Multi-snowflake network interconnection planning method based on diffusion model and multi-port SOP

Abstract

The invention relates to a multi-snowflake network interconnection planning method based on a diffusion model and a multi-port SOP. The multi-snowflake network interconnection planning method comprises the steps of obtaining a topological structure, an operation state, photovoltaic power generation data and wind power generation data of each snowflake network unit; performing graph theory modeling on the multi-snowflake network topological structure, and identifying key interconnection nodes as SOP candidate positions; performing normalization processing on the operation state, photovoltaic power generation and wind power generation data, and inputting the data into a diffusion model to obtain a node load typical scene, a photovoltaic power generation typical scene and a wind power generation typical scene of each snowflake network unit; constructing a multi-snowflake network interconnection planning model based on the diffusion model and the multi-port SOP; inputting each typical scene of each snowflake network unit into a multi-snowflake network interconnection planning model, and solving the model to obtain a Pareto solution system of the model; and solving an equilibrium solution in the Pareto solution system, and outputting a multi-snowflake network interconnection planning result. The method can deal with complex and changeable operation scenes and consider the comprehensive cost of the power grid and the power supply reliability of the system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optimal configuration of multi-port intelligent soft switches in power distribution networks, and specifically relates to a multi-snowflake network interconnection planning method based on a diffusion model and multi-port SOP. BACKGROUND

[0002] With the rapid development of distributed energy and new power systems, the complexity of power grid operation and the demand for regulation have significantly increased. The widespread access of distributed energy, such as solar and wind energy, has improved the diversity and sustainability of energy supply, but has also exacerbated power and voltage fluctuations in the power grid, posing higher requirements for the operational safety and economy of the power grid. Traditional power grid regulation methods, such as transformer tap adjustment and reactive power compensation device adjustment, have limited regulation capacity and insufficient precision, making it difficult to meet the fast and accurate optimization operation requirements of modern power grids.

[0003] In recent years, "snowflake network" has gradually received widespread attention and application as an innovative power grid structure due to its flexible load transfer, high efficiency and effectiveness, and adaptability to more new elements of energy internet. The snowflake network consists of independent feeder clusters formed by the feeders of multiple substations, and realizes open-loop operation through ring network node devices such as ring network boxes, box-type transformers, and distribution rooms. Its unique network topology significantly improves the power supply reliability and self-healing capability of the power grid. However, how to realize the coordinated operation and comprehensive utilization of resources under such a complex network structure has become a key problem to be solved.

[0004] Current research mainly optimizes the configuration of snowflake networks through two-port SOPs in a single scenario. However, the planning results generated in a single scenario are difficult to cope with complex and variable actual operation scenarios. In addition, although two-port SOPs can realize the interconnection of different feeders within a single snowflake network unit, their regulation range is often limited to the two connected feeders. When multiple interconnected snowflake network units need to be optimized, using two-port SOPs cannot fully exploit the regulation potential of multi-snowflake network interconnection.

[0005] Chinese patent CN117728406B discloses a snowflake network structure power grid intelligent soft switch (SOP) and energy storage combined optimization configuration method based on voltage sensitivity, which realizes resource optimization within a single snowflake network unit by constructing a multi-objective model of comprehensive cost and system load deviation. However, this scheme has the following limitations:

[0006] 1. Limitations of single scenario assumption: The existing method only considers a certain typical daily data of node load in a single snowflake network unit when planning the snowflake network, and the planning results are difficult to cope with complex and variable actual operation scenarios.

[0007] 2. Limitations of single snowflake network unit: Existing methods only address SOP and energy storage configuration within a single snowflake network structure, without addressing the collaborative optimization problem in the multi-snowflake network interconnection scenario. In practical applications of multi-snowflake network interconnection, the power interaction and load balancing requirements between different units are more complex, and traditional two-port SOP or single unit optimization models cannot meet the global regulation requirements.

[0008] 3. Device function limitations: The regulation range of the two-port SOP in existing technologies is limited to adjacent feeders, and cannot achieve cross-regional power flexible scheduling between multi-snowflake network units, resulting in limited resource utilization, and the expansion capability of multi-port SOP is not fully utilized.

[0009] 4. Insufficient model complexity: Existing optimization models do not consider the topological coupling constraints and cross-unit power flow coordination problems brought by multi-snowflake network interconnection, making it difficult to balance the safety and economy of the interconnected system. SUMMARY

[0010] The purpose of the present application is to provide a multi-snowflake network interconnection planning method based on diffusion model and multi-port SOP, which can cope with complex and variable operating scenarios and take into account the overall cost of the power grid and the system power supply reliability.

[0011] To achieve the above purpose, the technical scheme adopted by the present application is: a multi-snowflake network interconnection planning method based on diffusion model and multi-port SOP, comprising the following steps:

[0012] S1, obtaining the topological structure data, operating state data, node load data, photovoltaic power generation data and wind power generation data of each snowflake network unit;

[0013] S2, graph theory modeling is performed on the topological structure data of the multi-snowflake network, and key interconnection nodes are identified as SOP candidate positions;

[0014] S3, the node load data, photovoltaic power generation data and wind power generation data are normalized respectively, and then input into the diffusion model to obtain the node load typical scenario, photovoltaic power generation typical scenario and wind power generation typical scenario of each snowflake network unit;

[0015] S4, a multi-snowflake network interconnection planning model based on diffusion model and multi-port SOP is constructed, the optimization objective of the multi-snowflake network interconnection planning model is to minimize the linear weighted combination of the annual comprehensive cost of multi-snowflake network interconnection devices and the system power supply reliability, and the constraint conditions consist of multi-snowflake network interconnection constraints, snowflake unit power flow constraints, snowflake unit operation constraints and SOP operation constraints;

[0016] S5, input the node load typical scene, the photovoltaic power generation typical scene and the wind power generation typical scene of each Xueliang unit obtained in step S3 into the multi-Xueliang interconnected planning model constructed in step S4, solve the model, and obtain a Pareto solution system of the model;

[0017] S6, sort the Pareto solution system based on the entropy weight method, select an equilibrium solution in the Pareto solution system, and output the multi-Xueliang interconnected planning result based on the diffusion model.

[0018] Further, in step S1, the topology structure data of the Xueliang unit includes the type, position, number basic information of the power supply point, transformer, switch, line device of the Xueliang unit and the connection relationship between each device; and the operation state data of the Xueliang unit includes the load data of each node in the Xueliang unit and the real-time operation state of each device.

[0019] Further, the implementation method of step S2 is:

[0020] Modeling the topology structure data of the multi-Xueliang by graph theory, and identifying the key interconnected nodes as SOP candidate positions

[0021] Map the physical structure of the Xueliang to a weighted directed graph G=(V, E, W), wherein V represents the set of all electrical nodes in the Xueliang, E represents the line connection relationship, and W represents the electrical details of the edge, including line impedance, capacity limit and node load power;

[0022] Combined with the electrical characteristics and the topology structure, define a comprehensive node importance index S i , and the specific expression is as follows:

[0023]

[0024] , wherein C B (i) represents the betweenness centrality of node i, which is used to measure the hub nature of the node in the network; P i flow represents the normalized power flow density of node i, which reflects its key role in power transmission; ΔV i represents the difference between the voltage of node i and the rated voltage, which is used to evaluate the voltage stability demand; V base represents the rated voltage; ω1, ω2 and ω3 are weight coefficients;

[0025] Select the top N nodes as candidate SOP access positions. i

[0026] Further, in step S3, the diffusion model mainly consists of diffusion and reverse processes; in the diffusion process, assume that the original data sample is z0, and the sample generated after the nth diffusion is z​n , assuming the prior distribution is a standard Gaussian distribution From z n-1 to z n The diffusion process is expressed as:

[0027]

[0028] In the formula, is the scaling factor of the nth diffusion, z n The probability distribution of z

[0029]

[0030] Let Then the relationship between z n and z0is as follows:

[0031]

[0032] The probability distribution of z n is transformed into:

[0033]

[0034] In the reverse process, the diffusion model fits the conditional probability distribution p θ (z n-1 |z n ) through a neural network, which is expressed as:

[0035]

[0036] In the formula, p θ (z n-1 |z n ) is a Gaussian distribution obtained through learnable parameters θ, which is expressed as:

[0037]

[0038] Further, in step S4, the optimization objective of the multi-snowflake network interconnection planning model is to minimize the value of the objective function f, and the expression of the optimization objective is:

[0039] min f = αC com + βξ reliability (8)

[0040] Wherein,

[0041]

[0042] Wherein, C com is the annual comprehensive cost of multi-snowflake network interconnection equipment, and ξ reliabilityFor system power supply reliability, α and β are weight coefficients, and α + β = 1; N snow is the total number of snowflake network units; is the SOP investment cost of the i th snowflake network unit; is the SOP operation and maintenance cost of the i th snowflake network unit; is the power supply loss cost of the i th snowflake network unit; is the power purchase cost of the i th snowflake network unit from the upper-level power grid; is the interaction power cost of the i th snowflake network unit with other units;

[0043] The expression of is:

[0044]

[0045] In the formula, N SOP,i is the number of SOPs to be installed in the i th snowflake network unit; c i,k,sop is the unit capacity investment cost of the k th SOP of the i th snowflake network unit; S i,k,SOP is the capacity of the k th SOP of the i th snowflake network unit;

[0046] The expression of is:

[0047]

[0048] In the formula, η i is the operation and maintenance cost coefficient of the i th snowflake network unit;

[0049] The expression of is:

[0050]

[0051] In the formula, c i is the electricity price of the i th snowflake network unit; N i,t is the total number of time periods of the i th snowflake network unit; Ω b,i is the set of all branches in the i th snowflake network unit; r ab is the resistance value of branch ab; I ab,t is the current amplitude of branch ab in the t th time period; N nopower,i is the total number of non-power supply nodes in the i th snowflake network unit; is the active power loss of the SOP at the j th node of the i th snowflake network unit in the t th time period;

[0052] The expression of is:

[0053]

[0054] where λ i,t is the electricity purchase price from the upper-level power grid, is the electricity purchase power from the upper-level power grid at time t, and Δt represents the time step;

[0055] The expression of ξ

[0056]

[0057] where C i,ex is the unit electricity cost of interaction between units, is the interaction power between units;

[0058] ξ reliability The lower the value of ξ

[0059] ξ reliability = β1ξ node + β2ξ line (15)

[0060] wherein,

[0061]

[0062] wherein, ξ node represents the node reliability of the multi-snowflake interconnected system; ξ line represents the line reliability of the multi-snowflake interconnected system; β1 and β2 are weight coefficients; N snow represents the total number of snowflake units; m k represents the number of nodes in the kth snowflake unit; represents the load of node i in the kth snowflake unit at time t; represents the transformer capacity of node i in the kth snowflake unit; represents the transformer expansion capacity of node i in the kth snowflake unit; l k represents the number of lines in the kth snowflake unit; represents the load of line i in the kth snowflake unit at time t; represents the capacity of line i in the kth snowflake unit; represents the expansion capacity of line i in the kth snowflake unit.

[0063] Further, in step S4, the expression of the multi-snowflake interconnected constraint is:

[0064]

[0065] wherein, m represents the grid load of the kth snowflake unit; k n represents the number of nodes in the kth snowflake unit; P represents the regular load power of the jth snowflake unit; P represents the flexible load power of the jth snowflake unit; P represents the energy storage charging and discharging power of the jth snowflake unit; P represents the distributed power output power of the jth snowflake unit; P represents the interaction power of the kth snowflake unit with other units; Pmax represents the maximum value of the interaction power between snowflake units.

[0066] Further, in step S4, the expression of the snowflake unit power flow constraint is:

[0067]

[0068] wherein r ij,k and x ij,k are the resistance and reactance of branch ij in the kth snowflake network unit; P ij,k is the active power of node i flowing to node j in the kth snowflake network unit, Q ij,k is the reactive power of node i flowing to node j in the kth snowflake network unit; P i,k is the sum of the active power injected at node i in the kth snowflake network unit, are the active power injected by SOP and consumed by load at node i in the kth snowflake network unit, respectively; Q i,k is the sum of the reactive power injected at node i in the kth snowflake network unit, are the reactive power injected by SOP and consumed by load at node i in the kth snowflake network unit, respectively.

[0069] Further, in step S4, the expression of the snowflake unit operation constraint is:

[0070]

[0071] wherein U min,k and U max,k are the minimum allowed node voltage value and the maximum allowed node voltage value of the kth snowflake network unit, respectively; I max,k is the maximum allowed branch current value of the kth snowflake network unit.

[0072] Further, in step S4, the SOP operation constraint is:

[0073]

[0074] wherein, and are the active power and the reactive power of the SOP connected to the node i in the kth snowflake network unit, respectively; is the loss of the SOP connected to the node i in the kth snowflake network unit; is the loss coefficient of the SOP connected to the node i in the kth snowflake network unit; is the capacity of the SOP connected to the node i.

[0075] Further, the multi-snowflake network interconnection planning result output in step S6 is compared with a planning result not based on the diffusion model, so as to realize the evaluation of the planning effect.

[0076] Compared with the prior art, the present application has the following beneficial effects:

[0077] 1. The diffusion model is used to realize the generation of typical scenes such as node load, photovoltaic power generation and wind power generation, solves the limitation that the traditional planning method only considers a single scene, and greatly improves the applicability of the planning result in dealing with complex and variable operation scenes.

[0078] 2. Through the accurate power flow control of the multi-port SOP, the method can realize the flexible scheduling of power flow between multiple snowflake network units, effectively reduce the line loss, and improve the energy utilization efficiency. At the same time, the operation cost, system power supply reliability and other multi-objectives are considered for optimization, which helps to reduce the overall operation cost of the power grid and improve the economic benefits.

[0079] 3. The "snowflake network" structure itself has high power supply reliability and self-healing ability, and the introduction of the present method will further strengthen this advantage. By optimizing the operation strategy of the multi-port SOP, the power flow path can be quickly adjusted when a fault occurs, ensuring the continuous power supply of key loads and reducing the power outage time and range. BRIEF DESCRIPTION OF DRAWINGS

[0080] Figure 1 is a schematic diagram of the "snowflake network" structure power grid in the embodiment of the present application;

[0081] Figure 2 is a schematic diagram of the typical scene of node load in the embodiment of the present application;

[0082] Figure 3 is a schematic diagram of the typical scene of photovoltaic power generation in the embodiment of the present application;

[0083] Figure 4 is a schematic diagram of the typical scene of wind power generation in the embodiment of the present application;

[0084] Figure 5 is the solution result of the Pareto solution system of the model in the embodiment of the present application;

[0085] Figure 6is the full-day voltage change curve diagram of each node of the snowflake unit 1 after accessing the multi-port SOP in the embodiment of the present application;

[0086] Figure 7 is the full-day voltage change curve diagram of each node of the snowflake unit 2 after accessing the multi-port SOP in the embodiment of the present application;

[0087] Figure 8 is the active power output curve diagram of the multi-port SOP in the embodiment of the present application;

[0088] Figure 9 is the reactive power output curve diagram of the multi-port SOP in the embodiment of the present application;

[0089] Figure 10 is the flow chart of the multi-snowflake network interconnection planning method based on the diffusion model and the multi-port SOP provided by the embodiment of the present application. DETAILED DESCRIPTION

[0090] The multi-snowflake network interconnection planning method based on the diffusion model and the multi-port SOP provided by the present application will be further described in detail below in combination with the accompanying drawings and specific embodiments. The advantages and features of the present application will be more apparent according to the following description. It should be noted that the accompanying drawings are very simplified and all use non-precise proportions, only for the purpose of facilitating and clarifying the description of the embodiments of the present application, and are not used to limit the scope of the embodiments of the present application, and therefore do not have substantial technical significance. Any modification of the structure, change of the proportional relationship or adjustment of the size, without affecting the effects and purposes that can be achieved by the present application, should still fall within the scope of the technical content disclosed by the present application.

[0091] It should be noted that in the present application, the relationship terms such as and the like are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations. Moreover, the term "comprises", "includes" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes the elements listed explicitly, but also includes other elements not listed explicitly or inherent to such process, method, article or equipment.

[0092] As shown in Figure 10 , the present application provides a multi-snowflake network interconnection planning method based on the diffusion model and the multi-port SOP, which is used for optimizing the configuration of the multi-port intelligent soft switch connecting each snowflake network unit; comprising the steps of:

[0093] S1. Obtain the topology data, operating status data, node load data, photovoltaic power generation data, wind power generation data, etc. of each snowflake network unit; wherein, the topology data of each snowflake network unit includes the basic information such as the type, location, and number of the power supply point, transformer, switch, line and other equipment of the snowflake network unit, as well as the connection relationship between each equipment; the operating status data includes at least the load data of each node in the snowflake network unit and the real-time operating status of each equipment.

[0094] S2. Perform graph theory modeling on the multi-snowflake network topology data to identify key interconnected nodes as candidate SOP locations.

[0095] The physical structure of the snowflake network is mapped to a weighted directed graph G = (V, E, W), where V contains all electrical nodes in the snowflake network, E represents the line connection relationship, and W represents the electrical details of the edges, including line impedance, capacity limits, and node load power.

[0096] Combining electrical characteristics and topology, a comprehensive node importance index S is defined. i The specific expression is as follows:

[0097]

[0098] Among them, C B (i) denotes the betweenness centrality of node i, used to measure the pivotal nature of a node in the network; P i flow ΔV represents the normalized power current density of node i, reflecting its criticality in power transmission; i V represents the voltage deviation at node i (the difference from the rated voltage), used to assess voltage stability requirements; base This represents the rated voltage; ω1, ω2, and ω3 are all weighting coefficients.

[0099] Select S i The top N nodes are selected as candidate SOP access locations, where N is set according to the size of the snowflake network and is usually an inter-unit communication node or a high-load node.

[0100] S3. Normalize the node load data, photovoltaic power generation data, and wind power generation data respectively, and then input them into the diffusion model to obtain the typical node load scenarios, photovoltaic power generation scenarios, and wind power generation scenarios for each snowflake grid unit.

[0101] The diffusion model mainly consists of two processes: diffusion and reverse diffusion. In the diffusion process, assuming the original data sample is z0, the sample generated after the nth diffusion is z... n Assume the prior distribution is a standard Gaussian distribution. From z n-1 To z nThe diffusion process can be expressed as:

[0102]

[0103] wherein, is a scaling factor, z n The probability distribution of z

[0104]

[0105] Let Then z n The relationship between z

[0106]

[0107] z n The probability distribution of z

[0108]

[0109] In the reverse process, the diffusion model fits the conditional probability distribution p θ (z n-1 |z n ) through the neural network, the expression of which is as follows:

[0110]

[0111] wherein, p θ (z n-1 |z n ) is a Gaussian distribution obtained through learnable parameters θ, the expression of which is as follows:

[0112]

[0113] S4, a multi-snowflake network interconnection planning model based on a diffusion model and a multi-port SOP is constructed. The model includes two parts of an objective function and constraint conditions. The objective function is divided into two parts of objective function 1 and objective function 2, the objective function 1 is the annual comprehensive cost of multi-snowflake network interconnection, and the objective function 2 is the system power supply reliability; the constraint conditions are composed of multi-snowflake network interconnection constraints, snowflake unit power flow constraints, snowflake unit operation constraints, and SOP operation constraints.

[0114] The expression of the objective function is as follows:

[0115] minf=αC com +βξ reliability (40)

[0116] wherein,

[0117]

[0118] In formulas (40)-(41), N snow is the total number of snowflake network units; C com is the annual investment and operation cost of the multi-snowflake network interconnected device; is the SOP investment cost of the ith snowflake network unit; is the SOP operation and maintenance cost of the ith snowflake network unit; is the power supply loss cost of the ith snowflake network unit; is the purchase cost of the ith snowflake network unit from the upper-level power grid; is the interaction power cost of the ith snowflake network unit with other units; ξ deviation,i is the feeder load deviation of the ith snowflake network unit; α and β are weight coefficients, and α+β=1.

[0119] The expression of is:

[0120]

[0121] In formula (42), N SOP,i is the number of SOPs to be installed in the ith snowflake network unit; c i,k,sop is the unit capacity investment cost of the kth SOP of the ith snowflake network unit; S i,k,SOP is the capacity of the kth SOP of the ith snowflake network unit.

[0122] The expression of is:

[0123]

[0124] In formula (43), η i is the operation and maintenance cost coefficient of the ith snowflake network unit.

[0125] The expression of is:

[0126]

[0127] In formula (44), c i is the electricity price of the ith snowflake network unit; N i,t is the total number of time periods of the ith snowflake network unit; Ω b,i is the set of all branches in the ith snowflake network unit; r ab is the resistance value of branch ab; I ab,t is the current amplitude of branch ab in the tth time period; N nopower,i is the total number of non-power supply nodes in the ith snowflake network unit; The active power loss of the SOP at the jth node of the ith snowflake unit at the tth time period.

[0128] The expression of is:

[0129]

[0130] where λ i,t is the electricity purchase price from the upper-level power grid, is the electricity purchase power from the upper-level power grid at time t, and Δt represents the time step, which is 1 h.

[0131] The expression of is:

[0132]

[0133] where C i,ex is the unit electricity exchange cost between units, is the exchange power between units.

[0134] The lower the value of ξ reliability is, the better the system power supply reliability is, and the expression of is:

[0135] The lower the value of ξ reliability is, the better the system power supply reliability is, and the expression of is: node + β2ξ line (47)

[0136] where,

[0137]

[0138] In formulas (47)-(49), ξ node represents the node reliability of the multi-snowflake grid interconnection system; ξ line represents the line reliability of the multi-snowflake grid interconnection system; β1 and β2 are weight coefficients; N snow represents the number of snowflake units; m k represents the number of nodes in the kth snowflake unit; represents the load of node i in the kth snowflake unit at time t; ξ k T,i0represents the transformer capacity of node i in the kth snowflake unit; represents the transformer expansion capacity of node i in the kth snowflake unit; l k represents the number of lines in the kth snowflake unit; represents the load of line i in the kth snowflake unit at time t; represents the capacity of line i in the kth snowflake unit; represents the expansion capacity of line i in the kth snowflake unit.

[0139] Each snowflake network unit is independent of each other, and can be interconnected through SOP. Each snowflake network unit realizes the real-time self-balancing of source-load-storage power within the unit, and the insufficient part relies on the power supply of the upper-level power grid. The expression of the multi-snowflake network interconnection constraint is:

[0140]

[0141] In the formula, m k represents the number of nodes in the kth snowflake unit; represents the network load of the kth snowflake unit; represents the interaction power of the kth snowflake unit with other units; represents the conventional load power of the jth snowflake unit; represents the distributed power output power of the jth snowflake unit; represents the energy storage charging and discharging power of the jth snowflake unit; represents the flexible load power of the jth snowflake unit; is the maximum value of the interaction power between snowflake units.

[0142] The expression of the snowflake unit power flow constraint is:

[0143]

[0144] In formulas (52)-(57), r ij,k and x ij,k are the resistance and reactance of branch ij in the kth snowflake network unit; P ij,k is the active power of node i flowing to node j in the kth snowflake network unit, Q ij,k is the reactive power of node i flowing to node j in the kth snowflake network unit; P i,k is the sum of active power injected at node i in the kth snowflake network unit, are the active power injected by SOP and consumed by load at node i in the kth snowflake network unit, respectively; Q i,k is the sum of reactive power injected at node i in the kth snowflake network unit, are the reactive power injected by SOP and consumed by load at node i in the kth snowflake network unit, respectively.

[0145] The expression of the snowflake unit operation constraint is:

[0146]

[0147] In formulas (58)-(59), U min,k and U max,k are the minimum and maximum allowed node voltage values of the kth snowflake network unit, respectively; Imax,k is the maximum allowed branch current value of the kth snowflake network unit.

[0148] The SOP operation constraints are:

[0149]

[0150] in formulas (60)-(64), and are the active power and reactive power of the SOP transmission connected to node i in the kth snowflake network unit, respectively; is the loss of the SOP connected to node i in the kth snowflake network unit; is the loss coefficient of the SOP connected to node i in the kth snowflake network unit; is the capacity of the SOP connected to node i.

[0151] S5, input the node load typical scenario, photovoltaic power generation typical scenario and wind power generation typical scenario of each snowflake network unit in step S3 into the multi-snowflake network interconnection planning model constructed in step S4, solve the model, and obtain the Pareto solution set of the model.

[0152] S6, sort the Pareto solution set based on the entropy weight method, select the equilibrium solution in the Pareto solution set, and output the multi-snowflake network interconnection planning result based on the diffusion model.

[0153] S7, compare the multi-snowflake network interconnection planning result output in step S6 with the planning result not based on the diffusion model, so as to realize the evaluation of the planning effect.

[0154] The present application provides a specific embodiment, including the following steps:

[0155] Step A, obtain the topological structure data, operation state data, photovoltaic power generation data, wind power generation data, etc. of each snowflake network unit.

[0156] In this embodiment, the topological structure data, operation state data, photovoltaic power generation data, and wind power generation data of each snowflake network unit come from a city in northern China. The schematic diagram of the "snowflake network" structure power grid is shown in Figure 1 . The length, resistance, and reactance parameters of each line are shown in Table 1.

[0157] Table 1 Line parameters of "snowflake network" structure power grid

[0158]

[0159]

[0160] Step B, perform graph theory modeling on the multi-snowflake network topological data, and identify key interconnection nodes as SOP candidate positions.

[0161] In the embodiment, the SOP candidate positions are the 9th, 10th and 25th nodes of the snowflake network unit 1, and the 24th, 38th and 39th nodes of the snowflake network unit 2.

[0162] Step C: The node load data, photovoltaic power generation data and wind power generation data in step A are normalized respectively, and then input into the diffusion model to obtain the node load typical scenario, photovoltaic power generation typical scenario and wind power generation typical scenario of each snowflake network unit.

[0163] In the embodiment, the node load typical scenario (taking the node 6 in the snowflake unit 1 as an example), photovoltaic power generation typical scenario and wind power generation typical scenario generated by the diffusion model are as shown in Figure 2 、 Figure 3 、 Figure 4 .

[0164] Step D: A multi-snowflake network interconnection planning model based on the diffusion model and the multi-port SOP is constructed. The model includes two parts of an objective function and a constraint condition. The objective function includes two parts of an objective function 1 and an objective function 2, the objective function 1 is the annual comprehensive cost of the multi-snowflake network interconnection, and the objective function 2 is the system power supply reliability; the constraint condition is composed of a multi-snowflake network interconnection constraint, a snowflake unit power flow constraint, a snowflake unit operation constraint and an SOP operation constraint.

[0165] In the embodiment, the system operation parameters are as follows: the maximum allowed node voltage of the system is 1.05 p.u., the minimum allowed node voltage of the system is 0.95 p.u., and the maximum allowed branch current is 0.4 kA. The SOP related parameters are as follows: the economic service life of the SOP is 20 years, the unit capacity investment cost of the SOP is 1000 yuan / kVA, the SOP operation and maintenance cost coefficient is 0.01, the loss coefficient of the SOP is 0.02, and the discount rate of the SOP is 0.08.

[0166] Step E: The node load typical scenario, photovoltaic power generation typical scenario and wind power generation typical scenario of each snowflake network unit in step C are input into the multi-snowflake network interconnection planning model constructed in step D, the model is solved, and the Pareto solution set of the model is obtained.

[0167] In the embodiment, the Pareto solution set of the model is as shown in Figure 5 .

[0168] Step F: The Pareto solution set is sorted based on the entropy weight method, the equilibrium solution in the Pareto solution set is selected, and the multi-snowflake network interconnection planning result based on the diffusion model is output.

[0169] In the embodiment, after the multi-port SOP is accessed, the all-day voltage variation curve of each node of the snowflake unit 1 is as shown in Figure 6The full-day voltage variation curves of each node of the snowflake unit 2 are shown in FIG. 2B. Figure 7 The active power output curve of the multi-port SOP is shown in FIG. 3B. Figure 8 The reactive power output curve of the multi-port SOP is shown in FIG. 3C. Figure 9 The access capacity of the multi-port SOP is 2.8 MVA, the annual investment cost of the multi-port SOP is 142,600 yuan, the annual operation and maintenance cost of the multi-port SOP is 14,000 yuan, and the annual power supply loss cost of the distribution network is 200,800 yuan. The annual comprehensive cost of the multi-port SOP is 35,740 yuan. The system power supply reliability is 33.71.

[0170] Step G, compare the multi-snowflake network interconnection planning result output by step F with the planning result not based on the diffusion model to realize the evaluation of the planning effect.

[0171] In this embodiment, the comparison between the multi-snowflake network interconnection planning result based on the diffusion model (result 1) and the planning result not based on the diffusion model (result 2) is shown in the following table.

[0172] Table 2 Typical solution set of optimization configuration result

[0173]

[0174] As shown in the above table, the multi-snowflake network interconnection planning result based on the diffusion model considers various typical scenarios that the system may encounter during operation, and realizes a high system power supply reliability under the condition of minimizing the annual comprehensive cost as much as possible. The planning result not based on the diffusion model only considers general scenarios encountered during system operation, so the power supply reliability of the planning result is low, and it is difficult to effectively respond to various emergencies.

[0175] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0176] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0177] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0178] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0179] The above description is only preferred embodiments of the present application, not intended to limit other forms of the present application. Any person familiar with the art can make changes or modifications to the above-mentioned disclosed technical content as equivalent embodiments. However, any simple modification, equivalent change and modification of the above-mentioned embodiments without departing from the technical solution of the present application, according to the technical essence of the present application, still belongs to the protection scope of the present application.

Claims

1. A multi-flake network interconnection planning method based on a diffusion model and a multi-port SOP, characterized in that, The method comprises the following steps: S1, acquiring topology structure data, running state data, node load data, photovoltaic power generation data and wind power generation data of each snowflake network unit; S2, performing graph theory modeling on the topology structure data of the multiple snowflake networks, and identifying key interconnected nodes as SOP candidate positions; S3, performing normalization processing on the node load data, photovoltaic power generation data and wind power generation data respectively, and then inputting the data into a diffusion model to obtain node load typical scenarios, photovoltaic power generation typical scenarios and wind power generation typical scenarios of each snowflake network unit; S4, constructing a multiple snowflake network interconnection planning model based on the diffusion model and a multi-port SOP, wherein an optimization objective of the multiple snowflake network interconnection planning model is to minimize a linear weighted combination of annual comprehensive cost of multiple snowflake network interconnection equipment and system power supply reliability, and constraint conditions are composed of multiple snowflake network interconnection constraints, snowflake unit power flow constraints, snowflake unit operation constraints and SOP operation constraints; S5, inputting the node load typical scenarios, photovoltaic power generation typical scenarios and wind power generation typical scenarios of each snowflake network unit obtained in step S3 into the multiple snowflake network interconnection planning model constructed in step S4, solving the model, and obtaining a Pareto solution system of the model; S6, sorting the Pareto solution system based on an entropy weight method, selecting an equilibrium solution in the Pareto solution system, and outputting a multiple snowflake network interconnection planning result based on the diffusion model.

2. The method of claim 1, wherein, In step S1, the topology structure data of the snowflake network unit includes basic information such as types, positions and numbers of power supply points, transformers, switches and line devices of the snowflake network unit and connection relationships between the devices, and the running state data of the snowflake network unit includes load data of each node in the snowflake network unit and real-time running states of the devices.

3. The method of claim 1, wherein, The implementation method of step S2 is as follows: The physical structure of the snowflake network is mapped into a weighted directed graph G=(V, E, W), wherein V represents a set of all electrical nodes in the snowflake network, E represents line connection relationships, and W represents electrical details of edges, including line impedance, capacity limit and node load power. In combination with the electrical characteristics and the topology structure, a comprehensive node importance index S is defined i The specific expression is as follows: where C B (i) represents the betweenness centrality of node i, which measures the hub nature of the node in the network; P i flow represents the normalized power flow density of node i, reflecting its key role in power transmission; ΔV i represents the difference between the voltage of node i and the rated voltage, which is used to evaluate the voltage stability demand; V base represents the rated voltage; ω1, ω2, ω3 are weight coefficients; Select S i The top N ranked nodes are selected as candidate SOP access locations.

4. The method of claim 1, wherein, In step S3, the diffusion model mainly consists of two processes of diffusion and reverse; in the diffusion process, assuming that the original data sample is z0, the sample generated by the n-th diffusion is zn n , assuming that the prior distribution is a standard Gaussian distribution The diffusion process from z n-1 to z n is expressed as: wherein is the scaling factor for the n-th diffusion, z n The probability distribution of z is as follows: Let Then z n The relationship of z0 is as follows: z n The probability distribution of the sum of the two random variables X and Y is given by: wherein is the accumulated scaling factor, representing the proportion of the original data that remains after n steps of diffusion; I represents the identity matrix; In the backward pass, the diffusion model fits the conditional probability distribution p θ (z n-1 |z n ) through a neural network, whose expression is as follows: In the formula, p θ (z0) represents the probability distribution of the original data obtained from the diffusion model; p(z) N ) represents the prior distribution of the diffusion process; p θ (z n-1 |z n Let θ be a Gaussian distribution obtained through learnable parameters, and its expression is as follows:

5. The method of claim 1, wherein, In step S4, the optimization objective of the multiple snowflake network interconnection planning model is to minimize the value of the objective function f, and the expression of the optimization objective is as follows: minf = aC com + βξ reliability (8) Wherein, wherein, C com is the annual comprehensive cost of the snowflake network interconnection device, ξ reliability is the system power supply reliability, and α and β are weight coefficients, and α+β=1; N snow is the total number of snowflake network units; is the SOP investment cost of the i-th snowflake network unit; is the SOP operation and maintenance cost of the i-th snowflake network unit; is the power supply loss cost of the i-th snowflake network unit; is the power purchase cost of the i-th snowflake network unit from the upper-level power grid; is the interactive power cost of the i-th snowflake network unit with other units; The expression is: In the formula, N sop,i is the number of SOPs to be installed in the i-th snowflake network unit; c i,k,sop is the unit capacity investment cost of the k-th SOP of the i-th snowflake network unit; S i,k,sop is the capacity of the k-th SOP of the i-th snowflake network unit. The expression is: wherein η i is the operating and maintenance cost coefficient of the i-th snowflake network unit; The expression is: where c i is the electricity price of the ith snowflake network unit; N i,t is the total number of time periods in the ith snowflake network unit; Ω b,i is the set of all branches in the ith snowflake network unit; r ab is the resistance value of branch ab; I ab,t is the current amplitude of branch ab in the tth time period; N nopower,i is the total number of non-power supply nodes in the ith snowflake network unit; is the active loss of the SOP at the jth node of the ith snowflake network unit in the tth time period; The expression is: In the formula, T represents the total number of time periods, λ i,t the electricity purchase price from the upper-level power grid, P(t) represents the power purchased from the upper-level power grid at time t, and Δt represents the time step. The expression is: In the formula, C i,ex is the unit cost of the power exchanged between units, is the power exchanged between units. ξ reliability The lower the value of the number represents the better the system power supply reliability, and the expression is: ξ reliability = β1ξ node + β2ξ line (15) Wherein, wherein, ξ node represents the node reliability of the multi-snowflake network interconnection system; ξ line represents the line reliability of the multi-snowflake network interconnection system; β1 and β2 are weight coefficients; N snow represents the total number of snowflake units; m k represents the number of nodes in the kth snowflake unit; represents the load of node i in the kth snowflake unit at time t; represents the transformer capacity of node i in the kth snowflake unit; represents the transformer expansion capacity of node i in the kth snowflake unit; l k represents the number of lines in the kth snowflake unit; represents the load of line i in the kth snowflake unit at time t; represents the capacity of line i in the kth snowflake unit; represents the expansion capacity of line i in the kth snowflake unit.

6. The method of claim 1, wherein, In step S4, the expression of the multiple snowflake network interconnection constraint is as follows: wherein K t1 represents the simultaneous rate of conventional load; K t2 represents the simultaneous rate of flexible load; K t3 represents the simultaneous rate of energy storage; represents the grid supply load of the kth snowflake unit; m k represents the number of nodes in the kth snowflake unit; represents the conventional load electric power of the jth snowflake unit; represents the flexible load electric power of the jth snowflake unit; represents the energy storage charging and discharging power of the jth snowflake unit; represents the distributed power output power of the jth snowflake unit; represents the interaction power between the kth snowflake unit and other units; represents the interaction power between the kth snowflake unit and other units at time t; is the maximum value of the interaction power between snowflake units.

7. The method of claim 1, wherein, In step S4, the expression of the snowflake unit power flow constraint is as follows: where I t,ji,k is the current of line ji in the kth snowflake network unit at time t, assuming the positive direction of current is from node j to node i; Ω b is the set of all branches in the snowflake network unit; P t,iq,k represents the active power of node i flowing to node q in the kth snowflake network unit at time t; Q t,iq,k represents the reactive power of node i flowing to node q in the kth snowflake network unit at time t; r ij,k and x ij,k are the resistance and reactance of branch ij in the kth snowflake network unit; U t,i,k represents the voltage value of the ith node in the kth snowflake network unit at time t; P t,ij,k is the active power of node i flowing to node j in the kth snowflake network unit at time t, Q t,ij,k is the reactive power of node i flowing to node j in the kth snowflake network unit at time t; P t,i,k is the sum of the active power injected at node i in the kth snowflake network unit at time t, are the active power injected by SOP and consumed by load at node i in the kth snowflake network unit at time t, respectively; Q t,i,k is the sum of the reactive power injected at node i in the kth snowflake network unit at time t, are the reactive power injected by SOP and consumed by load at node i in the kth snowflake network unit at time t, respectively.

8. The method of claim 1, wherein, In step S4, the expression of the snowflake unit operation constraint is as follows: wherein U i,k represents the voltage value of the i-th node of the k-th snowflake network unit, U min,k and U max,k are the minimum and maximum allowed node voltage values of the k-th snowflake network unit, respectively; I ij,k represents the current value of the branch ij in the k-th snowflake network unit; I max,k is the maximum allowed branch current value of the k-th snowflake network unit.

9. The method of claim 1, wherein, In step S4, the SOP operation constraint is as follows: wherein, and are the active and reactive power of the SOP connected to node i in the kth snowflake cell, respectively; is the loss of the SOP connected to node i in the kth snowflake cell; is the loss coefficient of the SOP connected to node i in the kth snowflake cell; is the capacity of the SOP connected to node i.

10. The method of claim 1, wherein, The multiple snowflake network interconnection planning result output in step S6 is compared with a planning result not based on the diffusion model to realize evaluation of planning effect.

Images