Active power distribution network voltage control method and device adaptive to topology change

Abstract

The invention discloses an active power distribution network voltage control method and an active power distribution network voltage control device adaptive to topological change, which are used for solving the problem that a voltage control strategy cannot be output due to ineffective modeling caused by incapability of well adapting to topological change in the prior art. The active power distribution network comprises a first voltage level power distribution network and a plurality of second voltage level power distribution networks; constructing a voltage control model of the active power distribution network according to a coupling relationship between the first voltage level power distribution network and the plurality of second voltage level power distribution networks; carrying out optimization solution on the voltage control model, and taking a solved boundary variable optimal solution as a voltage control strategy when power grid dispatching is carried out on the first voltage level power distribution network and the plurality of second voltage level power distribution networks; and in the optimization solution process, performing power flow model updating based on the element matrix for the target second voltage level power distribution network with topological change to obtain a new linear power flow model, and performing power flow calculation of the target second voltage level power distribution network based on the new linear power flow model.

Description

Technical Field

[0001] The present invention relates to the technical field of power system operation, and in particular, to an active distribution network voltage control method adaptable to topological changes, an active distribution network voltage control device adaptable to topological changes, an electronic device, and a storage medium. Background Art

[0002] With the increasing popularity of Renewable Energy Sources (RES) in the distribution network, large-scale changes in the Power Flow (PF) have reshaped the operating characteristics of the power grid. Especially under extreme weather conditions, sudden changes in the power of renewable energy can extend the influence range of the power flow beyond the distribution network, thus increasing the risks of system overload and overvoltage. In addition, extreme weather may also cause active or passive changes in the network topology, resulting in large-scale power flow transfer. Therefore, the voltage control of the active distribution network under extreme weather conditions is crucial for the power grid with a high proportion of renewable energy.

[0003] Currently, distributed computing methods can be used for the coordination between high-voltage distribution networks and medium- and low-voltage distribution networks. In the distributed iterative optimization method of the active distribution network, various decoupling methods have also been widely applied. However, traditional physics-based models rely heavily on accurate physical parameters, which are difficult to obtain in medium- and low-voltage distribution networks, making it difficult to effectively carry out the coordinated optimization between high-voltage distribution networks and distribution networks.

[0004] Data-driven methods have become effective solutions to problems such as inaccurate parameters in the distribution network. For example, a data-driven voltage control strategy using a convolutional neural network is used to solve the voltage violation problem in the distribution network. Another example is to adopt a deep reinforcement learning method to optimize the day-ahead scheduling strategy of the Active Distribution Network (ADN) to maintain safe and efficient operation. However, the optimization models established by these methods are all black-box models and lack interpretability of the physical mechanism. Especially under extreme weather conditions, the topology of the active distribution network may undergo active or passive reconfiguration, and due to the lack of operation data in the new topology, the above modeling methods will become invalid. Summary of the Invention

[0005] The present invention provides an active distribution network voltage control method adaptable to topological changes, an active distribution network voltage control device adaptable to topological changes, an electronic device, and a storage medium, which are used to solve or partially solve the technical problem that the current active distribution network voltage control method cannot well adapt to topological changes, resulting in invalid voltage control modeling and thus unable to output a voltage control strategy.

[0006] The present invention provides a method for voltage control of an active distribution network adaptable to topological changes. The active distribution network includes a first-voltage-level distribution network and a plurality of second-voltage-level distribution networks. The method includes:

[0007] Construct a voltage control model of the active distribution network according to the coupling relationship between the first-voltage-level distribution network and the plurality of second-voltage-level distribution networks;

[0008] Optimize and solve the voltage control model, and use the optimal solution of the boundary variables obtained by the solution as the voltage control strategy when performing power grid dispatching on the first-voltage-level distribution network and the plurality of second-voltage-level distribution networks;

[0009] During the optimization and solution process, update the power flow model based on the meta-matrix for the target second-voltage-level distribution network with topological changes to obtain a new linear power flow model, and perform power flow calculation on the target second-voltage-level distribution network based on the new linear power flow model.

[0010] Optionally, the constructing a voltage control model of the active distribution network according to the coupling relationship between the first-voltage-level distribution network and the plurality of second-voltage-level distribution networks includes:

[0011] Construct a first objective function of the first-voltage-level distribution network with the goal of minimizing the network active power loss cost, the action cost of discrete voltage regulating equipment, and the regulation cost of on-load tap-changer;

[0012] Based on the first objective function and the pre-constructed first multi-constraint conditions, construct a first voltage optimization model of the first-voltage-level distribution network;

[0013] For each of the second-voltage-level distribution networks, considering the sudden drop in the output of a high proportion of distributed power sources and the wide-area power flow changes caused by topological reconstruction under extreme weather conditions, construct a second objective function of the second-voltage-level distribution network;

[0014] Based on the second objective function and the pre-constructed second multi-constraint conditions, construct a second voltage optimization model of the second-voltage-level distribution network;

[0015] Considering the coupling relationship between the first-voltage-level distribution network and the plurality of second-voltage-level distribution networks, construct boundary constraints between the first-voltage-level distribution network and each of the second-voltage-level distribution networks respectively;

[0016] Take the first voltage optimization model, each of the second voltage optimization models, and each of the boundary constraints as the voltage control model of the active distribution network.

[0017] Optionally, each of the second-voltage-level distribution networks corresponds to a pre-trained meta-matrix; the construction process of the meta-matrix includes:

[0018] Obtain the historical topology information graph and the historical linear power flow model of the second-voltage-level distribution network;

[0019] Construct a historical adjacency matrix according to the historical topology information graph data;

[0020] Construct a retraining sample set and a meta-matrix mapping relationship model according to the historical linear power flow model and the historical adjacency matrix;

[0021] Substitute the retraining sample set into the meta-matrix mapping relationship model, and solve the meta-matrix mapping relationship model after substituting the retraining sample set according to the least squares regression and the generalized inverse matrix to obtain the meta-matrix of the second-voltage-level distribution network.

[0022] Optionally, the updating of the power flow model based on the meta-matrix for the target second-voltage-level distribution network with topological changes to obtain a new linear power flow model includes:

[0023] Obtain the new adjacency matrix of the target second-voltage-level distribution network with topological changes;

[0024] Extract the column vectors of the adjacency matrix from the new adjacency matrix, and substitute the column vectors of the adjacency matrix into the meta-matrix mapping relationship model to solve for the column vectors of the linear power flow matrix;

[0025] Re-arrange the matrix elements of the column vectors of the linear power flow matrix to obtain the new linear power flow model of the target second-voltage-level distribution network under the new topology.

[0026] Optionally, the method further includes:

[0027] Obtain the historical operation data of the second-voltage-level distribution network;

[0028] Construct the historical linear power flow model of the second-voltage-level distribution network through state space mapping according to the historical operation data.

[0029] Optionally, the historical operation data includes the distribution network node voltage, distributed power source power, energy storage power, load power, and distribution network supply power; the constructing of the historical linear power flow model of the second-voltage-level distribution network through state space mapping according to the historical operation data includes:

[0030] Construct a net power model according to the distributed power source power, the energy storage power, and the load power;

[0031] According to the net power model, construct a plurality of input samples, respectively perform state - space dimensionality elevation on each of the input samples to obtain corresponding dimensionality - elevated input samples, and construct a dimensionality - elevated input sample set based on each of the dimensionality - elevated input samples;

[0032] According to the node voltage of the distribution network and the power supplied by the distribution network, construct a plurality of output samples, and construct an output sample set based on each of the output samples;

[0033] In combination with least - squares data - driven training, construct the historical linear power flow model of the second - voltage - level distribution network according to the dimensionality - elevated input sample set and the output sample set.

[0034] Optionally, the optimization solution of the voltage control model includes:

[0035] Take the first - voltage optimization model as the main problem, take each of the second - voltage optimization models as sub - problems, consider each of the boundary constraints, and perform multi - round optimization solutions on the main problem and each of the sub - problems respectively through generalized Benders decomposition. After each round of solution, exchange the optimized solutions of the boundary variables between the first - voltage - level distribution network and each of the second - voltage - level distribution networks until the optimal solution of the boundary variables is obtained.

[0036] The present invention also provides an active distribution network voltage control device with adaptive topology change. The active distribution network includes a first - voltage - level distribution network and a plurality of second - voltage - level distribution networks; the device includes:

[0037] A voltage control model construction unit for constructing the voltage control model of the active distribution network according to the coupling relationship between the first - voltage - level distribution network and the plurality of second - voltage - level distribution networks;

[0038] An optimization solution unit for performing optimization solution on the voltage control model and taking the obtained optimal solution of the boundary variables as the voltage control strategy when performing power grid scheduling on the first - voltage - level distribution network and the plurality of second - voltage - level distribution networks;

[0039] A power flow model update unit for, during the optimization solution process, performing power flow model update based on the meta - matrix for the target second - voltage - level distribution network with topology change to obtain a new linear power flow model, and performing power flow calculation on the target second - voltage - level distribution network based on the new linear power flow model.

[0040] The present invention also provides an electronic device, and the device includes a processor and a memory:

[0041] The memory is used to store program code and transmit the program code to the processor;

[0042] The processor is configured to execute the active distribution network voltage control method with adaptive topology change according to the instructions in the program code as described in any one of the above.

[0043] The present invention also provides a computer-readable storage medium for storing program code for executing the active distribution network voltage control method with adaptive topology change according to any one of the above.

[0044] It can be seen from the above technical solutions that the present invention has the following advantages:

[0045] There is provided an active distribution network voltage control method with adaptive topology change. The active distribution network includes a first-voltage-level distribution network and a plurality of second-voltage-level distribution networks; first, a voltage control model of the active distribution network is constructed according to the coupling relationship between the first-voltage-level distribution network and the plurality of second-voltage-level distribution networks for subsequent optimization and solution of model problems; then, the voltage control model is optimized and solved, and the optimal solution of the boundary variables obtained by the solution is used as the voltage control strategy during the power grid dispatching of the first-voltage-level distribution network and the plurality of second-voltage-level distribution networks. Thus, through the optimization and solution of the constructed model, the optimal solution of the model can be quickly obtained and used as the voltage control strategy in the optimal case for subsequent power grid dispatching of each distribution network to achieve the global voltage control of the active distribution network; during the optimization and solution process, for the target second-voltage-level distribution network with topology change, the power flow model is updated based on the meta-matrix to obtain a new linear power flow model, and the power flow calculation of the target second-voltage-level distribution network is performed based on the new linear power flow model. Thus, during the solution process, even if the topology of the distribution network changes, the power flow model is updated based on the meta-matrix, and the linearized power flow model under the new topology can be obtained without training data. On the one hand, it can avoid the problems of power flow model failure and calculation failure caused by topology change, and on the other hand, it can avoid retraining the power flow model of the distribution network while maintaining the safe operation of the distribution network voltage, greatly saving computing resources and training time. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0047] Figure 1 It is a flowchart of the steps of an active distribution network voltage control method with adaptive topology change;

[0048] Figure 2 Schematic diagram of an adaptive dimension-augmented power flow architecture for a medium- and low-voltage distribution network

[0049] Figure 3 Overall flowchart of a voltage control method for an active distribution network with adaptive topology changes

[0050] Figure 4 Schematic diagram of the topology structure of a high-voltage distribution network and a medium- and low-voltage distribution network

[0051] Figure 5 Visualization diagram of the actual voltage of a medium- and low-voltage distribution network and the voltage distribution obtained after adaptive dimension augmentation after topology reconstruction caused by extreme weather

[0052] Figure 6 Visualization diagram of the effects before and after voltage optimization of a medium- and low-voltage distribution network after topology reconstruction caused by extreme weather

[0053] Figure 7 Structure block diagram of a voltage control device for an active distribution network with adaptive topology changes Specific implementation manners

[0054] The embodiments of the present invention provide a voltage control method for an active distribution network with adaptive topology changes, a voltage control device for an active distribution network with adaptive topology changes, an electronic device, and a storage medium, which are used to solve or partially solve the technical problem that the current voltage control method for an active distribution network cannot well adapt to topology changes, resulting in invalid voltage control modeling and thus unable to output a voltage control strategy

[0055] To make the invention purpose, features, and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention

[0056] As an example, the current coordination of a high-voltage distribution network and a medium- and low-voltage distribution network can adopt a distributed computing method. In the distributed iterative optimization method for an active distribution network, various decoupling methods have also been widely used. However, traditional physics-based models highly rely on accurate physical parameters, which are difficult to obtain in a medium- and low-voltage distribution network, making it difficult to effectively carry out the coordinated optimization between a high-voltage distribution network and a distribution network

[0057] Data-driven methods have become effective solutions to problems such as inaccurate parameters in distribution networks. For example, a data-driven voltage control strategy using a convolutional neural network is used to solve voltage violation problems in distribution networks. Another example is the use of deep reinforcement learning methods to optimize the day-ahead scheduling strategy of active distribution networks to maintain safe and efficient operation. However, the optimization models established by these methods are all black-box models and lack interpretability of physical mechanisms. Especially under extreme weather conditions, the topology of the active distribution network may undergo active or passive reconfiguration, and due to the lack of operation data in the new topology, the above modeling methods will become ineffective.

[0058] Therefore, one of the core inventive points of the embodiments of the present invention is: to propose an active distribution network voltage control method that adapts to topological changes in the medium- and low-voltage distribution network under a new topology. First, comprehensively consider the coupling relationship between the high-voltage distribution network and the medium- and low-voltage distribution network, and construct a voltage control model for the active distribution network. Secondly, according to the historical operation data of the medium- and low-voltage distribution network, through a data-driven method, based on the Koopman data-driven theory, the traditional non-linear power flow model is transformed into a high-dimensional linear power flow model that does not depend on impedance parameters. Aiming at the problem that the original power flow model fails due to topological changes in the medium- and low-voltage distribution network, an adaptive dimension augmentation power flow method is used to train the meta-matrix to describe the mapping relationship between the topology (adjacency matrix) and the linearized power flow model. Thus, through the adaptive dimension augmentation power flow considering topological switching, global PF sensitivity can be generated under the new topology, and the linearized power flow matrix is updated through the meta-matrix and the new adjacency matrix. Finally, based on the differential characteristics of the high-voltage distribution network and the medium- and low-voltage distribution network, generalized Benders decomposition (GBD) is used for coordinated solution to achieve global voltage control of the active distribution network.

[0059] By adopting the technical solution of the present invention, on the one hand, the accuracy of power flow calculation does not depend on the static model parameters of the distribution network. On the other hand, when the topology changes, the linearized power flow model under the new topology can be obtained without training data, so that while maintaining the safe operation of the distribution network voltage, the retraining of the distribution network power flow model can be avoided, greatly saving computing resources and training time. In addition, the generalized Benders decomposition, a distributed cooperative optimization algorithm adopted by the present invention, can quickly converge to the global optimal solution after a limited number of information exchanges, and is particularly suitable for solving voltage safety operation problems caused by sudden changes in distributed power sources and topological changes in the distribution network due to extreme weather.

[0060] Refer to Figure 1, showing the step flowchart of an active distribution network voltage control method with adaptive topology change provided by an embodiment of the present invention. The active distribution network includes a first voltage level distribution network and multiple second voltage level distribution networks; the method may specifically include the following steps:

[0061] Step 101, construct a voltage control model of the active distribution network according to the coupling relationship between the first voltage level distribution network and the multiple second voltage level distribution networks;

[0062] In practical applications, the first voltage level distribution network may be a high-voltage distribution network. The second voltage level distribution network may be a medium- and low-voltage distribution network. In this step, the coupling relationship between the high-voltage distribution network and the medium- and low-voltage distribution networks is mainly considered to establish a voltage control model of the active distribution network.

[0063] In some embodiments, the execution process of constructing a voltage control model of the active distribution network according to the coupling relationship between the first voltage level distribution network and the multiple second voltage level distribution networks includes the following sub-steps S01 to S06:

[0064] Step S01: Construct a first objective function of the first voltage level distribution network with the goal of minimizing the network active power loss cost, the discrete voltage regulation device action cost, and the on-load tap changer regulation cost;

[0065] Specifically, in the voltage optimization model of the high-voltage distribution network, its objective function is to minimize the network active power loss cost, the discrete voltage regulation device action cost, and the on-load tap changer regulation cost. To distinguish it from the objective function of the medium- and low-voltage distribution network, it is defined here as the first objective function, expressed as:

[0066] (1)

[0067] Where, is the set of lines representing the high-voltage distribution network; is the set of lines including OLTC (On-Load Tap Changer) lines; is the set of shunt capacitor banks; is the line current; is the line resistance value; is the tap position of the OLTC of the line , is its initial value; is the number of shunt capacitor access groups at node , is its initial value, , and respectively represent the active feed-in tariff, the cost of single-step regulation of the OLTC tap, and the cost of switching a single group of capacitors.

[0068] Step S02: Based on the first objective function and the pre-constructed first multi-constraint conditions, construct the first voltage optimization model for the first voltage level distribution network;

[0069] To distinguish from the multi-constraint conditions of the medium and low voltage distribution network, the multiple constraint conditions that the high voltage distribution network needs to follow are defined as the first multi-constraint conditions here. Specifically, the high voltage distribution network should follow the multiple constraint conditions shown in the following formulas (2)-(8):

[0070] (2)

[0071] (3)

[0072] (4)

[0073] (5)

[0074] (6)

[0075] (7)

[0076] (8)

[0077] Among them, is the set of SVC (Static Var Compensator); 、 respectively represent the SVC reactive power output power and its upper limit value of node ; is the reactive power capacity of a single group of capacitors at node ; is the upper limit of the number of capacitor groups put into operation at node ; is the adjustable transformation ratio of the OLTC on line ; 、 are respectively the standard transformation ratio and the regulation step of the OLTC in line ; is the regulation gear of the OLTC in line ; 、 are respectively the maximum gears for the OLTC to increase and decrease in line ; is the set of generators in the high voltage distribution network; and respectively represent the minimum active power and reactive power of the generator units in the high-voltage distribution network ; and respectively represent the maximum active power and reactive power of the generator node units ; and respectively represent the active power and reactive power flowing from node to node ; and respectively represent the active power and reactive power flowing from node to the sub-node ; is the set of nodes in the high-voltage distribution network; and respectively represent the resistance and reactance values of the line ; is the voltage amplitude at node ; is the voltage amplitude at node ; and are respectively the lower limit and upper limit of the square of the voltage at node ; and respectively represent the sets of the parent node and sub-nodes of node ; and represent the active power and reactive power at node ; represents the output reactive power of the capacitor bank at node .

[0078] Step S03: For each second-voltage-level distribution network, considering simultaneously the sudden drop in the output of a high proportion of distributed power sources and the wide-area power flow changes caused by topological reconstruction under extreme weather conditions, construct the second objective function of the second-voltage-level distribution network;

[0079] Specifically, in the voltage control model of the medium- and low-voltage distribution network, for the sudden drop in the output of a high proportion of distributed power sources and the wide-area power flow changes caused by topological reconstruction under extreme weather conditions, the objective function of voltage control (defined here as the second objective function to distinguish it from the objective function of the high-voltage distribution network) can be expressed as the following formula (9):

[0080] (9)

[0081] Wherein, is the set of nodes in the medium- and low-voltage distribution network; represents the voltage reference value; is the set of DG (Distributed Generation) nodes in the medium- and low-voltage distribution network; is the set of load nodes in the medium- and low-voltage distribution network; and respectively represent the adjustment powers of DG and load; 、 、 respectively represent the penalty coefficients of voltage deviation, DG curtailment, and load curtailment power.

[0082] Step S04: Based on the second objective function and the pre-constructed second multi-constraint conditions, construct the second voltage optimization model for the second voltage level distribution network;

[0083] To distinguish from the multi-constraint conditions of the high-voltage distribution network, the multiple constraint conditions that the medium- and low-voltage distribution network needs to follow are defined as the second multi-constraint conditions here. Specifically, the medium- and low-voltage distribution network should follow the multiple constraint conditions shown in the following equations (10)-(15):

[0084] (10)

[0085] (11)

[0086] (12)

[0087] (13)

[0088] (14)

[0089] (15)

[0090] Among them, represents the set of lines in the medium- and low-voltage distribution network; represents the voltage magnitude at node ; and here represent the lower and upper limits of the voltage magnitude respectively; represents the current magnitude on line ; and respectively represent the resistance and reactance of line ; and respectively represent the active power and reactive power at node ; 、 and respectively represent the active powers of DG, energy storage, and load at node ; , and respectively represent the reactive power of DG, energy storage, and load at node . and are the initial values of the active / reactive power injection of the load at node . and respectively represent the adjustment amounts of the active power and reactive power of the load at node . represents the power factor angle at the load node . is the set of nodes of the medium - low voltage distribution network energy storage system (ESS). and are the initial values of the active / reactive power injection of the energy storage at node . and respectively represent the adjustment amounts of the active power and reactive power of the energy storage at node . represents the rated capacity of the energy storage. and are the active / reactive power of DG at node . represents the rated capacity of DG.

[0091] Step S05: Considering the coupling relationship between the first - level voltage distribution network and multiple second - level voltage distribution networks, construct the boundary constraints between the first - level voltage distribution network and each second - level voltage distribution network respectively.

[0092] The following boundary constraints should be satisfied between the high - voltage distribution network and the medium - low voltage distribution network:

[0093] (16)

[0094] Among them, represents the voltage amplitude at the connection point between the high - voltage distribution network and the medium - low voltage distribution network . represents the node voltage amplitude of the root node of the medium - low voltage distribution network .

[0095] Step S06: Take the first - level voltage optimization model, each second - level voltage optimization model, and each boundary constraint as the voltage control model of the active distribution network.

[0096] The voltage control model of the active distribution network can be constructed through steps S01 to S05. Among them, equations (1)-(8) constitute the complete voltage emergency control model of the high-voltage distribution network. Equations (9)-(15) together with equation (31) constructed in the subsequent steps jointly constitute the complete voltage emergency control model of the medium- and low-voltage distribution network. Equation (16) describes the consistency condition of the interaction variables between the high-voltage distribution network and the medium- and low-voltage distribution network.

[0097] Based on the construction of the voltage control model of the above-mentioned active distribution network, in the embodiment of the present invention, in the medium- and low-voltage distribution network, the traditional non-linear power flow model is transformed into a high-dimensional linear model independent of impedance parameters based on the Koopman operator theory, and after the topology of the distribution network changes due to extreme weather, a new corresponding power flow model can be quickly generated.

[0098] In some embodiments, the historical operation data of the second voltage level distribution network (medium- and low-voltage distribution network) can be obtained first; then, according to the historical operation data, the historical linear power flow model of the second voltage level distribution network is constructed through state space mapping.

[0099] Further, each medium- and low-voltage distribution network collects data samples. Taking the th medium- and low-voltage distribution network as an example, the historical operation data of the medium- and low-voltage distribution network mainly includes the node voltage of the distribution network , the power of distributed power sources (DG active power and reactive power ), the power of energy storage (energy storage active power and reactive power ), the load power (load active power and reactive power ), and the supply power of the distribution network (the active power supplied to the th distribution network and reactive power ). Specifically, based on the operation data in the historical database of the medium- and low-voltage distribution network control center (including the root node voltage of the distribution network, DG, energy storage, and load data), through state space mapping and least squares method training, the node voltage , DG active power and reactive power , energy storage active power and reactive power , load active power and reactive power in each medium- and low-voltage distribution network are constructed offline through a high-dimensional linear state space mapping model.

[0100] In a specific implementation, according to historical operation data, the execution process of constructing a historical linear power flow model for a secondary voltage level distribution network through state space mapping includes the following sub-steps S11 to S14:

[0101] Step S11: Construct a net power model based on distributed power generation power, energy storage power, and load power;

[0102] The net power model (including net active power and reactive power ) in the i-th medium and low voltage distribution network is shown in the following formula (17):

[0103] (17)

[0104] Wherein, , and respectively represent the position matrices of the load, energy storage, and distributed power generation DG in the active distribution network.

[0105] Step S12: Based on the net power model, construct multiple input samples, perform state space dimensionality elevation on each input sample respectively to obtain corresponding elevated input samples, and construct an elevated input sample set based on each elevated input sample;

[0106] Taking the i-th sample as an example, the input sample can be defined as:

[0107] (18)

[0108] According to formula (18), the state space dimensionality elevation of the input sample can be expressed as:

[0109] (19)

[0110] Wherein, represents the augmented input variable of the j-th data sample in the i-th medium and low voltage distribution network. Assuming the dimension of the augmented input variable is , then can be defined as: , then can be defined as:

[0111] (20)

[0112] can be defined as:

[0113] (21)

[0114] Wherein, represents the one related to Base vectors of the same dimension; Representing the dimensional augmented input variable; Indicating the th medium - low voltage distribution network th data sample's th Euclidean distance; Indicating the th medium - low voltage distribution network th input sample.

[0115] The data - driven matrix can be obtained through data - driven training with historical samples in the medium - low voltage distribution network. Among them, the th set of input samples after dimensionality increase of the distribution network is:

[0116] (22)

[0117] Step S13: According to the node voltage and supply power of the distribution network, construct multiple output samples, and based on each output sample, construct an output sample set;

[0118] Still taking the th sample as an example, the output sample can be defined as:

[0119] (23)

[0120] The th output sample set of the medium - low voltage distribution network is:

[0121] (24)

[0122] Step S14: Combining least - squares data - driven training, according to the set of input samples after dimensionality increase and the output sample set, construct the historical linear power flow model of the second - level voltage distribution network.

[0123] Specifically, through least - squares data - driven training, construct the net active power and reactive power supplied to the medium - low voltage distribution network, voltage and the high - dimensional linear model between the net injected active power and the net injected reactive power at the nodes. The linearized power flow model (i.e., linearized power flow matrix or linear power flow matrix)

[0124] (25)

[0125] Among them, represents matrix transpose; represents matrix pseudo-inverse. Therefore, the linear function relationship of the th medium and low voltage distribution network can be defined as:

[0126] (26)

[0127] In a medium and low voltage distribution network with incomplete model parameters, the high-precision global linearized power flow equation can be obtained based on historical data samples through the above method.

[0128] On this basis, a more accurate sensitivity coefficient can be further derived as in:

[0129] (27)

[0130] (28)

[0131] (29)

[0132] where is the total number of nodes of the th medium and low voltage distribution network (i.e., the active distribution network ADN); , and respectively represent the elements corresponding to the output variables in , , and the input variable ; , and respectively correspond to the output variables , , and the th dimensionality-increased variable . The partial differential terms on the right side of equations (25)-(27) can be expressed as follows:

[0133] (30)

[0134] where represents the th -dimensional th basis element in the th medium and low voltage distribution network; is the th input variable in the th medium and low voltage distribution network; The th base element in the -dimensional medium and low voltage distribution network;

[0135] Based on the above derivation, the node voltage , the active power and reactive power supplied to the medium and low voltage distribution network sensitivity coefficient matrix , , .

[0136] In summary, the linear power flow equation and security constraints of the medium and low voltage distribution network can be linearly expressed as:

[0137] (31)

[0138] The embodiment of the present invention also derives the meta-matrix representing the mapping relationship between the topological graph data and the linear power flow model corresponding to each second voltage level distribution network (medium and low voltage distribution network) based on the adaptive dimension augmented power flow method. That is to say, each second voltage level distribution network can correspond to a pre-trained meta-matrix.

[0139] Figure 2 shows a schematic diagram of the adaptive dimension augmented power flow architecture of the medium and low voltage distribution network provided by the embodiment of the present invention. Combining Figure 2 , in some embodiments, the construction process of the meta-matrix includes the following sub-steps S21 to S24:

[0140] Step S21: Obtain the historical topological information graph and historical linear power flow model of the second voltage level distribution network;

[0141] For a medium and low voltage distribution network with nodes, assume the historical topological information graph (i.e., the graph data representing historical topological information) is . Among them, represents the set of buses, represents the undirected edges connecting adjacent buses in the medium and low voltage distribution network.

[0142] Step S22: Construct a historical adjacency matrix according to the historical topological information graph data;

[0143] The historical adjacency matrix (historical topological matrix) of the medium and low voltage distribution network

[0144] (32)

[0145] Step S23: Construct a retraining sample set and a meta-matrix mapping relationship model based on the historical linear power flow model and the historical adjacency matrix;

[0146] In the offline training stage, the meta-matrix can be used to formulate the adjacency matrix and the corresponding linear power flow model to establish the mapping relationship between them. The constructed meta-matrix mapping relationship model can be expressed as:

[0147] (33)

[0148] Wherein, represents the meta-matrix; and are the column vectors of the adjacency matrix and the linear power flow model respectively, that is:

[0149] (34)

[0150] represents the conversion of to the column vector in the form of .

[0151] The meta-matrix can be obtained by retraining with historical samples in the existing topology. The retraining sample set can be defined as:

[0152] (35)

[0153] Wherein, , represent the sample sets of the integer topology vector and the linear power flow vector respectively; represents the number of retraining samples.

[0154] Step S24: Substitute the retraining sample set into the meta-matrix mapping relationship model, and solve the meta-matrix mapping relationship model after substituting the retraining sample set according to the least squares regression and the generalized inverse matrix to obtain the meta-matrix of the secondary voltage level distribution network.

[0155] Specifically, the meta-matrix can be determined by least squares regression, and it can be expressed as:

[0156] (36)

[0157] By using the Moore-Penrose pseudoinverse of the generalized inverse matrix, It can be expressed as:

[0158] (37)

[0159] Step 102: Optimally solve the voltage control model, and use the optimal solution of the boundary variables obtained by the solution as the voltage control strategy when performing power grid scheduling for the first-level voltage distribution network and the multiple second-level voltage distribution networks.

[0160] In the embodiment of the present invention, through the generalized Benders decomposition algorithm, the high-voltage distribution network and the medium- and low-voltage distribution networks independently solve the master problem and the sub-problems respectively. Finally, the optimal solution of the boundary variables in the active distribution network optimization problem obtained by the solution is used as the strategy for scheduling the high-voltage distribution network and the medium- and low-voltage distribution networks. That is to say, the master problem corresponds to the problem of the high-voltage distribution network, that is, to solve the voltage optimization model of the high-voltage distribution network. The sub-problems correspond to the problems of each medium- and low-voltage distribution network, and the voltage optimization models of each medium- and low-voltage distribution network are solved respectively. The high-voltage distribution network and each medium-voltage distribution network independently solve the master problem and the sub-problems respectively, and transmit the optimization parameters of the boundary variables to the other party after obtaining the optimization solution in each round.

[0161] In a specific implementation, optimally solving the voltage control model includes: taking the first voltage optimization model as the master problem, taking each second voltage optimization model as the sub-problem respectively, considering each boundary constraint, and performing multi-round optimal solutions for the master problem and each sub-problem respectively through the generalized Benders decomposition. After the solution of each round is completed, exchange the optimal solutions of the boundary variables between the first-level voltage distribution network and each second-level voltage distribution network until the optimal solution of the boundary variables is obtained.

[0162] Step 103: During the optimization solution process, update the power flow model based on the meta-matrix for the target second-level voltage distribution network with topological changes to obtain a new linear power flow model, and perform power flow calculation for the target second-level voltage distribution network based on the new linear power flow model.

[0163] Combined with Figure 2 , in some embodiments, the execution process of updating the power flow model based on the meta-matrix for the target second-level voltage distribution network with topological changes to obtain a new linear power flow model includes the following sub-steps S31 to S33:

[0164] Step S31: Obtain the new adjacency matrix of the target second-level voltage distribution network with topological changes;

[0165] Step S32: Extract the column vectors of the adjacency matrix from the new adjacency matrix, and substitute the column vectors of the adjacency matrix into the meta-matrix mapping relationship model to solve the column vectors of the linear power flow matrix;

[0166] Given the column vectors of the new adjacency matrix of the linear power flow model under the new topology the column vectors of the linear power flow matrix can be updated according to the following formula:

[0167] (38)

[0168] Step S33: Rearrange the matrix elements of the column vectors of the linear power flow matrix to obtain a new linear power flow model of the target secondary voltage level distribution network under the new topology.

[0169] The linear power flow model of the medium and low voltage distribution network under the new topology is shown in the following formula:

[0170] (39)

[0171] In the embodiment of the present invention, a voltage control method for an active distribution network with adaptive topology change is proposed. First, considering the coupling relationship between the high-voltage distribution network and the medium and low-voltage distribution network, a voltage control model of the active distribution network is constructed. Secondly, according to the historical operation data of the medium and low-voltage distribution network, through a data-driven method, based on the Koopman data-driven theory, the traditional non-linear power flow model is transformed into a high-dimensional linear power flow model that does not depend on impedance parameters. Aiming at the problem that the original power flow model fails due to the topology change of the medium and low-voltage distribution network, an adaptive dimension augmented power flow method is used to train the meta-matrix to describe the mapping relationship between the topology (adjacency matrix) and the linearized power flow model. Thus, through the adaptive dimension augmented power flow considering topology switching, global PF sensitivity can be generated under the new topology, and the linearized power flow matrix is updated through the meta-matrix and the new adjacency matrix. On the one hand, the accuracy of power flow calculation does not depend on the static model parameters of the distribution network. On the other hand, when the topology changes, the linearized power flow model under the new topology can be obtained without training data, so that while maintaining the safe operation of the distribution network voltage, the retraining of the distribution network power flow model can be avoided, greatly saving computing resources and training time. Finally, based on the differential characteristics of the high-voltage distribution network and the medium and low-voltage distribution network, generalized Benders decomposition is used for coordinated solution to realize the global voltage control of the active distribution network. Thus, through the distributed cooperative optimization algorithm of generalized Benders decomposition, it can quickly converge to the global optimal solution after a limited number of information exchanges, and is especially suitable for solving the voltage safety operation problems caused by the sudden change of distributed power sources and the topology change of the distribution network due to extreme weather.

[0172] For better illustration, refer to Figure 3, showing the overall flowchart of an active distribution network voltage control method with adaptive topology change provided by an embodiment of the present invention. It should be noted that this embodiment only briefly describes the general process of active distribution network voltage control with adaptive topology change. The specific implementation process of each step can be understood by referring to the relevant content in the foregoing embodiments, and will not be elaborated here. It can be understood that the present invention is not limited thereto.

[0173] Step 301: Construct a voltage control model for the active distribution network according to the coupling relationship between the high-voltage distribution network and multiple medium- and low-voltage distribution networks;

[0174] Step 302: Obtain the historical topology information diagram and historical linear power flow model of the medium- and low-voltage distribution network. According to the historical topology information diagram data, construct a historical adjacency matrix, and according to the historical linear power flow model and the historical adjacency matrix, construct a retraining sample set and a meta-matrix mapping relationship model;

[0175] Step 303: Substitute the retraining sample set into the meta-matrix mapping relationship model, and solve the meta-matrix mapping relationship model after substituting the retraining sample set according to the least squares regression and the generalized inverse matrix to obtain the meta-matrix of the medium- and low-voltage distribution network;

[0176] Step 304: Optimize and solve the voltage control model through generalized Benders decomposition, and use the optimal solution of the boundary variables obtained by the solution as the voltage control strategy during the power grid dispatching of the high-voltage distribution network and multiple medium- and low-voltage distribution networks;

[0177] Step 305: During the optimization solution process, for the target medium- and low-voltage distribution network with topology change, update the power flow model based on the meta-matrix in combination with the adjacency matrix under the new topology to obtain a new linear power flow model, and perform the power flow calculation of the target medium- and low-voltage distribution network based on the new linear power flow model.

[0178] To enable those skilled in the art to better understand the technical solution of the present invention, the following uses a specific example to illustrate the embodiments of the present invention.

[0179] The schematic diagram of the topology structure of the active distribution network (high-voltage distribution network - medium- and low-voltage distribution network) in this example is as Figure 4 shown. Among them, three medium- and low-voltage distribution networks with large-scale distributed photovoltaic access are connected to the high-voltage distribution network through buses 3, 7, and 14. Select a certain boundary section for voltage optimization control of the active distribution network.

[0180] First is the scenario setting. Under extreme weather conditions, it may cause changes in the topology structure of the medium- and low-voltage distribution network, resulting in large-scale power flow transfer and voltage violation of the distribution network. In this example, the following three scenarios are set for case analysis and verification after the topology of the medium- and low-voltage distribution network mutates due to extreme weather:

[0181] Scenario 1: Natural voltage distribution without optimization;

[0182] Scenario 2: Independent optimization of high-voltage distribution network and medium- and low-voltage distribution network;

[0183] Scenario 3: Active distribution network optimization based on generalized Benders decomposition.

[0184] Next is the comparison of optimization results. After extreme weather causes topological reconstruction, the actual voltage of the medium- and low-voltage distribution network and the visualization diagram of the voltage distribution obtained after adaptive dimension augmentation are as Figure 5 shown, and the visualization diagram of the effect of the medium- and low-voltage distribution network voltage before and after optimization is as Figure 6 shown.

[0185] Combined with Figure 5 , after closing different numbers of tie lines, the actual voltage of the medium- and low-voltage distribution network 1 is compared with the result obtained by using the method proposed in the present invention. The results show that the adaptive dimension augmentation power flow method proposed in the present invention can accurately track the actual voltage distribution and effectively solve the optimization problem of the medium- and low-voltage distribution network.

[0186] Figure 6 It is a visualization comparison diagram of the voltage optimization effect of the medium- and low-voltage distribution network 1 under three scenarios. The results show that by using the method proposed in the present invention, not only the voltage over-limit problem is effectively solved, but also the voltage distribution is more balanced, significantly reducing the voltage deviation cost. In addition, compared with independent optimization, the coordinated optimization between the high-voltage distribution network and the medium- and low-voltage distribution network further improves the balance of the voltage distribution.

[0187] Referring to Figure 7 , a structural block diagram of an active distribution network voltage control device with adaptive topological changes provided by an embodiment of the present invention is shown. The active distribution network includes a first voltage level distribution network and a plurality of second voltage level distribution networks; the device may specifically include:

[0188] A voltage control model construction unit 701, configured to construct a voltage control model of the active distribution network according to the coupling relationship between the first voltage level distribution network and the plurality of second voltage level distribution networks;

[0189] An optimization solution unit 702, configured to perform optimization solution on the voltage control model, and use the optimal solution of the boundary variable obtained by the solution as a voltage control strategy when performing power grid scheduling on the first voltage level distribution network and the plurality of second voltage level distribution networks;

[0190] A power flow model updating unit 703, configured to update a power flow model based on a meta-matrix for a target secondary voltage level distribution network with topological changes during an optimization solution process, obtain a new linear power flow model, and perform power flow calculation for the target secondary voltage level distribution network based on the new linear power flow model.

[0191] In an alternative embodiment, the voltage control model construction unit 701 includes:

[0192] A first objective function construction unit, configured to construct a first objective function for the first voltage level distribution network with the goal of minimizing the network active power loss cost, the discrete voltage regulating device operation cost, and the on-load tap changer regulation cost;

[0193] A first voltage optimization model construction unit, configured to construct a first voltage optimization model for the first voltage level distribution network based on the first objective function and a pre-constructed first multi-constraint condition;

[0194] A second objective function construction unit, configured to construct a second objective function for each of the second voltage level distribution networks, taking into account the sudden drop in high-proportion distributed power generation output and the wide-area power flow changes caused by topological reconstruction under extreme weather conditions;

[0195] A second voltage optimization model construction unit, configured to construct a second voltage optimization model for the second voltage level distribution network based on the second objective function and a pre-constructed second multi-constraint condition;

[0196] A boundary constraint construction unit, configured to construct boundary constraints between the first voltage level distribution network and each of the second voltage level distribution networks by considering the coupling relationship between the first voltage level distribution network and the multiple second voltage level distribution networks;

[0197] A voltage control model integration unit, configured to use the first voltage optimization model, each of the second voltage optimization models, and each of the boundary constraints as the voltage control model of the active distribution network.

[0198] In an alternative embodiment, each of the second voltage level distribution networks corresponds to a pre-trained meta-matrix; the device further includes a meta-matrix construction unit, and the meta-matrix construction unit includes:

[0199] A historical topology data acquisition unit, configured to acquire the historical topology information map and the historical linear power flow model of the second voltage level distribution network;

[0200] A historical adjacency matrix construction unit, configured to construct a historical adjacency matrix based on the historical topology information map data;

[0201] A mapping relationship model construction unit, configured to construct a retraining sample set and a meta-matrix mapping relationship model according to the historical linear power flow model and the historical adjacency matrix;

[0202] A meta-matrix solving unit, configured to substitute the retraining sample set into the meta-matrix mapping relationship model, and solve the meta-matrix mapping relationship model after substituting the retraining sample set according to the least squares regression and the generalized inverse matrix, so as to obtain the meta-matrix of the second voltage level distribution network.

[0203] In an optional embodiment, the power flow model updating unit 703 includes:

[0204] A new adjacency matrix obtaining unit, configured to obtain a new adjacency matrix of a target second voltage level distribution network with topological changes;

[0205] A linear power flow matrix column vector solving unit, configured to extract an adjacency matrix column vector from the new adjacency matrix, and substitute the adjacency matrix column vector into the meta-matrix mapping relationship model to solve the linear power flow matrix column vector;

[0206] A matrix element rearrangement unit, configured to rearrange matrix elements of the linear power flow matrix column vector to obtain a new linear power flow model of the target second voltage level distribution network under the new topology.

[0207] In an optional embodiment, the device further includes:

[0208] A historical operation data obtaining unit, configured to obtain historical operation data of the second voltage level distribution network;

[0209] A historical linear power flow model construction unit, configured to construct a historical linear power flow model of the second voltage level distribution network through state space mapping according to the historical operation data.

[0210] In an optional embodiment, the historical operation data includes distribution network node voltage, distributed power source power, energy storage power, load power, and distribution network supply power; the historical linear power flow model construction unit includes:

[0211] A net power model construction unit, configured to construct a net power model according to the distributed power source power, the energy storage power, and the load power;

[0212] A dimensionality-raising input sample set construction unit, configured to construct a plurality of input samples according to the net power model, perform state space dimensionality-raising on each of the input samples respectively to obtain corresponding dimensionality-raising input samples, and construct a dimensionality-raising input sample set based on each of the dimensionality-raising input samples;

[0213] An output sample set construction unit is configured to construct a plurality of output samples according to the node voltages and supply powers of the distribution network, and construct an output sample set based on each of the output samples;

[0214] A historical linear power flow model construction subunit is configured to construct a historical linear power flow model of the distribution network at the second voltage level by combining least squares data-driven training according to the dimensionality-increased input sample set and the output sample set.

[0215] In an alternative embodiment, the optimization solving unit 702 is specifically configured to:

[0216] Take the first voltage optimization model as the main problem, take each of the second voltage optimization models as sub-problems, consider each of the boundary constraints, and perform multiple rounds of optimization solving on the main problem and each of the sub-problems respectively through generalized Benders decomposition. After each round of solving, exchange the optimized solutions of the boundary variables between the distribution network at the first voltage level and each of the distribution networks at the second voltage level until the optimal solution of the boundary variables is obtained.

[0217] For the apparatus embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For related parts, refer to the corresponding description in the foregoing method embodiment.

[0218] It should be noted that, in order to enable those skilled in the art to better distinguish data of the same type but with different actual meanings, in the embodiments of the present invention, some technical features are distinguished and described by using first and second. First and second are only used for data distinction and have no other special meanings. It can be understood that the present invention makes no limitation in this regard.

[0219] The embodiments of the present invention further provide an electronic device, which includes a processor and a memory:

[0220] The memory is used to store program codes and transmit the program codes to the processor;

[0221] The processor is configured to execute the active distribution network voltage control method with adaptive topology change according to any embodiment of the present invention according to the instructions in the program codes.

[0222] The embodiments of the present invention further provide a computer-readable storage medium, which is used to store program codes, and the program codes are used to execute the active distribution network voltage control method with adaptive topology change according to any embodiment of the present invention.

[0223] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, apparatuses, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0224] In several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of devices or units can be in electrical, mechanical, or other forms.

[0225] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0226] In addition, in each embodiment of the present invention, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0227] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. And the aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0228] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An active distribution network voltage control method for adapting to topological changes, characterized in that The active distribution network includes a first-voltage-level distribution network and multiple second-voltage-level distribution networks; the method includes: Construct a voltage control model of the active distribution network according to the coupling relationship between the first-voltage-level distribution network and the multiple second-voltage-level distribution networks; Optimize and solve the voltage control model, and use the optimal solution of the boundary variables obtained by the solution as the voltage control strategy when performing power grid dispatching on the first-voltage-level distribution network and the multiple second-voltage-level distribution networks; During the optimization and solution process, update the power flow model based on the meta-matrix for the target second-voltage-level distribution network with topological changes to obtain a new linear power flow model, and perform power flow calculation on the target second-voltage-level distribution network based on the new linear power flow model.

2. The active distribution network voltage control method with adaptive topology change according to claim 1, characterized in that, The constructing the voltage control model of the active distribution network according to the coupling relationship between the first-voltage-level distribution network and the multiple second-voltage-level distribution networks includes: Construct a first objective function of the first-voltage-level distribution network with the goal of minimizing the network active power loss cost, the action cost of discrete voltage regulating equipment, and the regulation cost of on-load tap-changer transformers; Construct a first voltage optimization model of the first-voltage-level distribution network based on the first objective function and the pre-constructed first multi-constraint conditions; For each of the second-voltage-level distribution networks, construct a second objective function of the second-voltage-level distribution network while considering the sudden drop in the output of a high proportion of distributed power sources and the wide-area power flow changes caused by topological reconstruction under extreme weather conditions; Construct a second voltage optimization model of the second-voltage-level distribution network based on the second objective function and the pre-constructed second multi-constraint conditions; Considering the coupling relationship between the first-voltage-level distribution network and the multiple second-voltage-level distribution networks, construct boundary constraints between the first-voltage-level distribution network and each of the second-voltage-level distribution networks respectively; Use the first voltage optimization model, each of the second voltage optimization models, and each of the boundary constraints as the voltage control model of the active distribution network.

3. The active distribution network voltage control method with adaptive topology change according to claim 2, characterized in that Each of the second-voltage-level distribution networks corresponds to a pre-trained meta-matrix; the construction process of the meta-matrix includes: Obtain the historical topological information graph and the historical linear power flow model of the second-voltage-level distribution network; Construct a historical adjacency matrix according to the historical topological information graph data; Construct a retraining sample set and a meta-matrix mapping relationship model according to the historical linear power flow model and the historical adjacency matrix; Substitute the retraining sample set into the meta-matrix mapping relationship model, and solve the meta-matrix mapping relationship model after substituting the retraining sample set according to the least squares regression and the generalized inverse matrix to obtain the meta-matrix of the second-voltage-level distribution network.

4. The active distribution network voltage control method with adaptive topology change according to claim 3, characterized in that, The updating the power flow model based on the meta-matrix for the target second-voltage-level distribution network with topological changes to obtain a new linear power flow model includes: Obtain the new adjacency matrix of the target second-voltage-level distribution network with topological changes; Extract the column vectors of the adjacency matrix from the new adjacency matrix, and substitute the column vectors of the adjacency matrix into the meta-matrix mapping relationship model to solve for the column vectors of the linear power flow matrix; Rearrange the matrix elements of the column vectors of the linear power flow matrix to obtain the new linear power flow model of the target secondary voltage level distribution network under the new topology.

5. The active distribution network voltage control method with adaptive topology change according to claim 3, characterized in that It further includes: Obtain the historical operation data of the secondary voltage level distribution network; Construct the historical linear power flow model of the secondary voltage level distribution network through state space mapping according to the historical operation data.

6. The active distribution network voltage control method with adaptive topology change according to claim 5, characterized in that, The historical operation data includes the node voltage of the distribution network, the power of distributed power sources, the power of energy storage, the load power, and the power supplied by the distribution network; the constructing the historical linear power flow model of the secondary voltage level distribution network through state space mapping according to the historical operation data includes: Construct a net power model according to the power of distributed power sources, the power of energy storage, and the load power; According to the net power model, construct multiple input samples, perform state space dimensionality elevation on each input sample respectively to obtain the corresponding dimensionality-elevated input samples, and construct a dimensionality-elevated input sample set based on each dimensionality-elevated input sample; Construct multiple output samples according to the node voltage of the distribution network and the power supplied by the distribution network, and construct an output sample set based on each output sample; Combined with least squares data-driven training, construct the historical linear power flow model of the secondary voltage level distribution network according to the dimensionality-elevated input sample set and the output sample set.

7. The active distribution network voltage control method with adaptive topology change according to any one of claims 2 to 6, characterized in that The optimizing and solving the voltage control model includes: Take the first voltage optimization model as the main problem, and take each of the second voltage optimization models as sub-problems. Considering each of the boundary constraints, perform multi-round optimizing and solving on the main problem and each sub-problem respectively through generalized Benders decomposition, and after each round of solving, exchange the optimized solutions of the boundary variables between the first voltage level distribution network and each of the second voltage level distribution networks until the optimal solution of the boundary variables is obtained.

8. An active distribution network voltage control device with adaptive topology change, characterized in that The active distribution network includes a first voltage level distribution network and multiple second voltage level distribution networks; the device includes: A voltage control model construction unit, configured to construct a voltage control model of the active distribution network according to the coupling relationship between the first voltage level distribution network and the multiple second voltage level distribution networks; An optimizing and solving unit, configured to optimize and solve the voltage control model, and use the obtained optimal solution of the boundary variables as the voltage control strategy when performing power grid dispatching on the first voltage level distribution network and the multiple second voltage level distribution networks; A power flow model updating unit, configured to, during the optimizing and solving process, perform power flow model updating based on the meta-matrix for the target second voltage level distribution network with topological changes to obtain a new linear power flow model, and perform power flow calculation on the target second voltage level distribution network based on the new linear power flow model.

9. An electronic device, characterized in that, The device includes a processor and a memory: The memory is used to store program codes and transmit the program codes to the processor; The processor is configured to execute the active distribution network voltage control method with adaptive topology change according to any one of claims 1-7 based on the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is configured to store program code for executing the active distribution network voltage control method with adaptive topology change according to any one of claims 1-7.

Images