Modeling method, device, equipment and storage medium for deep aggregation power grid model

Abstract

The embodiment of the present invention provides a modeling method, device, equipment and storage medium for a deep aggregation power grid model, which relates to the technical field of power grid planning. Among them, this modeling method includes steps S1 to S4. S1. Obtain system information of the power system. Among them, the system information includes plant information. S2. Construct an initial plant model based on the plant information. Among them, the initial plant model includes an initial model of a power plant and an initial model of a contact station. S3. Obtain the flow data of the power plant. S4. According to the flow data of the power plant, based on the long short-term memory network, solve the unknown quantity of the initial plant model to obtain the aggregated plant model. Among them, the aggregated plant model includes an aggregate model of a power plant and an aggregate model of a contact station. The aggregated plant model and the deep aggregation power grid model of the present invention can well combine the actual power grid operating conditions, thereby simulating the power grid more realistically. Using an LSTM network for solving has higher accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid planning, and in particular to a modeling method, device, equipment and storage medium for a model of an aggregated power grid. Background Art

[0002] Equivalence simplification is the basis of economic dispatch and control of power grids and the core of power grid energy management system. Traditional power grid economic dispatch and control has always adopted the hierarchical and separate method of power generation, transmission, distribution and consumption, that is, each level of power grid simplifies the upper level (i.e. high voltage level, the same below) power grid into power source and the lower level (i.e. low voltage level, the same below) power grid into load, thereby implementing centralized dispatch and control strategies in power grids at all levels.

[0003] Renewable energy generation is massively integrated into power grids of various voltage levels in the form of decentralized, centralized and microgrids, making the boundary between transmission and distribution gradually unclear. In order to safely, reliably and maximize the consumption of renewable energy generation, it is necessary to coordinate power grids of various voltage levels such as generation, transmission and distribution, whether it is advance scheduling or online control. Establishing a real-time and accurate simplified equivalent model of the external power grid (hereinafter referred to as the external grid) is the key to solving this problem. In this regard, in the prior art, a topological equivalent method with known external grid parameters and a state estimation method with unknown external grid parameters are proposed.

[0004] Although rich results have been achieved in the equivalent simplification methods of the external network, these equivalent simplification methods have added or deleted some active quantities (such as generator active power, generator reactive power, reactive compensation equipment output, etc.) and passive quantities (such as load and network structure, etc.) of the external network, thus posing challenges to the precise control of active and passive quantities of the external network in the process of overall coordinated dispatch and control of the power grid.

[0005] In view of this, the applicant filed this application after studying the existing technology. Summary of the invention

[0006] The present invention provides a modeling method, device, equipment and storage medium for a model of an aggregated power grid to improve at least one of the above technical problems.

[0007] First,

[0008] An embodiment of the present invention provides a modeling method for a deep aggregation power grid model, which includes steps S1 to S4.

[0009] S1. Obtaining system information of the power system, wherein the system information includes plant information.

[0010] S2. Construct an initial plant model based on the plant information, wherein the initial plant model includes an initial model of a power plant and an initial model of a connection station.

[0011] S3. Obtain the power flow data of the power plant.

[0012] S4. According to the power plant's flow data, based on the long short-term memory network, the unknown quantity of the initial plant model is solved to obtain the aggregated plant model. The aggregated plant model includes the power plant aggregation model and the interconnection station aggregation model.

[0013] Second,

[0014] An embodiment of the present invention provides a modeling device for a deep aggregated power grid model, which includes:

[0015] The system information acquisition module is used to acquire the system information of the power system, wherein the system information includes plant information.

[0016] The plant model acquisition module is used to construct an initial plant model according to the plant information, wherein the initial plant model includes an initial model of a power plant and an initial model of a connection station.

[0017] The flow data acquisition module is used to obtain the flow data of the power plant.

[0018] The plant model solving module is used to solve the unknown quantities of the initial plant model based on the power plant flow data and the long short-term memory network to obtain the aggregated plant model. The aggregated plant model includes the power plant aggregation model and the interconnection station aggregation model.

[0019] Thirdly,

[0020] An embodiment of the present invention provides a modeling device for a deep aggregated power grid model, which includes a processor, a memory, and a computer program stored in the memory. The computer program can be executed by the processor to implement the modeling method as described in any paragraph of the first aspect.

[0021] Fourthly,

[0022] An embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the modeling method of the deep aggregated power grid model as described in any paragraph of the first aspect.

[0023] By adopting the above technical solution, the present invention can achieve the following technical effects:

[0024] The proposed aggregated plant model and deep aggregated power grid model in the embodiment of the present invention can well combine the actual power grid working conditions, thereby simulating the power grid more realistically. The aggregated plant model and deep aggregated power grid model are solved using an LSTM network with higher accuracy.

[0025] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 It is a flowchart of the modeling method of the first embodiment of the present invention.

[0028] Figure 2 It is a schematic diagram of the power system.

[0029] Figure 3 is a schematic diagram of an aggregated power grid.

[0030] Figure 4 This is the internal structure of LSTM.

[0031] Figure 5 Schematic diagram of the structure of a modeling device according to a second embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0034] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0035] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0036] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0037] The "first\second" mentioned in the embodiments is only to distinguish similar objects, and does not represent a specific order for the objects. It is understandable that the "first\second" can be interchanged with the specific order or sequence where permitted. It should be understood that the objects distinguished by "first\second" can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than those illustrated or described herein.

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0039] Embodiment 1:

[0040] See also Figures 1 to 4 The first embodiment of the present invention provides a modeling method of a deep aggregation power grid model, which can be performed by a modeling device of a deep aggregation power grid model (hereinafter referred to as a modeling device). In particular, it is performed by one or more processors in the modeling device to implement steps S1 to S4.

[0041] S1. Obtaining system information of the power system, wherein the system information includes plant information.

[0042] Specifically, when the plant is a power plant, the plant information includes the number of generators in the power plant, the active power output of generator i Reactive power output by generator i The reactive power output of the reactive compensation equipment that can be switched on and off in the power plant k Active load of power plant k Reactive load of power plant k

[0043] When the plant station is a contact station, the plant station information also includes reactive power flowing out of the contact station s and active power flowing into the contact station s.

[0044] It is understandable that the modeling device may be an electronic device with computing capabilities, such as a portable notebook computer, a desktop computer, a server, a smart phone, or a tablet computer.

[0045] S2. Construct an initial plant model based on the plant information, wherein the initial plant model includes an initial model of a power plant and an initial model of a connection station.

[0046] The inventor proposed the concept of aggregated power grid in the literature ("Concept, Control Method and Application of Aggregated Power Grid Based on Panoramic Measurement", Hao Guangtao, Han Xueshan, Lin Zhenheng, et al. [J]. Electric Power Automation Equipment, 2019, 39(10): 187-193). Figure 2 As shown in Figure 1, the power system is an organic whole composed of transmission and distribution links of several voltage levels through power plants and lines. If the power plants are regarded as aggregated power plants gathered at one end of the line, then at a certain voltage level, the power grid composed of several aggregated power plants and their associated lines is an aggregated power grid.

[0047] According to this concept, the core of the formation of an aggregated power grid is to establish an aggregated plant model. Plants and substations can be divided into two categories according to their location and formation in the actual power system: one is the power plant (including renewable energy power station) and substation located at both ends (i.e. the head end and the tail end) of the power system; the other is the contact station located in the middle of the power system.

[0048] For power plant k, its active and passive quantities are obviously known (substations have the same analysis method). According to the power balance relationship between power generation and power consumption in power plant k, the active power output of power plant k at time t can be and reactive power Expressed as:

[0049]

[0050] Where: N k represents the total number of generators in power plant k; represents the active power output by generator i at time t; represents the reactive power output by generator i at time t; It represents the reactive power output by the reactive compensation equipment that can be switched on and off at power plant k at time t; represents the active load of power plant k at time t; represents the reactive load of power plant k at time t; represents the active line loss of branch k of the power plant at time t; Represents the reactive line loss of branch k of the power plant at time t.

[0051] Then, the formula (1) is expressed in active and reactive forms respectively to obtain the initial model of the power plant:

[0052]

[0053] In the formula, represents the active power output of power plant k at time t, N k represents the total number of generators in power plant k, represents the active power output by generator i at time t, represents the active load of power plant k at time t, represents the active line loss of branch k of the power plant at time t, represents the reactive power output of power plant k at time t, represents the reactive power output by generator i at time t, It indicates the reactive power output of the reactive compensation equipment that can be switched on and off at power plant k at time t, represents the reactive load of power plant k at time t, Represents the reactive line loss of branch k of the power plant at time t.

[0054] Assuming that time t is the advance time, to achieve advance decision-making, its internal active quantity must be known Adjustable range and passive volume Numeric value.

[0055] In formula (2) For power plant k, the controllable information (such as the active and reactive output ranges of the generators, the ramp rate, etc.) is known; the active load and reactive load It can be obtained by predicting the time t-Δt before time t; only active line loss on the branch and reactive line loss Unknown, if we change the Expressed as a function of active quantity, equation (2) realizes an aggregation model that eliminates unknown quantities and retains active quantities to participate in external scheduling and control.

[0056] For the contact station s, it is located between power grids of different voltage levels in the power system and plays a role in communication and coordination

[20] . It has four parts: power generation, power consumption, incoming power, and outgoing power. According to the power balance relationship between these four parts, the initial model of the contact station can be obtained:

[0057]

[0058] In the formula, represents the active power flowing out of the interconnection station s at time t, represents the active power flowing into the contact station s at time t, N s Indicates the total number of generators in the contact station, represents the active power output by generator i at time t, represents the active load of the interconnection station s at time t, represents the active line loss of branch s of the contact station at time t, represents the reactive power flowing out of the interconnection station s at time t, represents the reactive power flowing into the interconnection station s at time t, represents the reactive power output by generator i at time t, It indicates the reactive power output of the reactive compensation equipment that can be switched on at the contact station s at time t. represents the reactive load of the interconnection station s at time t, Represents the reactive line loss of branch s of the interconnection station at time t.

[0059] In formula (3), the power flowing out of the interconnection station s and the power flowing into the interconnection station s belong to different voltage level grids. It is the decentralized coordination quantity when multiple voltage level grids are coordinated (that is, when a certain voltage level grid is locally autonomous, the outflow or inflow power of other voltage levels can be considered known).

[0060] It can be seen that only the active line loss in equation (3) and reactive line loss If it is expressed as a function of active quantity at the leading time t, a contact station aggregation model is formed that eliminates unknown quantities and retains active quantities to participate in scheduling and control.

[0061] S3. Obtain the power flow data of the power plant.

[0062] S4. According to the power plant's flow data, based on the long short-term memory network, the unknown quantity of the initial plant model is solved to obtain the aggregated plant model. The aggregated plant model includes the power plant aggregation model and the interconnection station aggregation model.

[0063] It is understandable that the basis for the global coordinated dispatch and control of the power system is to establish an accurate equivalent simplified model. In this regard, the embodiment of the present invention proposes an initial plant model that is more in line with the actual power grid. In addition, by solving the initial plant model through the LSTM network, a more accurate solution can be obtained.

[0064] Based on the above embodiment, in an optional embodiment of the present invention, the system information also includes line information, wherein the line information includes a voltage level.

[0065] Preferably, the modeling method further includes steps S5 to S8.

[0066] S5. Construct an aggregated power grid according to the line information. The aggregated power grid model includes a source aggregated power grid and a connection aggregated power grid.

[0067] The inventor proposed the concept of aggregated power grid in the literature ("Concept, Control Method and Application of Aggregated Power Grid Based on Panoramic Measurement", Hao Guangtao, Han Xueshan, Lin Zhenheng, et al. [J]. Electric Power Automation Equipment, 2019, 39(10): 187-193). Figure 2 As shown in Figure 1, the power system is an organic whole composed of transmission and distribution links of several voltage levels through power plants and lines. If the power plants are regarded as aggregated power plants gathered at one end of the line, then at a certain voltage level, the power grid composed of several aggregated power plants and their associated lines is as follows: Figure 3 The aggregated grid shown.

[0068] Specifically, Figure 2 The power system shown in FIG. 1 is formed according to the initial model of the power plant and the initial model of the interconnection station in step S2, as well as the voltage level of the power grid line. Figure 3 The aggregated grid shown.

[0069] S6. Constructing an initial aggregated grid model according to the aggregated grid model, wherein the initial aggregated grid model includes a source initial aggregated grid model and a connection initial aggregated grid model.

[0070] Specifically, for some aggregated power grids, in their long-term, process-based operation, the active power of their transmission lines and node voltages do not exceed safety limits in any scenario, and their essence is a single busbar model. Therefore, in advance decision-making (such as advance dispatch and control), the aggregated power plants connected to them can be aggregated again to form a deep aggregated power grid model.

[0071] Aggregate grids can be divided into two categories according to their location in the power system: one is the aggregate grid located at the headend and the end, and the other is the aggregate grid located in the middle.

[0072] For the source aggregated grid K (with the same load characteristics, but with different active and passive quantities), it is assumed that it contains m power plants (including substations) and n interconnection stations. According to the power balance relationship between power generation and power consumption, the source initial aggregated grid model of the aggregated grid K at time t is:

[0073]

[0074] In the formula, represents the active power flowing out of the aggregated grid K at time t, m is the number of power plants in the aggregated grid K, N k represents the total number of generators in power plant k, represents the active power output by generator i at time t, represents the active load of power plant k at time t, and n is the aggregated power grid K The number of contact stations, N s Indicates the total number of generators in the contact station, represents the active load of the interconnection station s at time t, It means that the total active line loss of the aggregated power grid K at time t is the sum of the active line loss of the plant branch and the line. represents the reactive power flowing out of the aggregated grid K at time t, represents the reactive power output by generator i at time t, It indicates the reactive power output of the reactive compensation equipment that can be switched on and off at power plant k at time t, represents the reactive load of power plant k at time t, It indicates the reactive power output of the reactive compensation equipment that can be switched on at the contact station s at time t. represents the reactive load of the interconnection station s at time t, It represents the total reactive line loss of the aggregated power grid K at time t, which is the sum of the reactive line losses of the plant branch and the line.

[0075] From formula (4), we can see that only and If it is an unknown quantity and is expressed as an explicit function of the active quantity at the leading time t, a deep aggregation power grid model can be formed, which retains the active quantity and eliminates the unknown quantity from participating in the coordinated dispatch and control of the power grid.

[0076] For the interconnected aggregated grid S, assuming that it contains m power plants (including substations) and n interconnected stations, then according to the power balance relationship between power generation and power consumption, the interconnected initial aggregated grid model of the aggregated grid S at time t can be obtained as follows:

[0077]

[0078] In the formula, represents the active power flowing out of the aggregated grid S at time t, represents the active power flowing into the aggregated grid S at time t, and m is the aggregated grid K The number of power plants, N k represents the total number of generators in power plant k, represents the active power output by generator i at time t, represents the active load of power plant k at time t, n is the number of interconnection stations in the aggregated power grid K, N s Indicates the total number of generators in the contact station, represents the active load of the interconnection station s at time t, It means that the total active line loss of the aggregated power grid at time t is the sum of the active line loss of the plant branch and the line. represents the reactive power flowing out of the aggregated grid S at time t, represents the reactive power flowing into the aggregated grid S at time t, represents the reactive power output by generator i at time t, It indicates the reactive power output of the reactive compensation equipment that can be switched on and off at power plant k at time t, represents the reactive load of power plant k at time t, n is the number of interconnection stations in the aggregated power grid K, It indicates the reactive power output of the reactive compensation equipment that can be switched on at the contact station s at time t. represents the reactive load of the interconnection station s at time t, It means that the total reactive line loss of the aggregated power grid S at time t is the sum of the reactive line losses of the plant branch and the line.

[0079] The active power flowing out of the aggregated grid S in formula (5) is and reactive Active power flowing into the aggregated grid S and reactive It is the decentralized coordination quantity in the process of global coordinated dispatch and control of the power grid, that is, when a certain aggregated power grid is locally autonomous, the outflow or inflow power of other aggregated power grids can be considered known. It can be seen that, similar to equations (2) to (4), only Is an unknown quantity.

[0080] S7. Obtain power flow data of the aggregated power grid.

[0081] S8. According to the power flow data of the aggregated power grid, based on the long short-term memory network, the unknown quantity of the initial aggregated power grid model is solved to obtain a deep aggregated power grid model. The deep aggregated power grid model includes a source deep aggregated power grid model and a connection deep aggregated power grid model.

[0082] It should be noted that, as mentioned above, the unknown quantities in equations (2) to (5) are active line loss and reactive line loss. If these unknown quantities are expressed as functions of generator power, they can be combined with the corresponding generator power in equations (2) to (5) to form a model in which both active and passive quantities are known. For the external power flowing in equations (3) and (5), it should be regarded as power generation, and the outflowing power should be regarded as load.

[0083] Specifically, step S4 solves the initial plant model, and step S8 solves the initial aggregated power grid model, and the specific steps are the same. The following is an explanation of step S4.

[0084] Based on the above embodiment, in an optional embodiment of the present invention, step S4 specifically includes steps S41 to S44.

[0085] S41. According to the initial plant model, obtain the active line loss P L and reactive line loss Q L The solution model:

[0086]

[0087]

[0088] Where P G represents the active power column vector of the generator node, Q G Represents the reactive power column vector of the generator node, B′ GG Indicates B GG The first derivative of GG Indicates B GG The second derivative of G0 Indicates B G0 The first derivative of G0 Indicates B G0 The second derivative of 00 Indicates B 00 The first derivative of 00 Indicates B 00 The second derivative of G0* Indicates B G0* The first derivative of G0* Indicates B G0* The second derivative of A G Indicated by A Gi (i=1,2,…,N G ) is a complex coefficient column vector composed of elements, E represents the unit matrix, k i represents the linear complex coefficient, N D Indicates the total number of load nodes in the aggregation, N G represents the total number of generator nodes in the aggregation, Z represents the impedance matrix of nodes in the aggregation, I 0 Indicated by The column vector, I′ 0 Indicates I 0 The first derivative of , T represents the transpose, and * represents the conjugate complex number.

[0089] When solving the initial plant model or the initial aggregated grid model, it is necessary to first express the unknown quantity in the formula as a function of the generator power. That is, the unknown quantity is expressed as an explicit function of the active quantity.

[0090] Specifically, let the total number of nodes in a certain aggregation (aggregation plant or aggregation power grid) be N, then the total active line loss P of the aggregation is L , total reactive line loss Q L It can be expressed as:

[0091] S L =P L +jQ L =I T Z T I * (6)

[0092] Where: SL represents the complex power line loss in the aggregation; I represents the column vector of the node injection current phasor in the aggregation, I T Transpose it, I * is its conjugate; Z represents the node impedance matrix within the aggregation, Z T Transpose it.

[0093] The node injection current phasor I in equation (6) is expressed as two parts: the generator node and the load node:

[0094] I=I G +I D (7)

[0095] Where: I G I represents the column vector of the current phasor injected into the generator node; D Column vector representing the phasor of the current injected into the load node.

[0096] Generally speaking, the current phasor injected into the load node i is Has a linear non-uniform relationship with the total load within the aggregate:

[0097]

[0098] Where: represents the sum of current phasors of all load nodes in the aggregation; k i represents a linear complex coefficient; Represents a random number.

[0099] The sum of all load node current phasors in formula (8) The relationship between the load node and generator node current is as follows:

[0100]

[0101] Where: N D Represents the total number of load nodes in the aggregation; N G Indicates the total number of generator nodes in the aggregate; represents the current phasor injected into the generator node i.

[0102] Substituting equation (9) into equation (8) and expressing it in matrix form is:

[0103] I D =-KI G +I 0 (10)

[0104] Where: I D The column vector representing the load node injected current phasor, whose elements are I GThe column vector representing the generator node injected current phasor, whose elements are I 0 Indicated by Column vector composed of .

[0105] Substituting formula (10) into formula (7) and combining similar terms, we get:

[0106] I=(EK)I G +I 0 (11)

[0107] Where: E represents the unit matrix.

[0108] Since the generator node injects current phasor and the node injection power S Gi , voltage phasor There is the following relationship between them:

[0109]

[0110] Where: P Gi represents the active power injected by node i; Q Gi represents the reactive power injected by node i; V i * Represents the voltage phasor of node i conjugation; Indicates S Gi conjugation of A Gi represents the linearized complex coefficient; A i0 Indicates the complex offset.

[0111] Expressing formula (12) in matrix form, we get:

[0112]

[0113] Where: A G Indicated by A Gi (i=1,2,…,N G ) is a complex coefficient column vector consisting of elements; Represented by the node complex power conjugate The matrix composed of 0 Indicates the node offset A i0 (i=1,2,…,N G ) is a complex bias matrix.

[0114] Substituting formula (13) into formula (11), we get:

[0115]

[0116] Where: I′ 0 =(EK)A0 +I 0 .

[0117] Substituting formula (14) into formula (6), we get:

[0118]

[0119] Expanding formula (15) yields:

[0120]

[0121] Where: P G , Q G They represent the active power column vector and reactive power column vector of the generator node respectively, and the superscript T represents the transpose; other parameters are shown in formula (17):

[0122]

[0123] In the formula, the superscript * indicates conjugation.

[0124] From the above derivation process, it can be seen that formula (16) does not use any assumptions. By expanding it, it can be seen that: This formula expresses the relationship between active (or reactive) line loss and aggregated internal N G Active power square of generator Generator active power combination P Gi P Gj …P Gk (i,j,k=1,2,…,N G ; i≠j≠k), generator reactive power square Reactive power combination Q Gi Q Gj …Q Gk (i,j,k=1,2,…,N G ; i≠j≠k), active power P Gi P Gj …P Gk (i,j,k=1,2,…,N G ; i≠j≠k) and reactive power Q Gi Q Gj …Q Gk (i,j,k=1,2,…,N G ; i≠j≠k) combination, generator active power P Gi (i=1,2,…,N G ), generator reactive power Q Gi (i=1,2,…,N G )’s functional relationship.

[0125] If the above coefficients are solved, the line loss in equations (2) to (5) can be expressed as an explicit function of the active power and reactive power of the generator, and the corresponding aggregation model can be obtained.

[0126] S42, obtaining the power plant's flow data, wherein the flow data includes N historical event sections.

[0127] S43, according to the power plant's flow data, with active line loss P L and reactive line loss Q L The solution model is the object. G The square of the active power of the generator, the combination of the active power of the generator, the square of the reactive power of the generator, the combination of the reactive power, the combination of the active and reactive power, the active power of the generator, and the reactive power of the generator are taken as input. The corresponding active line loss and reactive line loss are taken as output to form a training set.

[0128] S44, input the input data in the training set into the LSTM network to obtain the predicted value, and use the back propagation algorithm to learn and train the LSTM weights according to the error between the predicted value and the true value to obtain the active line loss P L and reactive line loss Q L The coefficients of the solved model are obtained to obtain the aggregated plant model.

[0129] It should be noted that the basis for the global coordinated dispatch and control of the power system is to establish an accurate equivalent simplified model. In this regard, the embodiments of the present invention propose an aggregated power grid model and a deep aggregated power grid model that are more in line with the actual power grid. And the LSTM network is used to solve the aggregated power grid and deep aggregated power grid models, which can obtain a more accurate solution.

[0130] Specifically, as can be seen from formula (16), the model has many coefficients to be solved. In order to obtain a high-precision solution, this paper uses the long short-term memory network (LSTM) in Appendix A to solve it. The LSTM network is composed of multiple LSTM units, and the input and output of each LSTM unit can be a multidimensional data matrix, thereby realizing learning from historical data. Based on this, the embodiment of the present invention is based on the flow data of N historical time sections (greater than the number of generators in the aggregation) to solve the parameters of formula (16), and its process is described as follows.

[0131] 1) Based on the historical tide data of N time sections, take the model of formula (16) as the object, and replace N in formula (16) with G Active power square of generator Generator active power combination P Gi P Gj …P Gk (i,j,k=1,2,…,NG ; i≠j≠k), generator reactive power square Reactive power combination Q Gi Q Gj …Q Gk (i,j,k=1,2,…,N G ; i≠j≠k), active power P Gi P Gj …P Gk (i,j,k=1,2,…,N G ; i≠j≠k) and reactive power Q Gi Q Gj …Q Gk (i,j,k=1,2,…,N G ; i≠j≠k) combination, generator active power P Gi (i=1,2,…,N G ), generator reactive power Q Gi (i=1,2,…,N G ) as input, and the corresponding active line loss and reactive line loss as output to form a training set.

[0132] 2) Input the input data in the training set into the LSTM network to obtain the predicted value. The LSTM weights are trained using the back propagation algorithm based on the error between the predicted value and the true value.

[0133] 3) After training, the coefficients in formula (16) can be obtained.

[0134] 4) Select the input data in the test set and the trained coefficients for calculation to obtain the predicted active line loss and reactive line loss values, and then test the effect of the predicted value.

[0135] Appendix A

[0136] LSTM is a type of recurrent neural network that has the ability to memorize historical experience information and learn rules from it. Its unit structure is as follows: Figure 4 As shown. Figure 4 It can be seen that LSTM contains three gate control units: forget gate, input gate and output gate, and their calculation methods are as follows:

[0137] First, the LSTM unit receives the current input information x through the forget gate t and the implicit information h at the previous moment t-1 , the output f is obtained by filtering with the activation function σ t :

[0138] f t =σ(w fx x t +w fh h t-1+b f ) (A1)

[0139] Where: w fx and w fh represents the weight of the forget gate; b f represents the bias; σ represents the activation function.

[0140] Secondly, the input gate filters the current input information x t , the implicit information h of the previous moment t-1 , get the filter value i t and new knowledge

[0141]

[0142] Where: w ix 、w ih 、w cx 、w ch represents the input gate weight; b i and b c represents the bias; σ represents the activation function.

[0143] Third, by filtering the value i t and new knowledge Get the updated information C at time t t :

[0144]

[0145] By C t Get the current state of the output gate h t :

[0146]

[0147] Where: w ox 、w oh represents the output gate weight; b o represents the bias; σ represents the activation function.

[0148] Finally, the current state information h of the output gate is filtered t , get the output value of the LSTM network:

[0149] y t =σ(w y h t +b y ) (A5)

[0150] Where: w y represents weight; b y represents the bias; σ represents the activation function.

[0151]

[0152] Where: P 1 t represents the active power output by station 1 at time t; represents the active load of station 1 at time t; represents the active line loss of station 1 at time t; represents the reactive power output by station 1 at time t; represents the reactive load of station 1 at time t; Represents the reactive line loss of station 1 at time t; other symbols are the same as before.

[0153] The unknown quantity in formula (A6) is and According to formula (16), the functional relationship between the unknown quantity and the two generators at station 1 is established:

[0154]

[0155] Where: e 1 is the coefficient to be determined.

[0156]

[0157] Where: e′ 1 is the coefficient to be determined.

[0158] Expand equations (A7) to (A8) and combine similar terms to obtain:

[0159]

[0160] Where: a 1 =a 12 +a 21 .

[0161]

[0162] Where: a′ 1 =a′ 12 +a′ 21 .

[0163]

[0164] Where: represents the active power flowing out of the aggregated grid 2 through station 5 at time t; represents the active power flowing into the aggregated grid 2 through station 2 at time t; represents the reactive power generated by the generator of station 3 at time t; represents the total active load in the aggregated power grid 2 at time t, which is the sum of the active loads of stations 2, 4, and 5; represents the total active line loss in the aggregated power grid 2 at time t; represents the reactive power flowing out of the aggregated grid 2 through station 5 at time t; represents the reactive power flowing into the aggregated grid 2 through station 2 at time t; represents the reactive power generated by the generator of station 3 at time t; represents the total reactive load in the aggregated power grid 2 at time t, which is the sum of the reactive loads of stations 2, 4, and 5; Represents the total reactive line loss in the aggregated power grid 2 at time t.

[0165] The unknown quantity in formula (A11) is and According to formula (16), the unknown quantity and the generator active power are established. and Reactive and The functional relationship between them is:

[0166]

[0167] In the formula: a, b, c, d, e, f, g, h, i, j, k, s and a′, b′, c′, d′, e′, f′, g′, h′, i′, j′, k′, s′ represent the coefficients to be determined.

[0168] Embodiment 2

[0169] like Figure 5 As shown, an embodiment of the present invention provides a modeling device for a deep aggregation power grid model, which includes:

[0170] The system information acquisition module 1 is used to acquire the system information of the power system, wherein the system information includes plant information.

[0171] The plant model acquisition module 2 is used to construct an initial plant model according to the plant information, wherein the initial plant model includes an initial model of a power plant and an initial model of a connection station.

[0172] The flow data acquisition module 3 is used to acquire the flow data of the power plant.

[0173] The plant model solving module 4 is used to solve the unknown quantity of the initial plant model based on the long short-term memory network according to the power plant flow data, and obtain the aggregated plant model. The aggregated plant model includes the power plant aggregation model and the interconnection station aggregation model.

[0174] Based on the above embodiment, in an optional embodiment of the present invention, the plant model solving module 4 includes:

[0175] The solution model acquisition unit is used to obtain the active line loss P according to the initial plant model. L and reactive line loss Q L The solution model:

[0176]

[0177]

[0178] Where P G represents the active power column vector of the generator node, Q G Represents the reactive power column vector of the generator node, B′ GG Indicates B GG The first derivative of GG Indicates B GG The second derivative of G0 Indicates B G0 The first derivative of G0 Indicates B G0 The second derivative of 00 Indicates B 00 The first derivative of 00 Indicates B 00 The second derivative of G0* Indicates B G0* The first derivative of G0* Indicates B G0* The second derivative of A G Indicated by A Gi (i=1,2,…,N G ) is a complex coefficient column vector composed of elements, E represents the unit matrix, k i represents the linear complex coefficient, N D Indicates the total number of load nodes in the aggregation, N G represents the total number of generator nodes in the aggregation, Z represents the impedance matrix of nodes in the aggregation, I 0 Indicated by The column vector, I′ 0 Indicates I 0 The first derivative of , T represents the transpose, and * represents the conjugate complex number.

[0179] The flow data acquisition unit is used to acquire the flow data of the power plant, wherein the flow data includes N historical event sections.

[0180] The training set acquisition unit is used to obtain the active line loss P according to the power flow data of the power plant. L and reactive line loss Q L The solution model is the object. GThe square of the active power of the generator, the combination of the active power of the generator, the square of the reactive power of the generator, the combination of the reactive power, the combination of the active and reactive power, the active power of the generator, and the reactive power of the generator are taken as input. The corresponding active line loss and reactive line loss are taken as output to form a training set.

[0181] The plant model solving unit is used to input the input data in the training set into the LSTM network to obtain the predicted value, and use the back propagation algorithm to learn and train the LSTM weights according to the error between the predicted value and the true value to obtain the active line loss P L and reactive line loss Q L The coefficients of the solved model are obtained to obtain the aggregated plant model.

[0182] Based on the above embodiment, in an optional embodiment of the present invention, the modeling method further includes:

[0183] The aggregation grid construction module is used to construct an aggregation grid according to line information. The aggregation grid model includes a source aggregation grid and a connection aggregation grid.

[0184] The power grid model building module is used to build an initial aggregated power grid model according to the aggregated power grid model. The initial aggregated power grid model includes a source initial aggregated power grid model and a connection initial aggregated power grid model.

[0185] The power grid data acquisition module is used to obtain the power flow data of the aggregated power grid.

[0186] The power grid model solving module is used to solve the unknowns of the initial aggregated power grid model based on the long short-term memory network according to the power flow data of the aggregated power grid, and obtain the deep aggregated power grid model. The deep aggregated power grid model includes the source deep aggregated power grid model and the connection deep aggregated power grid model.

[0187] Embodiment 3

[0188] The embodiment of the present invention provides a modeling device for a deep aggregated power grid model, which includes a processor, a memory, and a computer program stored in the memory. The computer program can be executed by the processor to implement the modeling method described in any paragraph of the first embodiment.

[0189] Embodiment 4:

[0190] An embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the modeling method of the deep aggregation power grid model as described in any paragraph of the first embodiment.

[0191] In several embodiments provided in the embodiments of the present invention, it should be understood that the disclosed apparatus and method can also be implemented in other ways. The apparatus and method embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the apparatus, method and computer program product according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, program segment or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0192] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0193] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention is essentially or partly contributed to the prior art or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, an electronic device, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk. It should be noted that, in this article, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements inherent to such a process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0194] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A modeling method for a deep aggregation power grid model, characterized in that: Include: Acquire system information of the power system; wherein the system information includes plant information; According to the plant information, an initial plant model is constructed; wherein the initial plant model includes an initial model of a power plant and an initial model of a connection station; Obtain power flow data for power plants; According to the power flow data of the power plant, based on the long short-term memory network, the unknown quantity of the initial plant model is solved to obtain the aggregated plant model; wherein the aggregated plant model includes the power plant aggregation model and the contact station aggregation model; The initial model of the power plant is , where express Moment Power Plant The output active power, Indicates power plant Total number of generators, express Time generator Output active power, express Moment Power Plant Active load, express Moment Power Plant Active line loss of branch circuit, express Moment Power Plant Reactive power output to the outside, express Time generator Output reactive power, express Moment Power Plant The reactive power output of the reactive compensation equipment can be switched on and off. express Moment Power Plant The reactive load, express Moment Power Plant Reactive line loss of the branch circuit; The initial model of the contact station is , where express Time outflow contact station The active power, express Time flow into the contact station The active power, Indicates the total number of generators in the contact station, express Time generator Output active power, express Time contact station Active load, express Time contact station Active line loss of branch circuit, express Time outflow contact station The reactive power, express Time flow into the contact station The reactive power, express Time generator Output reactive power, express Time contact station The reactive power output of the reactive compensation equipment can be switched on and off. express Time contact station The reactive load, express Time contact station Reactive line loss of the branch circuit; The plant information includes the number of generators in the power plant; the unknown quantity of the initial plant model is the active line loss and reactive line loss ; According to the power flow data of the power plant, based on the long short-term memory network, the unknown quantity of the initial power plant model is solved to obtain the aggregated power plant model, which specifically includes: According to the initial plant model, the active line loss is obtained. and reactive line loss The solution model: In the formula, represents the active power column vector of the generator node, represents the reactive power column vector of the generator node, express The first-order derivative of express The second-order derivative of express The first-order derivative of express The second-order derivative of express The first-order derivative of express The second-order derivative of express The first-order derivative of express The second-order derivative of Indicated by is a complex coefficient column vector consisting of elements, represents the identity matrix, , represents the linear complex coefficient, Indicates the total number of load nodes in the aggregation, Represents the total number of generator nodes in the aggregation, represents the node impedance matrix within the aggregation, Indicated by The column vector composed of express The first-order derivative of Indicates transposition, represents a conjugate complex number; Acquire the power flow data of the power plant; wherein the power flow data includes N historical event sections; According to the power flow data of the power plant, the active line loss and reactive line loss The solution model of The square of the active power of each generator, the combination of the active power of each generator, the square of the reactive power of each generator, the combination of reactive power, the combination of active and reactive power, the active power of each generator, and the reactive power of each generator are used as inputs; the corresponding active line loss and reactive line loss are used as outputs to form a training set; The input data in the training set is input into the LSTM network to obtain the predicted value, and the LSTM weights are trained using the back propagation algorithm according to the error between the predicted value and the true value to obtain the active line loss and reactive line loss The coefficients of the solution model are obtained to obtain the aggregation plant model; The system information also includes line information; wherein the line information includes voltage level; The modeling method also includes: According to the line information, an aggregated power grid is constructed; wherein the aggregated power grid model includes a source aggregated power grid and a connection aggregated power grid; According to the aggregated power grid model, an initial aggregated power grid model is constructed; wherein the initial aggregated power grid model includes a source initial aggregated power grid model and a connection initial aggregated power grid model; Obtaining power flow data of aggregated power grid; According to the power flow data of the aggregated power grid, based on the long short-term memory network, the unknown quantity of the initial aggregated power grid model is solved to obtain a deep aggregated power grid model; wherein the deep aggregated power grid model includes a source deep aggregated power grid model and a connection deep aggregated power grid model; The source initial aggregation grid model is , where express Time outflow aggregation grid The active power, For the aggregated grid The number of power plants, Indicates power plant Total number of generators, express Time generator Output active power, express Moment Power Plant Active load, For the aggregated grid The number of contact points Indicates the total number of generators in the contact station, express Time contact station Active load, express Aggregate the power grid at all times The total active line loss is the sum of the active line loss of the plant branch and the line. express Time outflow aggregation grid The reactive power, express Time generator Output reactive power, express Moment Power Plant The reactive power output of the reactive compensation equipment can be switched on and off. express Moment Power Plant The reactive load, express Time contact station The reactive power output of the reactive compensation equipment can be switched on and off. express Time contact station The reactive load, express Aggregate the power grid at all times The total reactive line loss is the sum of the reactive line losses of the plant branch and the line; The initial aggregation grid model of connectivity is: , where express Time outflow aggregation grid The active power, express Flowing into the aggregated grid The active power, For the aggregated grid The number of power plants, Indicates power plant Total number of generators, express Time generator Output active power, express Moment Power Plant Active load, For the aggregated grid The number of contact points Indicates the total number of generators in the contact station, express Time contact station Active load, express The total active line loss of the aggregated power grid at any given moment is the sum of the active line losses of the plant branch and the line. express Time outflow aggregation grid The reactive power, express Flowing into the aggregated grid The reactive power, express Time generator Output reactive power, express Moment Power Plant The reactive power output of the reactive compensation equipment can be switched on and off. express Moment Power Plant The reactive load, For the aggregated grid The number of contact points express Time contact station The reactive power output of the reactive compensation equipment can be switched on and off. express Time contact station The reactive load, express Aggregate the power grid at all times The total reactive line loss is the sum of the reactive line losses of the plant branch and the line.

2. A modeling device for a deep aggregation power grid model, characterized in that: Include: A system information acquisition module, used to acquire system information of the power system; wherein the system information includes plant information; A plant model acquisition module is used to construct an initial plant model according to the plant information; wherein the initial plant model includes an initial model of a power plant and an initial model of a connection station; A flow data acquisition module is used to acquire the flow data of the power plant; A plant model solving module is used to solve the unknown quantity of the initial plant model based on the long short-term memory network according to the power plant's flow data, and obtain an aggregated plant model; wherein the aggregated plant model includes a power plant aggregation model and a tie station aggregation model; The initial model of the power plant is , where express Moment Power Plant The output active power, Indicates power plant Total number of generators, express Time generator Output active power, express Moment Power Plant Active load, express Moment Power Plant Active line loss of branch circuit, express Moment Power Plant Reactive power output to the outside, express Time generator Output reactive power, express Moment Power Plant The reactive power output of the reactive compensation equipment can be switched on and off. express Moment Power Plant The reactive load, express Moment Power Plant Reactive line loss of the branch circuit; The initial model of the contact station is , where express Time outflow contact station The active power, express Time flow into the contact station The active power, Indicates the total number of generators in the contact station, express Time generator Output active power, express Time contact station Active load, express Time contact station Active line loss of branch circuit, express Time outflow contact station The reactive power, express Time flow into the contact station The reactive power, express Time generator Output reactive power, express Time contact station The reactive power output of the reactive compensation equipment can be switched on and off. express Time contact station The reactive load, express Time contact station Reactive line loss of the branch circuit; The plant information includes the number of generators in the power plant; the unknown quantity of the initial plant model is the active line loss and reactive line loss ; The plant model solving module includes: The solution model acquisition unit is used to obtain the active line loss according to the initial plant model. and reactive line loss The solution model: In the formula, represents the active power column vector of the generator node, represents the reactive power column vector of the generator node, express The first-order derivative of express The second-order derivative of express The first-order derivative of express The second-order derivative of express The first-order derivative of express The second-order derivative of express The first-order derivative of express The second-order derivative of Indicated by is a complex coefficient column vector consisting of elements, represents the identity matrix, , represents the linear complex coefficient, Indicates the total number of load nodes in the aggregation, Represents the total number of generator nodes in the aggregation, represents the node impedance matrix within the aggregation, Indicated by The column vector composed of express The first-order derivative of Indicates transposition, represents a conjugate complex number; A flow data acquisition unit, used to acquire flow data of a power plant; wherein the flow data includes N historical event sections; The training set acquisition unit is used to obtain the active line loss according to the power flow data of the power plant. and reactive line loss The solution model of The square of the active power of each generator, the combination of the active power of each generator, the square of the reactive power of each generator, the combination of reactive power, the combination of active and reactive power, the active power of each generator, and the reactive power of each generator are used as inputs; the corresponding active line loss and reactive line loss are used as outputs to form a training set; The plant model solving unit is used to input the input data in the training set into the LSTM network to obtain the predicted value, and use the back propagation algorithm to learn and train the LSTM weights according to the error between the predicted value and the true value to obtain the active line loss and reactive line loss The coefficients of the solved model are obtained to obtain the aggregated plant model; The system information also includes line information; wherein the line information includes voltage level; The modeling device also includes: The aggregated grid construction module is used to construct an aggregated grid according to the line information; wherein the aggregated grid model includes a source aggregated grid and a connection aggregated grid; A power grid model building module is used to build an initial aggregated power grid model according to the aggregated power grid model; wherein the initial aggregated power grid model includes a source initial aggregated power grid model and a connection initial aggregated power grid model; A power grid data acquisition module, used to obtain power flow data of the aggregated power grid; A power grid model solving module is used to solve the unknown quantities of the initial aggregated power grid model based on the long short-term memory network according to the power flow data of the aggregated power grid, and obtain a deep aggregated power grid model; wherein the deep aggregated power grid model includes a source deep aggregated power grid model and a connection deep aggregated power grid model; The source initial aggregation grid model is , where express Time outflow aggregation grid The active power, For the aggregated grid The number of power plants, Indicates power plant Total number of generators, express Time generator Output active power, express Moment Power Plant Active load, For the aggregated grid The number of contact points Indicates the total number of generators in the contact station, express Time contact station Active load, express Aggregate the power grid at all times The total active line loss is the sum of the active line loss of the plant branch and the line. express Time outflow aggregation grid The reactive power, express Time generator Output reactive power, express Moment Power Plant The reactive power output of the reactive compensation equipment can be switched on and off. express Moment Power Plant The reactive load, express Time contact station The reactive power output of the reactive compensation equipment can be switched on and off. express Time contact station The reactive load, express Aggregate the power grid at all times The total reactive line loss is the sum of the reactive line losses of the plant branch and the line; The initial aggregation grid model of connectivity is: , where express Time outflow aggregation grid The active power, express Flowing into the aggregated grid The active power, For the aggregated grid The number of power plants, Indicates power plant Total number of generators, express Time generator Output active power, express Moment Power Plant Active load, For the aggregated grid The number of contact points Indicates the total number of generators in the contact station, express Time contact station Active load, express The total active line loss of the aggregated power grid at any given moment is the sum of the active line losses of the plant branch and the line. express Time outflow aggregation grid The reactive power, express Flowing into the aggregated grid The reactive power, express Time generator Output reactive power, express Moment Power Plant The reactive power output of the reactive compensation equipment can be switched on and off. express Moment Power Plant The reactive load, For the aggregated grid The number of contact points express Time contact station The reactive power output of the reactive compensation equipment can be switched on and off. express Time contact station The reactive load, express Aggregate the power grid at all times The total reactive line loss is the sum of the reactive line losses of the plant branch and the line.

3. A modeling device for a deep aggregated power grid model, characterized in that: It includes a processor, a memory, and a computer program stored in the memory; the computer program can be executed by the processor to implement the modeling method of the deep aggregation power grid model as claimed in claim 1.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the modeling method of the deep aggregation power grid model as claimed in claim 1.

Images