Method and device for determining embedded direct current network connection topology

Abstract

The invention discloses an embedded DC network connection topology determination method and device, and the method comprises the steps: determining a target DC planning mode based on a multi-dimensional evaluation index; determining a target drop point based on the geographical environment parameters and the electrical parameters; according to the target direct current planning mode and the target drop point, a plurality of candidate topological structures are respectively designed according to a plurality of topological types, a multi-objective optimization function is constructed by taking the power transmission capacity, the line loss and the stability margin as optimization objectives, solving is carried out, and the optimal topology is determined in the plurality of candidate topological structures; and performing simulation verification on the optimal topology in a plurality of scenes to obtain a final topology. Through multi-dimensional direct current planning mode comparison, double-layer drop point quantitative screening, topological optimization and multi-scene adaptability verification, the embedded direct current network connection topology can be efficiently determined, the optimal adaptation of the embedded direct current and the provincial power grid is realized, and the requirements of large-scale new energy transmission and high-load density power supply in the provincial power grid are met.

Description

Technical Field

[0001] This invention relates to the field of power system planning technology, and specifically to a method and apparatus for determining the connection topology of an embedded DC network. Background Technology

[0002] With social development, provincial power grids face the dual pressures of large-scale integration of new energy sources and continuous load growth. Taking Jiangsu power grid as an example, in 2024, the province's installed wind and solar power capacity reached 84.86 million kilowatts, accounting for 40.4% of the total installed power capacity, with a maximum load of 147 million kilowatts. It exhibits characteristics of "north-to-south power transmission" and "reverse distribution of new energy-rich areas and load centers." Problems such as insufficient transmission capacity of river-crossing channels, excessive short-circuit current of 500kV grid, and weak regional mutual assistance capacity of 220kV grid are prominent.

[0003] Embedded DC technology, as a key means to improve the power grid's transmission capacity and flexibility, relies heavily on the selection of its landing point and the design of its network connection topology to determine the effectiveness of its application. However, existing solutions suffer from the following core problems: (1) DC planning methods are too simplistic: Currently, most rely on traditional AC enhancement or two-end DC methods, without considering the adaptability of multi-end embedded DC, and lack quantitative comparison models to support selection.

[0004] (2) Experience-based selection of landing points: Currently, the determination of landing points mostly relies on engineering experience and has not been systematically quantitative. This may lead to power flow blockage or insufficient voltage support after some landing points are put into operation.

[0005] (3) Poor network topology adaptability: Most of the current connection topologies adopt a fixed structure (such as chain or radial), which may not match the installation scenario and may lead to uncontrolled power flow transfer during faults (such as DC N-1 fault causing AC line overload).

[0006] (4) Incomplete verification of the solution: Currently, most of them only verify the normal operating conditions, the verification scenarios are not fully covered, and there is a lack of standardized simulation verification process, making it difficult to ensure the safety of the solution under complex operating conditions.

[0007] Therefore, how to efficiently determine the connection topology of embedded DC networks to meet the needs of provincial power grids for new energy transmission and high load density power supply is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0008] To address the shortcomings of the existing technologies, this invention provides a method and apparatus for determining the topology of an embedded DC network. By comparing multiple DC planning methods, performing dual-layer landing point quantification screening, optimizing the topology, and verifying adaptability in multiple scenarios, the method can efficiently determine the topology of the embedded DC network, achieving optimal adaptation between the embedded DC network and the provincial power grid, in order to meet the needs of large-scale new energy transmission and high load density power supply in the provincial AC / DC hybrid power grid.

[0009] In a first aspect, the present invention provides a method for determining the connection topology of an embedded DC network, comprising: Based on multi-dimensional evaluation indicators, the target DC planning method is determined; Determine the target landing point based on geographical and electrical parameters; Based on the target DC planning method and the target landing point, multiple candidate topologies are designed according to various topology types. With transmission capacity, line loss, and stability margin as optimization objectives, a multi-objective optimization function is constructed and solved to determine the optimal topology among the multiple candidate topologies. The optimal topology is simulated and verified in multiple scenarios to obtain the final topology, thus completing the determination of the embedded DC network connection topology.

[0010] Furthermore, based on geographical and electrical parameters, the target landing point is determined, including: Determine candidate landing points; For each candidate landing point, a membership matrix is ​​constructed based on the index values ​​of the geographical environment parameters and electrical parameters corresponding to that candidate landing point. Multi-level fuzzy operations are performed based on the membership matrix of the candidate landing point to obtain the comprehensive score of the candidate landing point. Candidate landing points whose overall score is greater than the score threshold are taken as the target landing points.

[0011] Furthermore, the geographical environmental parameters include: power transmission corridor resources, urban planning compatibility, and geological conditions; power transmission corridor resources include: existing corridor reuse rate and new corridor length; urban planning compatibility includes: distance from urban built-up areas and overlap with industrial parks; geological conditions include: foundation bearing capacity and seismic intensity. Electrical parameters include short-circuit current level, power flow distribution uniformity, and voltage stability margin; short-circuit current level includes short-circuit current value and short-circuit ratio during a fault; power flow distribution uniformity includes the standard deviation of line load rate and N-1 power flow transfer rate; voltage stability margin includes static voltage stability margin and transient voltage recovery time.

[0012] Furthermore, based on the index values ​​of the geographical environment parameters and electrical parameters corresponding to the candidate landing point, a membership matrix for the candidate landing point is constructed, including: Obtain the values ​​of each underlying indicator for the candidate landing point; Based on the stated values, determine the membership degree of the candidate landing point under each evaluation level of each underlying indicator, and construct the membership degree matrix of the candidate landing point.

[0013] Furthermore, based on the membership matrix of the candidate landing point, multi-level fuzzy operations are performed to obtain the comprehensive score of the candidate landing point, including: Based on the geographical environment parameters, electrical parameters, and the indicators they contain, determine the parameter weights and the weights of the indicators at each level. Based on the membership matrix of the candidate landing point and the weights of the lower-level indicators, the evaluation vector of the upper-level indicators is calculated. Based on the evaluation vectors and weights of the upper-level indicators, the evaluation vectors of the geographical environment parameters and the electrical parameters are calculated. Based on the evaluation vectors of geographical environmental parameters, electrical parameters, and parameter weights, the comprehensive evaluation vector of the candidate landing point is calculated. The comprehensive evaluation vector of the candidate landing point is used as the weight of the evaluation level. The weighted sum of the scores corresponding to each evaluation level is calculated to obtain the comprehensive score of the candidate landing point.

[0014] Furthermore, based on multi-dimensional evaluation indicators, the target DC planning method is determined, including: The weights of each dimension are determined using the analytic hierarchy process (AHP). The standardized scores of each DC planning method are obtained in each dimension. The DC planning methods include traditional AC enhancement, two-terminal DC, and multi-terminal embedded DC. For each DC planning method, a weighted sum is performed based on the weight of each dimension and the standardized score of the DC planning method in each dimension to obtain the comprehensive score of the DC planning method. The DC planning method with the highest comprehensive score will be selected as the target DC planning method. The multi-dimensional evaluation indicators include technical, economic, environmental, and safety dimensions. The technical dimensions include transmission capacity improvement rate, renewable energy absorption capacity, fault ride-through success rate, and voltage support strength. The economic dimensions include unit capacity investment cost, annual operating loss cost, and life-cycle benefits. The environmental dimensions include transmission corridor area, carbon emission reduction, and noise impact range. The safety dimensions include short-circuit current control effect, N-1 fault overload rate, and transient stability margin.

[0015] Furthermore, the topology types include chain topology, radial topology, ring topology, and mesh topology; Using transmission capacity, line loss, and stability margin as optimization objectives, a multi-objective optimization function is constructed and solved to determine the optimal topology from among various candidate topologies, including: A multi-objective optimization function is constructed with the objectives of maximizing transmission capacity, minimizing line loss, and maximizing stability margin, specifically expressed as follows:

[0016] Where F represents the multi-objective optimization function, α, β, and γ are the weights of transmission capacity, line loss, and stability margin, respectively, P is the transmission capacity, ΔP is the line loss, and K... stab For stability margin, P max For the transmission capacity threshold, ΔP max K is the line loss threshold. stab,max This is the stability margin threshold; Define the constraints, including transmission capacity constraints, line loss constraints, voltage stability constraints, and short-circuit current constraints; Under constraints, a multi-objective optimization function is solved to determine the optimal topology from the various candidate topologies.

[0017] Furthermore, under constraints, a multi-objective optimization function is solved to determine the optimal topology among the various candidate topologies, specifically including: A chaotic mapping method is used to initialize the particle swarm, where each particle represents a topological structure and the particle position represents the topological structure parameters. For each particle, based on its current position, the multi-objective value of the particle is determined, and the multi-objective optimization function value is calculated to obtain the current fitness of the particle. The multi-objective value includes the transmission capacity value, the line loss value, and the stability margin value. Based on the multi-objective values ​​of all particles, perform non-dominated sorting to determine the non-dominated solution set; Based on adaptive weights and neighborhood search mechanism, the particle velocity and position are updated. If the current fitness of the particle is greater than its historical best fitness, the individual historical best position of the particle is updated to the current position of the particle, and the non-dominated solution set is updated based on the current non-dominated sorting result. Repeat the above process of updating particle velocity and position, individual historical best position, and non-dominated solution set until the preset number of iterations is reached or the distance change of crowding in the non-dominated solution set is less than the preset threshold after N consecutive iterations. Determine the optimal topology based on the current non-dominated solution set.

[0018] Furthermore, the optimal topology is simulated and verified in multiple scenarios to obtain the final topology, including: For the aforementioned optimal topology, a power grid simulation model is constructed; The power flow distribution under normal operation, the power flow transfer under N-1 fault, the second-level restart capability under DC instantaneous fault, and the stability capability under renewable energy fluctuation are verified respectively to determine the final topology.

[0019] Secondly, the present invention also provides an apparatus for implementing the method for determining the connection topology of any of the embedded DC network described above, comprising: The first determination module is used to determine the target DC planning method based on multi-dimensional evaluation indicators; The second determination module is used to determine the target landing point based on geographical environmental parameters and electrical parameters; The topology optimization module is used to design multiple candidate topologies according to the target DC planning method and the target landing point, and to determine the optimal topology among the multiple candidate topologies with transmission capacity, line loss and stability margin as optimization objectives. The simulation verification module is used to perform simulation verification on the optimal topology in multiple scenarios to obtain the final topology and complete the determination of the embedded DC network connection topology.

[0020] The method and apparatus for determining the embedded DC network connection topology provided by the present invention have at least the following beneficial effects: (1) This invention can efficiently determine the connection topology of embedded DC network by comparing DC planning methods in multiple dimensions, quantitative screening of landing points in two layers, topology optimization and multi-scenario adaptability verification, and realize the optimal adaptation of embedded DC and provincial power grid to meet the needs of large-scale new energy transmission and high load density power supply in provincial AC-DC hybrid power grid. It solves the problems of the existing DC planning method being too simple, the landing point selection being based on experience, poor network topology adaptability, and incomplete scheme verification, and provides technical support for provincial power grid to solve problems such as excessive short-circuit current, power flow blockage, and insufficient new energy absorption.

[0021] (2) For several common DC planning methods in provincial power grids, a four-dimensional evaluation index system of "technology-economy-environment-safety" was constructed, and a multi-dimensional DC planning method comparison model was constructed. Through the analytic hierarchy process, the optimal DC planning method suitable for new energy transmission and high load density power supply scenarios was objectively and scientifically selected.

[0022] (3) Constructing a two-layer index system of “geographical environment parameters-electrical parameters” to screen the receiving and sending points can ensure that the landing points take into account both channel resources and power grid safety.

[0023] (4) Based on the selected landing point, combined with the scenarios of new energy transmission and load power reception, a variety of topologies are designed, and a multi-objective optimization function is constructed with the goal of "maximum transmission capacity, minimum line loss, and optimal stability margin". The optimal topology is solved by the particle swarm optimization algorithm to balance capacity, loss and stability, and the optimized design of network connection topology is realized, which can be dynamically adapted to the fluctuation of new energy output and load distribution characteristics.

[0024] (5) By constructing a power grid simulation model, the safety and adaptability of the optimal topology scheme in multiple scenarios are verified. In addition to normal operation scenarios, extreme scenarios are also included, realizing the adaptability verification in multiple scenarios. The verification is comprehensive, ensuring the safety of the scheme under complex working conditions and ensuring that the scheme meets engineering requirements. Attached Figure Description

[0025] Figure 1 A flowchart illustrating a method for determining the connection topology of an embedded DC network provided by the present invention; Figure 2 A flowchart illustrating a comparison of multi-dimensional DC planning methods according to one embodiment of the present invention; Figure 3 A schematic diagram of the two-layer screening process for landing points according to one embodiment of the present invention; Figure 4 A schematic diagram of a topology optimization process according to one embodiment of the present invention; Figure 5 This invention provides a structural block diagram of an embedded DC network connection topology determination device. Detailed Implementation

[0026] To better understand the above technical solutions, a detailed description of the solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0027] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0028] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0029] like Figure 1 As shown, the present invention provides a method for determining the connection topology of an embedded DC network, comprising the following steps: S1, based on multi-dimensional evaluation indicators, determines the target DC planning method; S2, based on geographical and electrical parameters, determines the target landing point; S3. Based on the target DC planning method and the target landing point, design multiple candidate topologies according to multiple topology types, and construct a multi-objective optimization function with the optimization objectives of transmission capacity, minimum line loss, and stability margin as the optimization objectives, and solve it to determine the optimal topology among the multiple candidate topologies. S4. Simulate and verify the optimal topology in multiple scenarios to obtain the final topology and complete the determination of the embedded DC network connection topology.

[0030] This invention, through multi-dimensional comparison of DC planning methods, dual-layer quantitative screening of landing points, topology optimization, and multi-scenario adaptability verification, can efficiently determine the connection topology of embedded DC networks, achieving optimal adaptation between embedded DC and provincial power grids to meet the needs of large-scale renewable energy transmission and high load density power supply in provincial AC / DC hybrid power grids. Based on multi-dimensional evaluation indicators, the optimal DC planning method can be selected through quantitative comparison. A dual-layer indicator system of "geographical environment parameters - electrical parameters" is constructed to screen sending and receiving end landing points, ensuring that landing points take into account both channel resources and grid security. This invention also combines renewable energy transmission and load receiving. In real-world scenarios, multiple topologies were designed with the goals of maximizing transmission capacity, minimizing line losses, and ensuring the highest stability margin. The optimal topology was determined from among these options to ensure that it is compatible with the scenario and can balance capacity, losses, and stability. Simulations were used to verify the safety and adaptability of the optimal topology in multiple scenarios, ensuring that the topology meets the requirements of complex operating conditions and engineering projects. This approach addresses the problems of existing DC planning methods being too simplistic, relying on experience in selecting landing points, having poor network topology adaptability, and lacking comprehensive solution verification. It provides technical support for provincial power grids to solve problems such as excessive short-circuit current, power flow congestion, and insufficient renewable energy absorption.

[0031] The following sections will provide a detailed explanation of each of the above steps.

[0032] (1) On the determination of DC planning method DC transmission planning methods include: traditional AC enhancement, two-terminal DC, and multi-terminal embedded DC. Traditional AC enhancement refers to converting existing AC lines into DC lines. Two-terminal DC refers to a DC transmission system consisting of two converter stations and DC lines. Multi-terminal embedded DC refers to a DC transmission system with three or more converter stations.

[0033] The multi-dimensional evaluation indicators include: technological, economic, environmental, and safety dimensions.

[0034] The technical dimensions include: transmission capacity improvement rate (capacity increase of embedded DC relative to AC), renewable energy absorption capacity (annual renewable energy absorption increment), fault ride-through success rate (probability of DC continuous operation under AC fault ≥95%), and voltage support strength (voltage stability margin of converter station bus ≥10%).

[0035] The economic dimensions include: unit capacity investment cost (RMB 10,000 / GW), annual operating loss cost, and total life cycle benefit. For example, the unit capacity investment cost of multi-terminal embedded DC is approximately RMB 30 million / GW, which is 15% lower than that of dual-terminal DC; regarding the annual operating loss cost, the DC line loss rate is ≤3%, which is 50% lower than that of AC lines; regarding the total life cycle benefit, the annual electricity cost savings are determined based on a 20-year lifespan.

[0036] Environmental dimensions include: transmission corridor footprint (km² / GW), carbon emission reduction (t / GW·a), and noise impact range. For example, multi-terminal embedded DC multiplexing of existing channels reduces footprint by 80%; DC converter station noise is ≤55dB, 10dB lower than AC substations.

[0037] Safety dimensions include: short-circuit current control effectiveness, N-1 fault overload rate, and transient stability margin. For example, after embedding, the AC grid short-circuit current is reduced by 10%-15%, the line load rate after a fault is ≤100%, and the power angle stability margin is ≥15°. Among these, short-circuit current control effectiveness refers to the ability to limit the short-circuit current within the allowable range, and is evaluated by indicators such as short-circuit current reduction rate, short-circuit current exceedance rate, and equipment withstand capability satisfaction rate.

[0038] like Figure 2 As shown, based on multi-dimensional evaluation indicators, the target DC planning method is determined, including: S11, using the analytic hierarchy process (AHP) to determine the weights of each dimension; S12, obtain the standardized scores of each DC planning method in each dimension; S13. For each DC planning method, a weighted sum is performed based on the weight of each dimension and the standardized score of the DC planning method in each dimension to obtain the comprehensive score of the DC planning method. S14, the DC planning method with the highest comprehensive score is taken as the target DC planning method.

[0039] The process of determining the weights of each dimension using the Analytic Hierarchy Process (AHP) is as follows: 1) Compare the importance of each dimension pairwise and construct a judgment matrix.

[0040] 2) Consistency test: Calculate the consistency ratio CR. If CR ≤ 0.1, it indicates that the judgment matrix is ​​consistent. CR = CI / RI, where CI is the consistency index, derived from the largest eigenvalue of the judgment matrix, and RI is the random consistency index, the value of which is related to the order of the judgment matrix.

[0041] 3) Process the judgment matrix to obtain the weight vector. The weight vector contains the weights for each dimension. For example, the technology dimension is more important than the economic dimension, so the resulting weight vector is W = [0.35, 0.25, 0.2, 0.2]. Specifically, the arithmetic mean, geometric mean, or eigenvalue method can be used to obtain the weight vector. Taking the eigenvalue method as an example, calculate the largest eigenvalue of the judgment matrix and its corresponding eigenvector, and normalize the eigenvector to obtain the weight vector. The weight for the technology dimension should be no less than 0.3, and the weight for the security dimension should be no less than 0.25.

[0042] Specifically, for each DC planning method, its comprehensive score can be calculated using the following formula:

[0043] Where S is the overall score of the DC planning method, and W k S represents the weight of the k-th dimension. k The standardized score (0-1 points) for this DC planning method in the k-th dimension.

[0044] The standardized scores of DC planning methods across various dimensions can be obtained through expert scoring or linear standardization, and this invention does not impose any limitations on this. Taking linear standardization as an example, for benefit-type indicators where a larger score is always better... S1 is the standardized score of a certain DC planning method on a certain indicator, and X is the raw score of the DC planning method on that indicator. min X represents the lowest raw score for this metric among all DC planning methods. max The highest raw score for this metric among all DC planning methods.

[0045] For example, in the scenario from northern Jiangsu to southern Jiangsu, the comprehensive score for traditional AC reinforcement is 0.52, the comprehensive score for two-terminal DC is 0.68, and the comprehensive score for multi-terminal embedded DC is 0.85. Therefore, multi-terminal embedded DC is determined to be the optimal DC planning method.

[0046] This embodiment constructs a four-dimensional evaluation index system of "technology-economy-environment-safety" for three common DC planning methods in provincial power grids. It realizes the construction of a multi-dimensional DC planning method comparison model, and through the analytic hierarchy process, it quantitatively compares and objectively and scientifically selects the optimal DC planning method suitable for new energy transmission and high load density power supply scenarios.

[0047] (2) Regarding the determination of the landing point Geographical environmental parameters include: power transmission corridor resources, urban planning compatibility, and geological conditions.

[0048] Transmission corridor resources include: the reuse rate of existing corridors and the length of newly added corridors. For example, an existing corridor reuse rate of ≥70% is preferred, the Sutong GIL corridor has a reuse rate of 100%, and a newly added corridor length of ≤50km is preferred.

[0049] Urban planning compatibility includes: distance from the urban built-up area and overlap with industrial parks. For example, a distance of ≥1km from the urban built-up area is preferred, and an overlap with industrial parks of ≤10% is preferred.

[0050] Geological conditions include: foundation bearing capacity and seismic intensity. For example, a foundation bearing capacity ≥ 200 kPa is considered excellent, and a seismic intensity ≤ VI is also considered excellent.

[0051] Electrical parameters include: short-circuit current level, power flow uniformity, and voltage stability margin.

[0052] Short-circuit current level includes: short-circuit current value during a fault and short-circuit ratio. For example, a short-circuit current value ≤ 50kA during a fault is preferred, and a short-circuit ratio SCR ≥ 3 is preferred.

[0053] Power flow distribution uniformity includes: standard deviation of line load rate and N-1 power flow transfer rate. For example, a standard deviation of line load rate ≤ 15% is considered excellent, and an N-1 power flow transfer rate ≤ 30% is considered excellent.

[0054] Voltage stability margin includes static voltage stability margin and transient voltage recovery time. For example, a static voltage stability margin ≥10% is preferred, and a transient voltage recovery time ≤0.1 is preferred.

[0055] As shown above, both geographical environmental parameters and electrical parameters contain two levels of indicators (upper-level indicators and lower-level indicators).

[0056] The target landing point can be determined through fuzzy comprehensive evaluation.

[0057] like Figure 3 As shown, the target landing point is determined based on geographical and electrical parameters, including: S21, determine the candidate landing point; S22, For each candidate landing point, construct the membership matrix of the candidate landing point based on the index values ​​of the geographical environment parameters and the electrical parameters corresponding to the candidate landing point; S23. Perform multi-level fuzzy operations based on the membership matrix of the candidate landing point to obtain the comprehensive score of the candidate landing point; S24, the candidate landing points with a comprehensive score greater than the score threshold are taken as the target landing points.

[0058] Furthermore, based on the index values ​​of the geographical environment parameters and electrical parameters corresponding to the candidate landing point, a membership matrix for the candidate landing point is constructed, including: Obtain the values ​​of each underlying indicator for the candidate landing point; Based on the stated values, determine the membership degree of the candidate landing point under each evaluation level of each underlying indicator, and construct the membership degree matrix of the candidate landing point.

[0059] For example, four evaluation levels can be set: Excellent, Good, Average, and Poor, with corresponding scores of 95, 85, 70, and 50 points, respectively. The degree to which a candidate endpoint belongs to different evaluation levels for a given indicator is called its membership degree. For example, the membership degree of candidate endpoint A in the indicator "Existing Channel Utilization Rate" (which is considered Excellent) is 0.9. The membership degree matrix of candidate endpoint A in the transmission corridor resources... for:

[0060] The first row represents the membership degree of candidate point A under each evaluation level of the underlying indicator "existing channel reuse rate" (from left to right: excellent, good, medium and poor). The second row represents the membership degree of candidate point A under each evaluation level of the underlying indicator "new channel length".

[0061] Furthermore, based on the membership matrix of the candidate landing point, multi-level fuzzy operations are performed to obtain the comprehensive score of the candidate landing point, including: Based on the geographical environment parameters, electrical parameters, and the indicators they contain, determine the parameter weights and the weights of the indicators at each level. Based on the membership matrix of the candidate landing point and the weights of the lower-level indicators, the evaluation vector of the upper-level indicators is calculated. Based on the evaluation vectors and weights of the upper-level indicators, the evaluation vectors of the geographical environment parameters and the electrical parameters are calculated. Based on the evaluation vectors of geographical environmental parameters, electrical parameters, and parameter weights, the comprehensive evaluation vector of the candidate landing point is calculated. The comprehensive evaluation vector of the candidate landing point is used as the weight of the evaluation level. The weighted sum of the scores corresponding to each evaluation level is calculated to obtain the comprehensive score of the candidate landing point.

[0062] Specifically, methods such as AHP can be used to determine the weights. For example, the weight of the geographical environment parameter is 0.4, and the weight of the electrical parameter is 0.6, so the parameter weight vector is [0.4 0.6]. For the geographical environment parameter, the weight of the transmission corridor resource is 0.4, the weight of the urban planning compatibility is 0.3, and the weight of the geological conditions is 0.3, so the index weight vector under the geographical environment parameter is [0.4 0.3 0.3]. For the transmission corridor resource, the weight of the existing channel reuse rate is 0.6, and the weight of the new channel length is 0.4. The weights of the bottom-level indicators can form the index weight vector of the upper-level indicators, so the index weight vector under the transmission corridor resource is [0.6 0.4].

[0063] For example, for candidate landing point A, the product of the index weight vector under the power transmission corridor resources and the membership degree matrix of candidate landing point A in the power transmission corridor resources is calculated, i.e. [0.6 0.4]×R, to obtain the evaluation vector L1 of the power transmission corridor resources. Similarly, the evaluation vector L2 of urban planning compatibility and the evaluation vector L3 of geological conditions can be obtained. These three evaluation vectors can form an index evaluation matrix.

[0064] Calculate the product of the index weight vector under the geographical environment parameters and the above index evaluation matrix, i.e. We obtain the evaluation vector P1 for the geographical environment parameters, and similarly, we can obtain the evaluation vector P2 for the electrical parameters. These two evaluation vectors can form a parameter evaluation matrix.

[0065] Next, the product of the parameter weight vector and the aforementioned parameter evaluation matrix is ​​calculated, i.e. The comprehensive evaluation vector of candidate landing point A is obtained. .

[0066] Finally, calculate b1×95+b2×85+b3×70+b4×50 to obtain the comprehensive score of candidate landing point A.

[0067] Specifically, candidate landing points can be determined based on actual circumstances and needs. For example, taking Jiangsu Province as an example, candidate landing points for the sending end are Yancheng (offshore wind power base), Huai'an (onshore photovoltaic base), and Yangzhou (new energy collection station), while candidate landing points for the receiving end are Suzhou (load center), Wuxi (industrial load), and Changzhou (urban load). After calculating the comprehensive score of each candidate landing point using the above method, candidate landing points with a comprehensive score greater than the score threshold are selected. Yancheng and Huai'an are chosen as sending end landing points, and Suzhou and Wuxi as receiving end landing points. The score threshold can be set according to actual circumstances. In actual implementation, preferred landing points and alternative landing points can also be distinguished based on the comprehensive score of the candidate landing points. For example, candidate landing points with a comprehensive score in the first range are considered preferred landing points, and candidate landing points with a comprehensive score in the second range are considered alternative landing points. The comprehensive score corresponding to the first range is greater than that of the second range.

[0068] This embodiment constructs a two-layer index system of "geographical environment parameters - electrical parameters" and adopts the fuzzy comprehensive evaluation method to obtain the comprehensive score of candidate landing points through multi-level fuzzy calculations from bottom to top, thereby screening landing points and ensuring that the landing points take into account both channel resources and power grid security.

[0069] In a specific example, the upper-level score of the candidate landing point is calculated according to the following formula: , of which S i W represents the upper-level score of the i-th candidate landing point, where n represents the total number of lower-level indicators. j Let P be the weight of the j-th underlying indicator. ij Let P be the standardized score of the i-th candidate landing point in the j-th underlying metric, 0≤P ij ≤1.

[0070] The standardized scores of candidate landing points on each underlying indicator can be obtained through expert scoring or linear standardization, and this invention does not impose any restrictions on this.

[0071] (3) Regarding topology optimization Topology types include: chain topology, radial topology, ring topology, and mesh topology.

[0072] Taking the selected sending and receiving points as examples, several candidate topologies are designed as follows: 1) Chain topology: Sending end (Yancheng) → intermediate landing point (Huaian) → receiving end (Suzhou → Wuxi), suitable for centralized transmission of new energy and linear load distribution scenarios. The advantages are simple structure and low investment, but the disadvantage is that a single line fault has a large impact range.

[0073] 2) Radial topology: Sending end (Yancheng, Huai'an) → Receiving end (Suzhou, Wuxi). Each sending end is independently connected to the receiving end. It is suitable for scenarios where new energy is distributed for transmission and loads are distributed for power reception. Its advantage is good fault isolation, but its disadvantage is low channel utilization.

[0074] 3) Ring topology: Sending end (Yancheng-Huaian) → Receiving end (Suzhou-Wuxi) → Sending end, forming a closed loop. It is suitable for scenarios with high anti-disturbance requirements, such as offshore wind power. Its advantage is that the power flow can be adjusted in both directions, but its disadvantage is that the control is complex.

[0075] 4) Mesh topology: The sending and receiving ends are fully interconnected (Yancheng-Suzhou, Yancheng-Wuxi, Huaian-Suzhou, Huaian-Wuxi), which is suitable for multi-source and multi-load mutual assistance scenarios. Its advantage is the highest flexibility, but its disadvantage is the large investment.

[0076] like Figure 4 As shown, a multi-objective optimization function is constructed and solved with transmission capacity, line loss, and stability margin as optimization objectives. The optimal topology is determined from various candidate topologies, including: S31, construct a multi-objective optimization function with the objectives of maximizing transmission capacity, minimizing line loss, and maximizing stability margin.

[0077] In a specific example, the multi-objective optimization function is:

[0078] Where F represents the multi-objective optimization function, α, β, and γ are the weights of transmission capacity, line loss, and stability margin, respectively, and α + β + γ = 1; P is the transmission capacity, ΔP is the line loss, characterized by the line's power loss rate, and K... stab For stability margin, it is characterized by the stability margin of the receiving-end bus voltage, P max For the transmission capacity threshold, ΔP max K is the line loss threshold. stab,max This represents the stability margin threshold. In this example, α, β, and γ are set to 0.4, 0.3, and 0.3, respectively. F ranges between 0 and 1; the closer F is to 1, the better the overall performance of the topology across multiple targets.

[0079] S32, define the constraints, including transmission capacity constraints, line loss constraints, voltage stability constraints, and short-circuit current constraints.

[0080] In a specific example, the constraints of the multi-objective optimization function are as follows: Transmission capacity constraint: P≤2GW (maximum capacity of a single DC line); Power loss constraint: ΔP≤5% (line loss rate); Voltage stability constraint: U≥0.92pu (receiving end bus voltage); Short-circuit current constraint: Isc≤50kA (AC bus short-circuit current).

[0081] S33, Under constraints, solve the multi-objective optimization function to determine the optimal topology from the various candidate topologies, specifically including: A chaotic mapping method is used to initialize the particle swarm, where each particle represents a topological structure and the particle position represents the topological structure parameters. For each particle, based on its current position, the multi-objective value of the particle is determined, and the multi-objective optimization function value is calculated to obtain the current fitness of the particle. The multi-objective value includes the transmission capacity value, the line loss value, and the stability margin threshold. Based on the multi-objective values ​​of all particles, perform non-dominated sorting to determine the non-dominated solution set; Based on adaptive weights and neighborhood search mechanism, the particle velocity and position are updated. If the current fitness of the particle is greater than its historical best fitness, the individual historical best position of the particle is updated to the current position of the particle, and the non-dominated solution set is updated based on the current non-dominated sorting result. Repeat the above process of updating particle velocity and position, individual historical best position, and non-dominated solution set until the preset number of iterations is reached or the distance change of crowding in the non-dominated solution set is less than the preset threshold after N consecutive iterations. Determine the optimal topology based on the current non-dominated solution set.

[0082] In a specific example, when initializing the particle swarm, the population size and number of iterations are set, for example, a population size of 50 and a number of iterations of 100. A Logistic mapping is used to generate the initial particle positions, ensuring a uniform particle distribution. Each particle represents a topology, and the particle position represents topology parameters. The position vector encodes discrete topology parameters (such as line connections, converter station capacity, etc.). Chaotic initialization results in a more uniform particle distribution, covering a wider solution space (such as ring, radial, and chain topologies), avoiding missing the global optimum.

[0083] For example, the initial position of the particle is generated using the following formula:

[0084] Where x0∈(0,1), the initial positions of 50 particles are generated iteratively. For the line connection relationship, if... If the value is 1, then take 1 (indicating connectivity); otherwise, take 0 (indicating disconnection). For converter station capacity: set x... n Mapping to discrete levels (e.g., x) n (∈[0,0.33]→1GW, [0.33,0.66]→1.5GW, [0.66,1]→2GW). Finally, 50 initial topologies (such as ring, radial, and chain) that satisfy the constraints are generated.

[0085] For each particle (topology), calculate the transmission capacity P and line loss Δ using power system simulation software (such as PSASP). P Stability margin K stab Given the target value, generate the target value triple (P, 1-Δ). P K stab These target values ​​are then substituted into the multi-objective optimization objective function to obtain the function value, which is used as the current fitness of the particle.

[0086] Based on the target value triples, all particles are non-dominatedly ordered. It can be understood that for particles A and B, if all targets of A (P, 1-Δ) are... P K stab (None of them are inferior to) And at least one objective is superior to If A dominates B, then A can be used to sort the particles into different dominance layers, with the first layer being the non-dominated solution set. A fast non-dominated sorting method such as NSGA-II can be employed to divide the particles into different dominance layers.

[0087] When updating particle velocities, the inertia weights are adjusted based on the particle's fitness and dispersion to balance the global and local search. This adaptive weighting mechanism is specifically expressed as follows:

[0088] Among them, v i (t+1) represents the updated velocity of particle i, v i (t) represents the current velocity of particle i, ω represents the adaptive weight, c1 represents the individual learning factor, c2 represents the social learning factor (c1=c2=2 can be taken), r1 and r2 are random numbers with values ​​ranging from [0,1], p i Let g be the individual historical best position of particle i. i For the globally optimal historical position, x i (t+1) represents the updated position of particle i, x i (t) represents the current position of particle i, ω0 is the initial weight, which is usually taken as a large value (e.g., 0.9) to ensure global search capability in the early stages of iteration, k is the fitness adjustment coefficient, used to control the degree of influence of fitness on weight, and F i F represents the fitness value of the current particle i. max F represents the maximum fitness value of the current population. min This represents the minimum fitness value of the current population. As a fitness adjustment term, when a particle is in a better position, it moves closer to its individual optimal position, refining the search in the current region; when a particle is in a worse position, it maintains its original velocity and explores new regions of the solution space to avoid missing the global optimum. i The average distance between the current particle and other particles reflects the degree of particle dispersion and is used to adjust the global search capability.max This represents the maximum dispersion distance of the current population (the distance between the most dispersed particles). As a dispersion adjustment term, when particles are more dispersed, the global search capability is enhanced, and particles continue to explore more areas, maintaining population diversity. When particles are more concentrated, the local search capability is enhanced, and particles move closer to the optimal solution, reducing the waste of computing resources in irrelevant areas.

[0089] When updating particle positions, a neighborhood search mechanism is used for discrete topology parameters to ensure position feasibility. For line connection relationships, if the updated position x... i +v i If the value is greater than 0.5, then take 1 (indicating connectivity); otherwise, take 0 (indicating disconnection). For converter station capacity, take the discrete value closest to the updated location, such as the updated location x. i +v i =1.2, then choose the one with the larger fitness (F value) between 1GW and 1.5GW. In this case, the position of the updated particle is still a feasible topology, avoiding the occurrence of infeasible solutions.

[0090] For each particle, the multi-objective optimization function value is calculated based on the particle's current position to obtain the particle's current fitness. If the particle's current fitness is greater than its historical best fitness, the particle's individual historical best position is updated to the particle's current position, and the global historical best position is updated based on the individual historical best positions of all particles. For each particle, if its current position is The value is greater than its individual historical best position p i of If the value is p, then p i Update to the current position (i.e., the particle has found a better topology).

[0091] The global best position is updated based on the individual best positions of all particles, including: selecting the particle with the largest fitness value from the individual best positions of all particles as the new global best position.

[0092] Adjusting particle velocity and position according to the above formula, updating the individual's historical best position and non-dominated solution set, and iterating until the number of iterations reaches a preset number or the change in the crowding distance of the non-dominated solution set is less than a preset threshold after N consecutive iterations (e.g., 3, 5), convergence is achieved. The crowding distance measures the diversity of the non-dominated solution set (the larger the crowding distance, the more dispersed the solution set), specifically expressed as:

[0093] Among them, CD i The distance represents the crowding level of particle i, M represents the number of targets (3 in this example), and f m(·) represents the m-th target value of the particle, f m_max f represents the maximum value of the m-th target. m_min This represents the minimum value of the m-th objective.

[0094] After iterative convergence, the topology corresponding to each particle in the non-dominated solution set is the optimal topology. For example, in the Jiangsu North to South Jiangsu scenario, the optimal topology is a ring (Yancheng-Huaian-Suzhou-Wuxi-Yancheng), with an F-value of 0.92, a transmission capacity of 2.8GW, a line loss rate of 2.8%, and a stability margin of 18°. The non-dominated solution set provides multiple optimal solutions, covering different objective priorities. For example, during peak periods, the solution with the largest transmission capacity can be selected to meet electricity demand; during off-peak periods, the solution with the smallest line loss can be selected to reduce operating costs; and during extreme weather, the solution with the largest stability margin can be selected to ensure system safety.

[0095] Based on the selected landing points and combined with the scenarios of new energy transmission and load receiving, this embodiment designs four topologies and constructs a multi-objective optimization function with the goal of "maximum transmission capacity, minimum loss, and optimal stability". The optimal topology is solved by optimization algorithm to balance capacity, loss and stability, and realize the optimized design of network connection topology, which can dynamically adapt to the fluctuation of new energy output and load distribution characteristics.

[0096] In another embodiment, the present invention further includes: performing topology adaptation based on the scenario type of renewable energy transmission and load reception. Specifically, the topology adaptation rules are as follows: radial topology is preferred for scenarios with distributed renewable energy transmission, ring topology is preferred for scenarios with dense load reception, and mesh topology is preferred for scenarios with multi-source and multi-load interconnection. When the renewable energy output fluctuates by more than 20% or the load changes by more than 15%, topology reconfiguration is triggered, and the grid is kept stable by switching topology branches or adjusting the power allocation of converter stations.

[0097] (4) Regarding multi-scenario verification The optimal topology was simulated and verified in multiple scenarios to obtain the final topology, including: For the aforementioned optimal topology, a power grid simulation model is constructed; The power flow distribution under normal operation, the power flow transfer under N-1 fault, the second-level restart capability under DC instantaneous fault, and the stability capability under new energy fluctuation are verified respectively to obtain the final topology that meets the requirements.

[0098] Specifically, power system analysis program PSASP or power system simulation software BPA can be used to construct a provincial power grid simulation model with embedded DC.

[0099] For example, when constructing the simulation model, the simulation tool used is BPA, the simulation step size is 0.01s, and the power grid scope targeted by the simulation includes: the new energy base in northern Jiangsu, the load center in southern Jiangsu, the 500kV and 220kV AC grid, and the ±200kV embedded DC. The key parameters of the simulation include: the DC converter station adopts the MMC topology, the short-circuit ratio at the sending end is 3.76, the short-circuit ratio at the receiving end is 5.01, and the smoothing reactor is 100mH.

[0100] Simulation verification scenarios include: 1) Normal operation scenario: Verify the power flow distribution under the conditions of large-scale new energy generation (70% wind power output and 60% photovoltaic power output) and peak load (130% base load in summer) to ensure that the line load rate is ≤80%.

[0101] For example, it has been verified that the current power grid simulation model has a uniform power flow distribution under the conditions of high renewable energy generation (70% wind power output and 60% photovoltaic power output) and peak load. The load rate of the Yancheng-Huaian DC line is 75%, and the load rate of the Suzhou-Wuxi DC line is 72%, both ≤80%. The receiving-end bus voltage is 0.95pu, which meets the requirements.

[0102] 2) N-1 fault scenario: Simulate DC line N-1 and AC line N-1 to verify that there is no overload after power flow transfer (load rate ≤100%) and voltage drop ≤10%.

[0103] For example, simulating the Yancheng-Suzhou DC line N-1, the power flow is transferred to the Yancheng-Wuxi and Huai'an-Suzhou lines. After the transfer, the load rate is 95%≤100%, the voltage drop is 8%≤10%, and it recovers in 0.05s.

[0104] 3) DC transient fault scenario: Verify the second-level restart capability when a DC transient fault occurs.

[0105] 4) New energy fluctuation scenario: Wind power output drops from 1.4GW to 0.5GW (fluctuation of 64%), topology dynamic adjustment (switching the power of the Huaian-Wuxi line to 0.8GW), frequency fluctuation ±0.2Hz≤±0.5Hz, voltage fluctuation ±3%≤±5%.

[0106] For example, it has been verified that the scheme meets the requirements of short-circuit current control (Isc≤Isc_limit), voltage fluctuation (ΔU≤5%), and frequency stability (f∈49.5-50.5Hz).

[0107] This embodiment verifies the safety and adaptability of the optimal topology scheme in multiple scenarios by constructing a power grid simulation model. In addition to normal operation scenarios, it also includes extreme scenarios, realizing the adaptability verification in multiple scenarios. The verification is comprehensive, ensuring the safety of the scheme under complex operating conditions and ensuring that the scheme meets engineering requirements.

[0108] The embedded DC network connection topology design scheme provided by this invention achieves optimal adaptation between embedded DC and provincial power grids through multi-dimensional planning comparison, dual-layer landing point quantification screening, topology optimization, and multi-scenario adaptability verification. It meets the needs of provincial power grids for new energy transmission and high load density power supply, improves the new energy absorption capacity of provincial power grids, reduces transmission losses, and provides technical support for scenarios such as inter-regional power transmission and cross-sectional transmission capacity enhancement of provincial power grids.

[0109] like Figure 5 As shown, the present invention also provides an apparatus for implementing the method for determining the connection topology of an embedded DC network as described above, comprising: The first determining module 501 is used to determine the target DC planning method based on multi-dimensional evaluation indicators; The second determining module 502 is used to determine the target landing point based on geographical environment parameters and electrical parameters; The topology optimization module 503 is used to design multiple candidate topology structures according to the target DC planning method and the target landing point, and to determine the optimal topology among the multiple candidate topology structures with transmission capacity, line loss and stability margin as optimization objectives. The simulation verification module 504 is used to perform simulation verification on the optimal topology in multiple scenarios to obtain the final topology and complete the determination of the embedded DC network connection topology.

[0110] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and modifications.

Claims

1. A method of determining an embedded DC network connection topology, characterized by, The method comprises the following steps: determining a target DC planning mode based on multi-dimensional evaluation indexes; determining a target landing point based on geographical environment parameters and electrical parameters; designing multiple candidate topological structures according to multiple topological types respectively based on the target DC planning mode and the target landing point, constructing a multi-objective optimization function with transmission capacity, line loss and stability margin as optimization objectives, and solving to determine an optimal topological structure from the multiple candidate topological structures; simulating and verifying the optimal topological structure under multiple scenarios to obtain a final topological structure and determine the embedded DC network connection topological structure.

2. The method of claim 1, wherein, The geographical environment parameters include transmission corridor resources, urban planning compatibility and geological conditions; the transmission corridor resources include existing channel multiplexing rate and newly added channel length; the urban planning compatibility includes distance from the urban built-up area and overlap degree with industrial parks; the geological conditions include foundation bearing capacity and seismic intensity; The electrical parameters include short-circuit current level, power flow distribution uniformity and voltage stability margin; the short-circuit current level includes short-circuit current value and short-circuit ratio during fault; the power flow distribution uniformity includes line load rate standard deviation and N-1 power flow transfer rate; the voltage stability margin includes static voltage stability margin and transient voltage recovery time.

3. The method of claim 2, wherein, The method of determining a target landing point based on geographical environment parameters and electrical parameters comprises the following steps: determining a candidate landing point; for each candidate landing point, constructing a membership matrix of the candidate landing point according to the index values of the geographical environment parameters and the index values of the electrical parameters corresponding to the candidate landing point; performing multi-level fuzzy operation based on the membership matrix of the candidate landing point to obtain a comprehensive score of the candidate landing point; taking the candidate landing point with a comprehensive score greater than a score threshold as the target landing point.

4. The method of claim 2, wherein, The method of constructing a membership matrix of the candidate landing point according to the index values of the geographical environment parameters and the index values of the electrical parameters comprises the following steps: obtaining the numerical values of each bottom-level index of the candidate landing point; determining the membership degrees of the candidate landing point in each evaluation level of each bottom-level index according to the numerical values to construct the membership matrix of the candidate landing point.

5. The method of claim 2, wherein, The method of performing multi-level fuzzy operation based on the membership matrix of the candidate landing point to obtain a comprehensive score of the candidate landing point comprises the following steps: determining parameter weights and bottom-level index weights according to the geographical environment parameters, the electrical parameters and each layer index contained therein; calculating an evaluation vector of the upper-level index according to the membership matrix of the candidate landing point and the bottom-level index weights; calculating an evaluation vector of the geographical environment parameters and an evaluation vector of the electrical parameters according to the evaluation vector of the upper-level index and the upper-level index weights; calculating a comprehensive evaluation vector of the candidate landing point according to the evaluation vector of the geographical environment parameters, the evaluation vector of the electrical parameters and the parameter weights; taking the comprehensive evaluation vector of the candidate landing point as the weight of the evaluation level, calculating the weighted sum of the scores corresponding to each evaluation level to obtain the comprehensive score of the candidate landing point.

6. The method of claim 1, wherein, The method of determining a target DC planning mode based on multi-dimensional evaluation indexes comprises the following steps: determining the weights of each dimension by analytic hierarchy process; respectively obtaining the standardized scores of each DC planning mode in each dimension, wherein the DC planning modes include traditional AC reinforcement, two-end DC and multi-end embedded DC; According to the weight of each dimension and the normalized score of the DC planning mode in each dimension, a weighted sum is performed to obtain a comprehensive score of the DC planning mode; The DC planning mode with the highest comprehensive score is taken as the target DC planning mode; The multi-dimensional evaluation index includes a technology dimension, an economic dimension, an environmental dimension, and a safety dimension; the technology dimension includes a transmission capacity improvement rate, a new energy consumption capacity, a fault ride-through success rate, and a voltage support strength; the economic dimension includes a unit capacity investment cost, an annual operation loss cost, and a full life cycle benefit; the environmental dimension includes a transmission corridor occupation area, a carbon emission reduction amount, and a noise influence range; and the safety dimension includes a short-circuit current control effect, an N-1 fault overload rate, and a transient stability margin.

7. The method of claim 1, wherein, The topology types include a chain topology, a radial topology, a ring topology, and a mesh topology. A multi-objective optimization function is constructed with the transmission capacity, the line loss, and the stability margin as the optimization objectives, and the optimal topology is determined from the multiple candidate topology structures by solving the multi-objective optimization function, including: A multi-objective optimization function is constructed with the maximum transmission capacity, the minimum line loss, and the highest stability margin as the objectives, and the optimal topology is determined from the multiple candidate topology structures by solving the multi-objective optimization function, including: ； Wherein, F represents a multi-objective optimization function, α, β, γ are weights of transmission capacity, line loss, and stability margin respectively, P is transmission capacity, △P is line loss, K stab is stability margin, P max is transmission capacity threshold, △P max is line loss threshold, K stab,max is stability margin threshold; The constraint conditions include a transmission capacity constraint, a line loss constraint, a voltage stability constraint, and a short-circuit current constraint. The optimal topology is determined from the multiple candidate topology structures by solving the multi-objective optimization function under the constraint conditions.

8. The method of claim 7, wherein, The optimal topology is determined from the multiple candidate topology structures by solving the multi-objective optimization function under the constraint conditions, including: The particle swarm is initialized by using a chaotic mapping method, and each particle represents a topology structure, and the particle position represents the topology structure parameters. For each particle, the multi-objective value of the particle is determined according to the current position of the particle, and the multi-objective optimization function value is calculated to obtain the current fitness of the particle, wherein the multi-objective value includes a transmission capacity value, a line loss value, and a stability margin value. The non-dominated solution set is determined by non-dominated sorting based on the multi-objective values of all particles. The particle speed and position are updated based on the adaptive weight and the neighborhood search mechanism, and if the current fitness of the particle is greater than its historical best fitness, the individual historical optimal position of the particle is updated to the current position of the particle, and the non-dominated solution set is updated based on the current non-dominated sorting result. The updating process of the particle speed and position, the individual historical optimal position, and the non-dominated solution set is repeated until the iteration reaches a preset number of times or the non-dominated solution set congestion distance changes less than a preset threshold for consecutive N times of iteration. The optimal topology is determined according to the current non-dominated solution set.

9. The method of claim 1, wherein, The optimal topology is simulated and verified under multiple scenarios to obtain a final topology, including: A power grid simulation model is constructed for the optimal topology; The final topology is determined by verifying the power flow distribution under normal operation scenarios, the power flow transfer under N-1 fault scenarios, the second-level restart capability under DC transient fault scenarios, and the stability capability under new energy fluctuation scenarios.

10. An apparatus for implementing the method of determining an embedded DC network connection topology according to any one of claims 1 to 9, characterized in that including: The first determination module is configured to determine the target DC planning mode based on the multi-dimensional evaluation index. A second determining module is configured to determine a target landing point based on the geographical environment parameter and the electrical parameter; A topology optimization module is configured to design a plurality of candidate topology structures according to a plurality of topology types based on the target DC planning mode and the target landing point, to construct a multi-objective optimization function with transmission capacity, line loss and stability margin as optimization objectives, and to solve the multi-objective optimization function to determine an optimal topology from the plurality of candidate topology structures; A simulation verification module is configured to simulate and verify the optimal topology under a plurality of scenarios to obtain a final topology and complete determination of the embedded DC network connection topology.

Images