to the end user. Traditional power distribution networks mostly adopt a one-way power supply structure. With the growth of power demand and the access of renewable energy, traditional power distribution networks face problems such as large load fluctuations, poor reliability, and unstable power supply. To solve these challenges, the honeycomb active power distribution network has emerged. The honeycomb power distribution network uses distributed power generation, energy storage devices, and intelligent control technology to achieve bidirectional power flow and flexibly respond to load fluctuations and power demand changes. Unlike traditional power distribution networks, the honeycomb power distribution network has high structural flexibility and reliability, and its unique topology structure enables it to effectively improve the stability and robustness of the power grid, especially when dealing with high proportions of distributed power access. The initial research direction of the honeycomb power distribution network focuses on its topology structure, gradually developing into multi-objective optimization, robustness analysis, and other directions. With the advancement of smart grid technology, the honeycomb power distribution network begins to integrate Internet of Things, big data, and artificial intelligence, providing new solutions for more efficient and intelligent power distribution network management.

[0003] Although existing honeycomb power distribution networks have achieved certain application results in multiple fields, existing topology optimization methods still have problems such as high computational complexity, poor optimization effect, and difficulty in adapting to dynamic changes. Therefore, how to efficiently select and optimize the topology structure of the honeycomb active power distribution network has become one of the current research hotspots. The present application is proposed to address the shortcomings of existing technologies and further improve the reliability, economy, and flexibility of power distribution networks in practical applications. SUMMARY

[0004] To address the shortcomings of existing technologies, the present application provides a method for selecting and optimizing the topology structure of a honeycomb active power distribution network based on graph theory. The method uses an improved spectral clustering algorithm to divide the microgrid (MG) cluster into a honeycomb power distribution network and establishes a honeycomb active power distribution network model based on a sparse matrix. The method maximizes the return on investment and system stability as the objective function for the Hive Smart Power Hub (HSPH) site selection and capacity determination, and uses the Tabu Search Algorithm (TSA) for optimization and solution. Finally, the method achieves stability and economic improvement of the honeycomb power distribution network and improves the accommodation capacity of new energy.

[0005] To achieve the above object, the application provides a kind of based on graph theory's honeycomb active power distribution network topology selection and optimization method, comprising the following steps:

[0006] Step S1, with line impedance and rated apparent power as constraints, construct multi-objective spectrum clustering weight factor;

[0007] Step S2, using improved spectrum clustering algorithm to divide distribution network, maximize the edge weight inside microgrid, minimize the weight between microgrids, and each microgrid cluster constitutes a honeycomb active power distribution network unit;

[0008] Step S3, propose a sparse matrix modeling method for honeycomb active power distribution network, abstract each microgrid as a single node, construct the mathematical representation of the connection relationship between HSPH and microgrid, and express the connection relationship between microgrid and HSPH by adjacency matrix;

[0009] Step S4, construct an optimization site selection scheme with maximizing HSPH investment return rate and system stability as objective functions, optimize the site selection strategy of HSPH, wherein the life cycle cost is divided into initial investment cost and later maintenance cost, and the total income mainly includes peak-valley electricity price arbitrage income and auxiliary service income;And construct a network connectivity index to evaluate the influence of site selection scheme on the stability of distribution network structure;

[0010] Step S5, based on microgrid net load time series data, combined with energy transmission efficiency constraint and deep discharge correction, optimize the energy storage capacity of honeycomb intelligent control power hub.

[0011] Step S6, optimize the site selection scheme and calculate the responding energy storage capacity by neighborhood operation and updating tabu list, and generate the optimal topology structure of honeycomb active power distribution network.

[0012] In the application, multi-objective weight factor is constructed with line impedance and rated apparent power as constraints, improved spectrum clustering algorithm is used to cluster and divide distribution network nodes, and the microgrid clusters after clustering are transitioned into honeycomb active power distribution network;Each microgrid is abstracted as a single node of honeycomb distribution network, the mathematical representation of the connection relationship between HSPH and microgrid is constructed, the connection relationship between microgrid and HSPH is represented by adjacency matrix, the sparse matrix modeling of honeycomb active power distribution network is realized;The site selection strategy of HSPH is optimized with maximizing HSPH investment return rate index and system stability index as objective functions, neighborhood operation and updating tabu list are used to avoid falling into local optimum during optimization search, and based on microgrid net load time series data, combined with energy transmission efficiency constraint and deep discharge correction, the energy storage capacity of honeycomb intelligent control power hub is optimized, and finally the optimal topology structure of honeycomb active power distribution network is generated. Through honeycomb topology and HSPH site selection and capacity optimization, the accommodation capacity of distribution network for new energy is improved, and the economy, reliability and flexibility of distribution network are increased. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 A flowchart of the method for selecting the honeycomb active power distribution network topology of the application;

[0014] Figure 2 A honeycomb active power distribution network model based on a sparse matrix of the application;

[0015] Figure 3 A schematic diagram of active power distribution network division and HSPH site selection and access of the application;

[0016] Figure 4 A schematic diagram of the honeycomb active power distribution network topology of the application;

[0017] Figure 5 An HSPH site selection and capacity optimization strategy based on tabu search of the application. DETAILED DESCRIPTION

[0018] The embodiments of the application will be further explained in conjunction with the accompanying drawings:

[0019] In one embodiment of the application, as shown in Figure 1 A graph theory-based honeycomb active power distribution network topology selection and optimization method, characterized in that the method comprises the following steps:

[0020] Step S1, constructing a multi-objective spectral clustering weight factor with line impedance and rated apparent power as constraints;

[0021] Step S2, dividing the power distribution network using an improved spectral clustering algorithm, maximizing the edge weight within the microgrid and minimizing the weight between microgrids, and each microgrid cluster constituting a honeycomb active power distribution network unit;

[0022] Step S3, abstracting each microgrid as a single node, constructing a mathematical representation of the connection relationship between HSPH and microgrids, representing the connection relationship between microgrids and HSPH through an adjacency matrix, and completing the sparse matrix modeling of the honeycomb active power distribution network;

[0023] Step S4, taking the maximum HSPH return on investment rate index and system stability index as the objective function. The return on investment rate index includes the life cycle cost and total revenue, where the life cycle cost is divided into initial investment cost and later maintenance cost, and the total revenue mainly includes peak-valley electricity price arbitrage revenue and auxiliary service revenue; the influence of the site selection scheme on the stability of the power distribution network structure is evaluated through the network connectivity index;

[0024] Step S5, optimizing the energy storage capacity of the honeycomb intelligent control power hub based on microgrid group net load time series data, combining energy transmission efficiency constraints and deep discharge correction.

[0025] Step S6, optimize the site selection strategy of HSPH through neighborhood operation and updating the tabu list, and generate the optimal topology result of the honeycomb active power distribution network.

[0026] Specifically, in step S1, multi-objective spectral clustering weight factors are constructed with line impedance and rated apparent power as constraints, and the specific process is as follows:

[0027] Considering the admittance and power carrying capacity of the line as two main electrical characteristics, the dynamic edge weight between node and node is defined as:

[0028] , wherein: represents the absolute value of the line admittance between node and node , represents the absolute value of the admittance of all lines in the distribution network; represents the maximum power carrying capacity of the line between node and node , represents the maximum power carrying capacity of all lines in the distribution network.

[0029] ;

[0030] , wherein: represents the weight of the line; is the admittance factor, is the power factor; is an adjustment factor, used to adjust the relative weight of the admittance factor and the power factor in the weight matrix .

[0031] The present application performs cluster division on the distribution network, and constructs a honeycomb active power distribution network, and the specific process is as follows:

[0032] Step S21, the improved spectral clustering algorithm realizes the node division of the distribution network by constructing a weighted adjacency matrix reflecting the electrical characteristics of the line and solving a normalized Laplacian matrix:

[0033] The degree matrix is defined as a diagonal matrix, and the diagonal element reflects the connection strength of the node in the network, and the calculation formula is as follows: ;

[0034] The Laplacian matrix is constructed and symmetrically normalized: ;

[0035] , wherein: is the inverse square root of the degree matrix; It is the weight matrix of the power distribution network; It is the Laplace matrix The symmetric normalized representation of .

[0036] Step S22: Based on the feature vector matrix, use the K-means clustering algorithm to partition the nodes, construct the feature vector matrix and normalize it, and select the feature vectors corresponding to the k smallest non-zero feature values. The feature matrix V is formed by columns;

[0037] In the formula: The preset number of clusters, For the first The center of each cluster, It is an indicator function, when the node Belongs to cluster hour Otherwise, it is 0;

[0038] Based on the cluster labels output by the K-means partitioning algorithm, the distribution network is divided into k subsets. The nodes contained in each subset and the connecting edges between the nodes constitute a candidate microgrid.

[0039] This invention proposes a sparse matrix modeling method for honeycomb active power distribution networks:

[0040] Step S31, the sparse matrix modeling method for honeycomb active distribution networks, is based on constructing a mathematical representation of the connection relationship between the HSPH and microgrids. Each microgrid is abstracted as a single node, ignoring its internal structure, and a connection matrix between the microgrid and the HSPH is constructed. The mapped honeycomb active distribution network topology can be viewed as an undirected, unweighted graph, and the connection relationship between the microgrid and the HSPH is represented by an adjacency matrix. Figure 2 This is the sparse matrix-based honeycomb distribution network model representation method used in this example.

[0041] Build 3D connection matrix ,in: ;

[0042] Step S32: Based on the connection matrix C, derive two additional matrices: the power interaction matrix and the fault response characteristic matrix. The power interaction matrix... , is defined as follows: when the connection matrix C has non-zero elements, Characterizes the maximum transmission power between the cellular intelligent power hub and the microgrid; flexible switching matrix Defined as when Forced during failure Disconnect the connection between the honeycomb intelligent power hub and the microgrid. During normal system operation... .

[0043] The application establishes an objective function of maximizing investment return rate and system stability, and the specific process is as follows:

[0044] In step S41, the life cycle cost index, the investment income index and the connectivity index are selected from the aspects of economy and stability to construct the objective function:

[0045] In the formula, LCC is the life cycle cost index, ROI is the investment income index, and C is the connectivity index. , is the index weight.

[0046] In step S42, the life cycle cost is calculated:

[0047] For the honeycomb active power distribution network, the life cycle cost is divided into initial investment cost and later maintenance cost.

[0048] In the formula, IC is the initial investment cost, including power construction cost and capacity construction cost. OC is the operation cost, RC is the recovery cost, and r is the standard discount rate or benchmark yield, and N is the total period number of cost calculation.

[0049] The initial investment cost usually includes all direct expenditures of project construction, which is composed of power-related cost (converter, transformer, etc.) and capacity-related cost (energy storage element), and the specific calculation formula is as follows:

[0050] In the formula, Pmax is the maximum output power of the energy storage system, unit is MW. Cp is the unit power cost of Pmax, indicating the construction cost of each MW power. Ct is the total storage capacity of the energy storage system, unit is kWh. Cc is the capacity unit cost, indicating the construction cost of each capacity.

[0051] The later maintenance cost is as follows:

[0052] In the formula, M is the cost of equipment maintenance and repair. R is the cost of energy storage element replacement. E is the cost of engineering management. ​​​​​​​​​​​​​​​​

[0053] Step S43, calculate the total income in the life cycle, the total income in the life cycle includes peak-valley electricity price arbitrage type income and auxiliary service type income, the formula is as follows: ;

[0054] Peak-valley electricity price arbitrage type income: ;

[0055] In the formula: represents the battery attenuation coefficient; represents the number of energy storage annual cycles; represents the single dischargeable capacity represents the first year peak-valley electricity price, considering the change of electricity price policy; represents the loss power per unit capacity; represents the energy conversion efficiency;

[0056] Auxiliary service type income: ;

[0057] In the formula: is a transaction characteristic coefficient, reflecting the situation of regional auxiliary service participating in the power market, the value range is (0, 1); represents the annual income of frequency modulation type service, represents the annual income of peak regulation type service;

[0058] ;

[0059] In the formula: represents the power capacity participating in frequency modulation, represents the frequency modulation response quality coefficient, represents the annual effective frequency modulation service time, represents the frequency modulation service remuneration standard, represents the annual peak regulation service cycle number, and respectively represent the charge and discharge compensation;

[0060] Step S44, calculate the connectivity index:

[0061] Two important indexes, algebraic connectivity and natural connectivity, are introduced to measure the overall stability and fault tolerance of the honeycomb intelligent control power hub network:

[0062] Connectivity gain: construct the algebraic connectivity index and natural connectivity index representing the stability of network connection: ;

[0063] In the formula: is a weight coefficient, , used to adjust the influence degree of the two sub-indices on the overall stability; is the second smallest eigenvalue of Laplacian matrix ; is the natural connectivity;

[0064] ;

[0065] wherein: is a vector orthogonal to the constant vector 1, The greater the value, the more closely connected the network nodes are, and the more robust the connection is, and the better connectivity can be maintained when some nodes or lines fail.

[0066] ;

[0067] wherein: is the eigenvalue based on the network adjacency matrix A, the network adjacency matrix A is used to represent the connection relationship of all nodes in the power distribution network, and V represents the total number of nodes of the honeycomb active power distribution network.

[0068] The application optimizes the results of the honeycomb intelligent control power hub site selection and capacity determination by using the tabu search algorithm to update the site selection scheme and calculate the corresponding energy storage capacity in iterations through neighborhood operation and tabu list updating.

[0069] Step S51, select a node with large load fluctuation, a storage node or a new energy node of photovoltaic as a candidate installation position of the energy storage-power exchange honeycomb intelligent control power hub.

[0070] Step S52, initialize the honeycomb intelligent control power hub site selection scheme based on the power distribution network division result through an algorithm, the site selection scheme includes the access position of the honeycomb intelligent control power hub and the feeder line of the newly built microgrid and the honeycomb intelligent control power hub, and the energy storage capacity is calculated according to the site selection scheme, the objective function is calculated and finally iteratively optimized; in the optimization process, the site selection and capacity determination result is back substituted through neighborhood operation, which can ensure that the site selection scheme is optimized according to the search radius and the tabu list when generated, and the situation of falling into a local optimal solution is avoided.

[0071] Figure 3 is the power distribution network division and base station site selection and access result of the present example. It can be seen from the figure that the power distribution network is divided into five microgrids according to the cluster division method, the five HSPHs are accessed into the microgrids in the figure through HSPH site selection, the microgrids are connected through HSPH, and the power feeder line is newly built to guarantee the connection of the HSPH and the microgrid. Figure 4 is a schematic diagram of the honeycomb active power distribution network topology in the present example, wherein Figure 3 is simplified, the honeycomb active power distribution network is modeled according to the sparse matrix in Figure 2 , and the required energy storage capacity of each HSPH is calculated and represented in the figure. Figure 5This example demonstrates the HSPH location and capacity optimization strategy based on tabu search, and details the optimization process.

[0072] The selection and optimization method of this invention includes a honeycomb intelligent power hub capacity selection strategy, the specific process of which is as follows:

[0073] Treat the microgrid topology as a whole and construct the net load time series matrix of the cluster within period T;

[0074] ;

[0075] In the formula: It is a honeycomb active distribution network unit at time t. Net load data, This represents the load data of the microgrid connected to the honeycomb intelligent power hub. and This represents the output data of distributed power sources;

[0076] Considering charging efficiency Discharge efficiency And the calculation of the depth of discharge index and theoretical capacity:

[0077] ;

[0078] In the formula: This indicates a power deficit in a honeycomb-shaped active distribution network unit; Depth of discharge (DOP) is an indicator of energy loss during the charging and discharging process. Constraints are used to avoid deep discharge damaging battery life, and calculation results ensure that the actual usable capacity is covered. ;

[0079] After each round of iterative optimization of the site selection strategy, the optimal capacity of the site selection result will be calculated to ensure the feasibility of each round of optimization results.

[0080] The location and capacity constraints of the honeycomb intelligent power hub of this invention are as follows:

[0081] To maintain the structural characteristics of the honeycomb active power distribution network, a decentralized honeycomb intelligent power hub configuration strategy is adopted, and the number of microgrids connected to each honeycomb intelligent power hub needs to be limited:

[0082] ;

[0083] In the formula: This is a binary variable representing the honeycomb intelligent power hub. Whether to connect to a microgrid; The maximum number of microgrids that can be connected to a single cellular smart power hub, to match the hexagonal structure of the cellular network. .

[0084] The fluctuation of distributed generation (DG) and the change of load may cause the voltage of nodes to exceed the limit, especially in the distribution network with high penetration of DG. The voltage amplitude of each node in the distribution network must be maintained within a safe range, and the voltage of nodes caused by the fluctuation of DG and the change of load must be prevented from exceeding the limit: ;

[0085] In the formula: is the voltage amplitude of node is the voltage amplitude of node and are the lower and upper limits of voltage, respectively; usually .

[0086] The output of distributed generation (DG) is limited by the rated capacity: ;

[0087] In the formula: is the active power of the output of distributed generation, is the reactive power of the output of distributed generation.

[0088] The power exchange capacity of the honeycomb intelligent control power hub is limited by the capacity of its converter: ;

[0089] In the formula: represents the power of the energy storage system equipped in the HSPH at time t; is the maximum charging power of the energy storage system, is the maximum discharging power of the energy storage system, is because a safety margin needs to be reserved when charging; represents the capacity of the converter equipped in the HSPH.

[0090] The sudden change of the power of the energy storage system may cause the risk of battery electrochemical polarization and thermal runaway, so the rate of change of its power needs to meet the maximum ramp rate limit: ;

[0091] In the formula: is the maximum ramp rate of the energy storage system.

[0092] The state of charge (SOC) of the energy storage system of the honeycomb intelligent control power hub needs to be dynamically updated according to the charging and discharging behavior: ;

[0093] Lithium batteries and other chemical batteries age faster when they are close to full charge or deep discharge. By setting a safety margin, the service life of the battery can be effectively extended, and the replacement cost can be reduced. The state of charge operating range constraint: ;

[0094] wherein: represents the battery discharge efficiency at time t, represents the battery charge efficiency at time t, represents the battery state of charge at time t.

[0095] To prolong the battery life, the state of charge should avoid approaching the upper and lower limits in actual operation: ;

[0096] wherein: is the safety margin value; is the rated capacity of the energy storage.

[0097] To simplify the calculation, the linearized DistFlow model is used to describe the power flow of the distribution network. The low-voltage distribution network line usually presents a resistive characteristic, so the active power flow is mainly affected by the difference in the node voltage amplitude, and the linearization of the active power balance constraint is:

[0098] ;

[0099] wherein: is the node net active load, is the line impedance.

[0100] The above examples are only used to illustrate the design idea and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the content of the present application and to implement it, and the protection scope of the present application is not limited to the above examples. Therefore, any equivalent changes or modifications made according to the principles and design ideas disclosed by the present application are within the protection scope of the present application.

Claims

1. A graph-based method for topology selection and optimization of a honeycomb active power distribution network, characterized in that, The method comprises the following steps: Step S1, constructing a multi-objective spectral clustering weight factor with line impedance and rated apparent power as constraints; Step S2, adopting an improved spectral clustering algorithm to perform cluster division on the distribution network to construct a honeycomb active distribution network distribution, maximize the internal edge weight of the microgrid, and minimize the weight between microgrids, and each microgrid cluster constitutes a honeycomb active distribution network unit; Step S3, abstracting each microgrid as a single node, constructing a mathematical representation of the connection relationship between the honeycomb intelligent control power hub and the microgrid, representing the connection relationship between the microgrid and the honeycomb intelligent control power hub through an adjacency matrix, and completing the sparse matrix modeling of the honeycomb active distribution network; Step S4, taking the maximum investment return rate index and system stability index of the honeycomb intelligent control power hub as the objective function; the investment return rate index includes the life cycle cost and the total income, wherein the life cycle cost is divided into initial investment cost and later maintenance cost, and the total income includes peak-valley electricity price arbitrage income and auxiliary service income; the influence of the site selection scheme on the stability of the distribution network structure is evaluated through the network connectivity index; Step S5, optimizing the energy storage capacity of the honeycomb intelligent control power hub based on the microgrid group net load time series data, combining the energy transmission efficiency constraint and the deep discharge correction; Step S6, optimizing the site selection strategy of the honeycomb intelligent control power hub through neighborhood operation and updating the taboo table, and generating the optimal topology result of the honeycomb active distribution network; In step S1, the multi-objective spectral clustering weight factor is constructed with line impedance and rated apparent power as constraints, and the specific process is as follows: Considering both admittance and power carrying capability of lines, the dynamic edge weight between node and node is defined as: where is the absolute value of line admittance between node and node , is the absolute value of all line admittances in the distribution network; is the maximum power carrying capability of the line between node and node , is the maximum power carrying capability of all lines in the distribution network. ； wherein: represents the weight of the line; is the admittance factor, is the power factor; is a scaling factor for scaling the relative weight of the admittance factor and the power factor in the weight matrix ; The cluster division is performed on the distribution network to construct a honeycomb active distribution network, and the specific process is as follows: In step S21, the improved spectral clustering algorithm realizes the node division of the distribution network by constructing a weighted adjacency matrix reflecting the electrical characteristics of the line and solving a normalized Laplacian matrix: Definition of the degree matrix is a diagonal matrix, the diagonal elements reflects the connection strength of the node in the network, and the calculation formula is as follows: ; Construct Laplacian matrix and symmetric normalize: ; wherein: is the inverse square root of the degree matrix; is the weight matrix of the power distribution network; is the symmetric normalized representation of the Laplacian matrix of the power distribution network. Step S22, based on the feature vector matrix, using K-means clustering algorithm for node division, constructing feature vector matrix and normalizing, selecting the first k minimum non-zero eigenvalue corresponding to the feature vector , form a feature matrix V by column In the formula: The preset number of clusters, For the first The center of each cluster, It is an indicator function, when the node Belongs to cluster hour Otherwise, it is 0; According to the cluster labels output by the K-means division algorithm, the distribution network is divided into k subsets, and each subset contains nodes and the connection edges between the nodes, which form a candidate microgrid; A sparse matrix modeling method of the honeycomb active distribution network is proposed: Step S31, constructing connectivity matrix wherein: ; Step S32, deriving two other matrices, power interaction matrix and fault response characteristic matrix, on the basis of the connection matrix C , defined as when the connection matrix C is non-zero elements, , representing the maximum transmission power between the honeycomb intelligent control power hub and the microgrid; flexible switching matrix , defined as when , disconnect the connection between the honeycomb intelligent control power hub and the microgrid during fault , the system is running normally .

2. The method of claim 1, wherein the method is based on graph theory. The objective function of maximizing the investment return rate and system stability is established, and the specific process is as follows: Step S41, from the aspects of economy and stability, select the life cycle cost index, investment income index and connectivity index to construct the objective function: ; wherein: is a full life cycle cost indicator, is an investment return indicator, is a connectivity indicator; , is an indicator weight; In step S42, the life cycle cost is calculated: For the honeycomb active power distribution network, the life cycle cost is divided into initial investment cost and later maintenance cost: ; In the formula: is the initial investment cost, including power construction cost and capacity construction cost; is the operation cost, is the recovery cost, is the standard discount rate or benchmark yield, and N is the total number of periods for cost calculation. Initial investment cost: ; In the formula: Pmax is the maximum output power of the energy storage system, unit ; Pc is the unit power cost of, representing the construction cost per unit of power; Etot is the total storage capacity of the energy storage system, unit ; Cc is the capacity unit cost, representing the construction cost per unit of capacity; Cost of later maintenance: ; In the formula: costs for equipment maintenance and repair; costs for replacement of energy storage elements; costs for project management; Step S43, total income in the whole life cycle is calculated, the total income in the whole life cycle includes peak-valley electricity price arbitrage income and auxiliary service income, and the formula is as follows: ; Peak valley electricity price arbitrage class income: ; In the formula: represents the battery attenuation coefficient; represents the number of annual energy storage cycles; represents the single-time dischargeable amount represents the first year peak-valley electricity price, considering the change of electricity price policy; represents the loss power per unit capacity; represents the energy conversion efficiency; Auxiliary service class revenue: ; In the formula, is a transaction characteristic coefficient, reflecting the participation of regional ancillary services in the electricity market, and the value range is (0, 1); represents the annual income of frequency modulation services, represents the annual income of peak regulation services; ； wherein: represents the power capacity participating in frequency modulation, represents the frequency modulation response quality coefficient, represents the annual effective frequency modulation service time, represents the frequency modulation service compensation standard, represents the annual frequency modulation service cycle times, and respectively represent the charge and discharge compensation; In step S44, the connectivity index is calculated: Two important indexes, algebraic connectivity and natural connectivity, are introduced to measure the overall stability and fault tolerance of the honeycomb intelligent control power hub network: Connectivity gain: Constructing algebraic and natural connectivity indices that represent the stability of network connectivity: ; In the formula: is a weight coefficient, for adjusting the degree of influence of the two sub-indicators on the overall stability; is the algebraic connectivity, representing the second smallest eigenvalue of the Laplacian matrix ; and is the geometric connectivity. ； In the formula: is a vector orthogonal to the constant vector 1, The larger the value is, the more robust the connection between the network nodes is, and the better connectivity can be maintained when some nodes or lines fail. ； In the formula: is an eigenvalue based on a network adjacency matrix A, the network adjacency matrix A is used to represent the connection relationship of all nodes in the power distribution network, and V represents the total number of nodes of the honeycomb active power distribution network.

3. The method of claim 2, wherein: the method further comprises: determining a plurality of candidate topologies for the active power distribution network; and selecting a topology from the plurality of candidate topologies based on the plurality of candidate topologies. Through neighborhood operation and updating the taboo table, the site selection scheme is updated in iteration and the corresponding energy storage capacity is calculated, and the taboo search algorithm is used to optimize the site selection and capacity determination result of the honeycomb intelligent control power hub, and the method is as follows: In step S51, the nodes with large load fluctuations, energy storage nodes or new energy nodes of photovoltaic are selected as the candidate installation positions of the energy storage-power exchange honeycomb intelligent control power hub; In step S52, the honeycomb intelligent control power hub site selection scheme is initialized based on the distribution network division result through the algorithm, the site selection scheme includes the access position of the honeycomb intelligent control power hub and the feeder line between the newly built microgrid and the honeycomb intelligent control power hub, the energy storage capacity is calculated according to the site selection scheme, the objective function is calculated, and finally iteration optimization is performed. In the optimization process, the results of site selection and capacity determination are back substituted through neighborhood operation.

4. The method of claim 3, wherein the method further comprises: determining a number of nodes in the active power distribution network; and determining a number of edges in the active power distribution network. The specific process is as follows: The microgrid topology is taken as a whole, and the net load time sequence matrix of the cluster within a period T is constructed. ； wherein: is the net load data of the cellular active distribution grid unit at time t, represents the load data of the microgrid connected to the cellular smart grid power hub, and represents the distributed power generation output data;​ Consider charging efficiency , discharging efficiency and depth of discharge index Calculate the depth of discharge index theoretical capacity: ； wherein: represents the power deficit of the honeycomb active power distribution network unit; represents the loss of the charge and discharge energy conversion process, the depth of discharge index constrains to avoid deep discharge damage to the battery life, the calculation result ensures that the actual available capacity is covered ; After the site selection strategy is optimized in each round, the optimal capacity of the site selection result is calculated to ensure the feasibility of the optimization result of each round.

5. The graph-based method for topology selection and optimization of a honeycomb active power distribution network according to claim 4, wherein: The constraints of the site selection and capacity determination of the honeycomb intelligent control power hub are as follows: In order to maintain the structural characteristics of the honeycomb active distribution network, the decentralized honeycomb intelligent control power hub configuration strategy is adopted, and the number of microgrids connected to each honeycomb intelligent control power hub needs to be limited. ； wherein: is a binary variable indicating whether the honeycomb smart power hub is connected to the microgrid; is the maximum number of microgrids that a single honeycomb smart power hub is allowed to connect to, matching the hexagonal structure of the honeycomb, ; The voltage amplitude of each node in the power distribution network must be maintained within a safe range, and the node voltage out-of-limit caused by the output fluctuation of the distributed power and the load change must be prevented: ; wherein: is the node node is the voltage amplitude at time t, and is the lower and upper voltage limit, respectively; The distributed power output is limited by the rated capacity: ; wherein: Pd, is the active power of the distributed power source output, Qd, is the reactive power of the distributed power source output; The power exchange capability of a cellular smart power hub is limited by the capacity of its power converters: ; In the formula: P(t) represents the power of the energy storage system equipped by the HSPH at time t; Pmax,charge is the maximum charging power of the energy storage system, Pmax,discharge is the maximum discharging power of the energy storage system, Pmax,charge is because a safety margin needs to be reserved during charging; Pmax,converter represents the capacity of the converter equipped by the HSPH; The power mutation of the energy storage system can cause the risk of battery electrochemical polarization and thermal runaway, and thus the power change rate of the energy storage system needs to meet the maximum ramp rate limit: ; In the formula: is the maximum ramp rate of the energy storage system; The state of charge of the honeycomb intelligent control power hub energy storage system needs to be dynamically updated according to the charging and discharging behavior: ; State of charge operating range constraints: ; In the formulae: represents the battery discharge efficiency at time t, represents the battery charge efficiency at time t, represents the battery state of charge at time t; To prolong the battery life, the state of charge should be avoided to approach the upper and lower limits in actual operation: ; In the formula: is a safety margin value; is the rated capacity of the energy storage; The linear DistFlow model is used to describe the power flow of the distribution network, and the active power balance constraint is linearized as follows: ； where: is the node net real load, is the line impedance.

Images