A method for adjusting a power grid structure and a related device

Abstract

The present application discloses a method for adjusting the grid structure of a power grid and a related device, which is used in the field of electric power. The method includes: obtaining the current system inertia corresponding to the current grid structure; calculating the frequency change rate based on the current system inertia and the difference between the power demand and the actual power supply; if the frequency change rate is greater than a preset first threshold, determining the target grid structure and the target constant corresponding to the target grid structure based on a pre-established deep learning model; the frequency change rate of the target grid structure is less than or equal to the first threshold; modifying the current constant applied in the current grid structure based on the target constant, and adjusting the current grid structure of the power grid to the target grid structure. Thus, by calculating the frequency change rate of the power grid, the phenomenon of imbalance in power supply and demand is discovered, and the target grid structure and the corresponding target constant that can improve the phenomenon are found through the deep learning model, and the current grid structure and the current constant are adjusted in time to improve the stability of the power grid.

Description

Technical Field

[0001] The present application relates to the field of electric power technology, and in particular to a method for adjusting a power grid structure and related devices. Background Art

[0002] As the scale of power systems continues to expand, the stability of the power grid faces increasingly severe challenges.

[0003] At present, in the process of power system operation, the grid structure of the power grid is usually fixed after setting. In the face of short-term fluctuations such as sudden load changes, changes in power generation and sudden failures, this fixed grid structure lacks flexibility and cannot be adjusted to these short-term fluctuations, which will lead to unstable grid frequency and voltage, resulting in economic losses such as power outages, damage to power generation equipment and power equipment, affecting the quality of life of the public and causing adverse social impacts.

[0004] Therefore, how to improve the stability of the power grid has become a problem that needs to be solved. Summary of the invention

[0005] Based on the above problems, the present application provides a method for adjusting the grid structure of a power grid and related devices, which can improve the stability of the power grid.

[0006] The embodiments of the present application disclose the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a method for adjusting a grid structure of a power grid, the method comprising:

[0008] Get the current system inertia corresponding to the current grid structure;

[0009] Calculating a frequency change rate based on the current system inertia and a difference between a power demand and an actual power supply;

[0010] If the frequency change rate is greater than a preset first threshold, a target grid structure and a target constant value corresponding to the target grid structure are determined based on a pre-established deep learning model; the frequency change rate of the target grid structure is less than or equal to the first threshold;

[0011] The current set values ​​applied in the current grid structure are modified based on the target set values, and the current grid structure of the power grid is adjusted to the target grid structure.

[0012] Optionally, determining the target grid structure based on a pre-established deep learning model includes:

[0013] Based on the pre-established deep learning model, the simulation system inertia of all optional grid structures is calculated in parallel, and the optional grid structure with the largest simulation system inertia is determined as the target grid structure.

[0014] Optionally, determining the target grid structure based on a pre-established deep learning model includes:

[0015] Based on the operating parameters of all optional grid structures, the frequency change rate and the difference change between the power generation and the power demand corresponding to each of the multiple adjustment schemes switched from the current grid structure to each optional grid structure are calculated through a pre-established deep learning model, and the scheme types to which the multiple adjustment schemes belong are obtained; the scheme types include stable and aggressive;

[0016] If a stable adjustment scheme exists, the target grid structure is determined from the optional grid structures corresponding to the stable adjustment scheme.

[0017] Optionally, before determining the target grid structure and the target value corresponding to the target grid structure based on the pre-established deep learning model, the method further includes:

[0018] If there is a faulty node, the faulty node is removed from the current grid structure; the node includes at least one of a power plant, a substation and a dispatching control center.

[0019] Optionally, determining the target grid structure based on a pre-established deep learning model includes:

[0020] Based on a pre-established deep learning model, and according to the matching situation between power generation and power demand in the current grid structure, a target grid structure that satisfies the balance condition between power generation and power demand is determined.

[0021] Optionally, the determining, based on a pre-established deep learning model and according to a matching condition between power generation and power demand in the current grid structure, a target grid structure that satisfies a balance condition between power generation and power demand includes:

[0022] Obtaining the power generation change rate of each power generation type grid node in the current grid structure;

[0023] If there are grid nodes whose power generation is reduced and whose power generation change rate is greater than a preset change rate threshold, then based on the pre-established deep learning model and according to the matching of power generation and power demand in the current grid structure, determine a target grid structure in which a power generation type grid grid is added on the basis of the current grid structure;

[0024] If there are grid nodes with increased power generation and a power generation change rate greater than a preset change rate threshold, then based on a pre-established deep learning model and according to the matching of power generation and power demand in the current grid structure, a target grid structure for adding a load-type power grid grid is determined based on the current grid structure.

[0025] Optionally, determining the target grid structure based on a pre-established deep learning model includes:

[0026] Based on a pre-established deep learning model, and according to the security priority of the current grid structure, a target network structure having a stability higher than a preset limit is determined.

[0027] Optionally, determining the target grid structure based on a pre-established deep learning model includes:

[0028] Based on the pre-established deep learning model and grid transfer function, the performance impact of multiple regulation schemes switching from the current grid structure to various optional grid structures on other grids in the power system is obtained;

[0029] Based on the pre-established deep learning model and the performance impact, a target grid structure is determined.

[0030] Optionally, the grid transfer function comprises a back propagation algorithm.

[0031] Optionally, determining a target value corresponding to the target grid structure based on a pre-established deep learning model includes:

[0032] Through a pre-established deep learning model, based on a typical value library and a target grid structure, a target constant value corresponding to the target grid structure is determined; the typical value library at least includes a correspondence between historical constant value data and a grid structure.

[0033] Optionally, adjusting the current grid structure to the target grid structure includes:

[0034] Based on the current grid structure and the target grid structure, determining a target action that a plurality of relays need to perform; the target action is opening or closing;

[0035] Based on the action time information of each relay, multiple relays are controlled to perform the target action to achieve adjustment from the current grid structure to the target grid structure; the action time information includes the transmission time of the target set value, the time required to write the target set value, the duration of the protection action, the time required for the relay to close, the time required for the relay to open, and the time required to return the status information.

[0036] Optionally, adjusting the current grid structure to the target grid structure includes:

[0037] Based on the target constants stored in the binary tree structure, parallel calculation is performed to obtain the action information of multiple relays that need to perform the target action during the process of adjusting the current grid structure to the target grid structure; the target action is opening or closing; the action information includes the time node when the relay is allowed to act and the order in which the multiple relays perform the actions;

[0038] Based on the action information of each of the plurality of relays, the plurality of relays are controlled to perform a target action, so as to adjust the current grid structure to the target grid structure.

[0039] In a second aspect, an embodiment of the present application provides a device for regulating a power grid structure, the device comprising: an acquisition module, a calculation module, a determination module and a regulation module;

[0040] The acquisition module is used to acquire the current system inertia corresponding to the current grid structure;

[0041] The calculation module is used to calculate the frequency change rate based on the current system inertia and the difference between the power demand and the actual power supply;

[0042] The determination module is used to determine the target grid structure and the target constant value corresponding to the target grid structure based on a pre-established deep learning model when the frequency change rate is greater than a preset first threshold; the frequency change rate of the target grid structure is less than or equal to the first threshold;

[0043] The adjustment module is used to modify the current set value applied in the current grid structure based on the target set value, and adjust the current grid structure to the target grid structure.

[0044] In a third aspect, an embodiment of the present application provides a regulating device for a power grid structure, the device comprising: a memory and a processor;

[0045] The memory is used to store program code and transmit the program code to the processor;

[0046] The processor is used to execute the steps of the method for adjusting the power grid structure described in any embodiment of the first aspect according to the program code.

[0047] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program runs on a regulating device of a power grid structure, the regulating device of the power grid structure performs the steps of the regulating method of the power grid structure as described in any embodiment of the first aspect.

[0048] Compared with the prior art, this application has the following beneficial effects:

[0049] The embodiment of the present application provides a method for adjusting the grid structure of a power grid, the method comprising: first, obtaining the current system inertia corresponding to the current grid structure; then, based on the current system inertia and the difference between the power demand and the actual power supply, calculating the frequency change rate; then, if the frequency change rate is greater than a preset first threshold, determining the target grid structure and the target constant corresponding to the target grid structure based on a pre-established deep learning model; the frequency change rate of the target grid structure is less than or equal to the first threshold; finally, modifying the current constant applied in the current grid structure based on the target constant, and adjusting the current grid structure of the power grid to the target grid structure. Thus, by calculating the frequency change rate of the power grid grid, the phenomenon of imbalance in power supply and demand and fluctuation in the power system can be discovered in time, and the target grid structure and the corresponding target constant that can improve this phenomenon can be found through the deep learning model, so that the current grid structure and the current constant can be adjusted in time, so as to better cope with various fluctuations in the power system and improve the stability of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A flow chart of a method for adjusting a power grid structure provided in an embodiment of the present application;

[0052] Figure 2 A flow chart of another method for adjusting the grid structure of a power grid provided in an embodiment of the present application;

[0053] Figure 3 A flow chart of a method for determining a target grid structure in the case of a sudden change in power generation provided in an embodiment of the present application;

[0054] Figure 4 A schematic diagram of a binary tree structure provided in an embodiment of the present application;

[0055] Figure 5 A schematic diagram of a regulating device for a power grid structure provided in an embodiment of the present application;

[0056] Figure 6 A structural diagram of a regulating device for a power grid structure provided in an embodiment of the present application. DETAILED DESCRIPTION

[0057] The power grid structure adjustment method and related devices provided in the present application can be used in the power field. The above is only an example and does not limit the application field of the power grid structure adjustment method and related devices provided in the present application.

[0058] The terms "first", "second", "third" and "fourth" etc. in the specification, claims and drawings of this application are used to distinguish different objects rather than to limit a specific order.

[0059] In the embodiments of the present application, the words "as an example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "as an example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the words "as an example" or "for example" is intended to present the relevant concepts in a specific way.

[0060] The terms used in the implementation section of this application are only used to explain the specific embodiments of this application and are not intended to limit this application.

[0061] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0062] See also Figure 1 , which is a flow chart of a method for adjusting a power grid structure provided in an embodiment of the present application, the method comprising:

[0063] S101: Obtain the current system inertia corresponding to the current grid structure.

[0064] In the power system, system inertia (also called rotational inertia or moment of inertia) is an important parameter to measure the sum of the kinetic energy of the rotors of all synchronous generators in the system. In the embodiment of the present application, the power grid is the part of the power system responsible for the transmission and distribution of electric energy, including but not limited to transmission lines, substations and distribution networks, etc.; the current system inertia is the total system inertia of all synchronous generators in the current grid structure. The larger the current system inertia, the stronger the resistance of the corresponding current grid structure to frequency changes, the smaller the frequency fluctuation, and the more stable it is during operation.

[0065] With the widespread use of renewable energy, the number of new energy units such as wind power generation and photovoltaic power generation in the power system has gradually increased. However, since new energy units are usually connected to the power grid through power electronic converters, they have decoupling characteristics, that is, the output of new energy units is not affected by the frequency changes of the power system, and the inertia coefficient is 0, which cannot provide inertia support for the power grid, which leads to a decrease in the frequency stability of the power grid when facing power fluctuations.

[0066] To this end, in the embodiments of the present application, for the new energy units, a grid-forming inverter can be introduced to simulate the behavior of a synchronous generator and provide inertia support similar to that of a synchronous generator to improve the stability of the power grid.

[0067] In the grid structure including new energy units, its equivalent inertia coefficient H eq It can be expressed as:

[0068]

[0069] Among them, S G is the total rated capacity of traditional synchronous units in the grid structure; S D is the total rated capacity of the renewable energy units in the grid structure; S L is the total load capacity in the grid structure; M is the total number of traditional synchronous units; N is the total number of new energy units; H i G is the inertia coefficient of the i-th traditional synchronous unit; S i G is the rated power of the i-th conventional synchronous unit; H j D is the inertia coefficient of the j-th new energy unit; S j D is the rated power of the jth new energy unit; H L is the equivalent inertia coefficient of the load.

[0070] S102: Calculate the frequency change rate based on the current system inertia and the difference between the power demand and the actual power supply.

[0071] Frequency Change Rate (RoCoF) indicates how many hertz the grid frequency changes per second. As an example, the frequency change rate can be calculated by the formula: Frequency Change Rate (Hz / s) = (50Hz / 2) × (Power Supply and Demand Imbalance (MW) / System Inertia (MVA·s)); where the power supply and demand imbalance is the difference between the power demand and the actual power supply, for example, the power supply and demand imbalance can be the absolute value of the difference between the power demand and the actual power supply.

[0072] S103: Determine whether the frequency change rate is greater than a preset first threshold, if so, execute step S104; if not, execute step S101.

[0073] In the case of an imbalance between electricity supply and demand, as the system inertia decreases, the frequency change rate will gradually increase, which may cause the frequency of the power system to drop to the frequency threshold set in the power system protection measure "Underfrequency Load Shedding" (UFLS), causing a major power outage.

[0074] As an example, when the frequency change rate is greater than the first threshold, it can be considered that the power system is facing large fluctuations and there is a risk that the frequency of the power system will drop to the "low-frequency load reduction" frequency threshold. It is necessary to execute steps S104~S105 to adjust the grid structure of the power grid so that the frequency change rate is less than or equal to the first threshold to maintain the stability of the power grid; when the frequency change rate is less than or equal to the first threshold, it can be considered that the power system is facing small fluctuations and there is no need to adjust the grid structure of the power grid. Steps S101~S103 can be repeated until the judgment result of step S103 is yes.

[0075] S104: Determine a target grid structure and a target constant corresponding to the target grid structure based on a pre-established deep learning model.

[0076] As an example, a pre-established deep learning model can simulate and generate all achievable grid structures in the power system as optional grid structures; or a grid structure library can be pre-established in the deep learning model, in which all optional grid structures are stored; a grid structure library including numerous sub-grids can also be pre-established in the deep learning model, and the sub-grids can be pre-classified according to multiple classification methods, so as to quickly obtain a target grid structure that meets the requirements.

[0077] For example, subgrids can be pre-divided into load-type subgrids and power generation-type subgrids according to their functions; subgrids can be pre-divided into photovoltaic terminal grids, thermal power terminal grids, hydropower terminal grids, and wind power terminal grids according to the type of power plant; subgrids can be pre-divided into large inertia subgrids with system inertia within the first interval, medium inertia subgrids with system inertia within the second interval, small inertia subgrids with system inertia within the third interval, and subgrids with system inertia of 0 according to the size of system inertia, etc. Therefore, when the current grid structure needs to be adjusted, one or more subgrids can be selected as needed to combine to obtain a target grid structure, or one or more subgrids can be selected and combined with the current grid structure to obtain a target grid structure, thereby improving the computing efficiency.

[0078] The deep learning model can select the optimal stability item from all optional grid structures as the target grid structure, and the frequency change rate of the target grid structure is less than or equal to the first threshold. For example, the simulation system inertia of all optional grid structures can be calculated in parallel through the deep learning model, and the optional network structure with the largest simulation system inertia is determined as the target network structure.

[0079] Exemplarily, a typical value library can be pre-established in the deep learning model, which includes at least the correspondence between historical constant value data and grid structure; wherein the constant value may include but is not limited to one or more of frequency deviation limit, frequency change rate limit, operating voltage range, voltage fluctuation amplitude, reactive power, maximum transmission power and relay protection parameters.

[0080] Therefore, the historical fixed value data establishes a reference sample for the target fixed value corresponding to the target grid structure. Through the deep learning model, an applicable target fixed value can be provided based on the target grid structure. For example, if the deep learning model determines that the target grid structure is isomorphic to a typical grid structure in the typical value library, the target fixed value with only a small difference from the fixed value of the typical grid structure will be output.

[0081] Optionally, the deep learning model can be trained using data in a typical value library. After obtaining the target network structure, the deep learning model directly outputs the target constant corresponding to the target network structure. The deep learning model can also be trained using the correspondence between other historical constant data and the grid structure. After obtaining the target network structure, the deep learning model refers to the data in the typical value library and outputs the target constant corresponding to the target network structure.

[0082] Changing the grid structure is divided into grid disconnection and grid connection. Its essence is to change the power flow and characteristic parameters of the grid. The calculation grid follows the chain rule, and its specific impact follows the back propagation algorithm. The farther away from the target grid, the smaller the correlation of the grid change, and the smaller the impact on the target grid.

[0083] In the context of deep learning, concepts such as the chain rule of neural networks in artificial intelligence are introduced and coexisted. Deep learning training is performed on parameter changes caused by grid structure changes, and errors are corrected through the back propagation algorithm, so that the parallel computing speed and accuracy of the deep learning model for real-time data can meet the needs of grid changes, that is, in the process of grid decoupling and grid connection, the safety of the target grid structure can be obtained through the trained deep learning model. Therefore, by defining the data characteristics of control parameters and production factors in the grid, forming data vectors, and extracting data vector characteristics, the deep learning model can be calculated based on data vectors, and can speed up the calculation speed by parallel computing the stability and adjustment feasibility of multiple grid structures, quickly obtain the appropriate target grid structure and its corresponding target value, and reduce the safety risks brought by grid structure transformation. Grid structure transformation continuously iterates and accumulates data assets. The related functions generated based on these data assets are used as the output layer to control more complex physical processes and equipment in the power system. In the solution before the emergence of artificial intelligence, without powerful computing power and neural network algorithms, it is impossible to achieve grid connection and decoupling of power grids. In the embodiment of the present application, the convolutional neural network is symbiotically integrated into the power system, so that deep learning algorithms and the like can symbiotically coexist in the power system, and the grid structure can be adjusted while ensuring safety, thereby improving the stability of the power grid.

[0084] As an example, in the modeling stage, the measured values ​​of grid data corresponding to various grid structures can be collected and classified, for example, according to the characteristics of frequency deviation limit, frequency change rate limit and system inertia; then, the classified data is fitted through mathematical methods such as exponential smoothing to better reflect the actual situation; then, data cleaning is performed, and key features that are helpful for subsequent analysis, that is, key production factors, are extracted from the original data; then, the difference between the fitted data and the original data is compared, and the error value is calculated; finally, the relationship between the production factors of different grids can be found and the model can be established. In the model training stage, the model can be trained with known data so that the model can learn the intrinsic connection between these factors. During the training process, a weight sharing mechanism can be used to ensure consistent results in different parts of the neural network; the back propagation algorithm can also be used for correction at the same time to adjust the model parameters to reduce errors and reduce information loss caused by numerical calculations.

[0085] In addition, in the embodiments of the present application, the hierarchical expression characteristics of the deep learning model can be used to use data at different levels to reflect the different properties of the grid structure. For example, low-level features reflect basic physical parameters, such as voltage, current, etc.; mid-level features reflect grid behavior characteristics, such as frequency response, etc.; high-level features reflect the stability and reliability of the grid, etc. As a result, the deep learning model can learn the multi-level features of the data and can achieve better performance in complex tasks such as selecting the target grid structure and calculating the target constant.

[0086] On the other hand, for specific functions and control logic within the power system, output functions can also be defined according to specific needs, such as control strategies, and specific operations or control logic can be implemented through these operators.

[0087] S105: Modify the current set value applied in the current grid structure based on the target set value, and adjust the current grid structure of the power grid to the target grid structure.

[0088] Specifically, the target setting can be sent to various nodes such as power plants, substations and dispatching control centers involved in the target power grid structure. After receiving the target setting, these nodes can verify the rationality of the target setting by region. If the verification is successful, the current setting will be modified to the target setting.

[0089] Optionally, for a power system with relays, the current grid structure is adjusted to the target grid structure, specifically, the current grid structure is adjusted to the target grid structure by controlling the relays to perform opening or closing operations. In this case, the target actions that need to be performed by multiple relays can be determined based on the current grid structure and the target grid structure, and then the multiple relays can be controlled to perform the target actions based on the action time information of each relay, so that the current grid structure is adjusted to the target grid structure, wherein the target action is opening or closing, and the action time information includes the transmission time S1 of the target set value, the time S2 required to write the target set value, the duration S3 of the protection action, the time S4 ​​required for the relay to close, the time S5 required for the relay to open, and the time S6 required to return the status information.

[0090] Specifically, the target setting value can be sent to each node first, and the transmission time of the target setting value is recorded as S1; then the target setting value is written into the protection device, and the time required to write the setting value is recorded as S2, where the protection device is used to modify the setting value; while sending the target setting value, the protection action can be executed, and the duration of the protection action is recorded as S3; the durations of S1, S2, and S3 need to satisfy S1 + S2 < S3. The time required for the relay to close is recorded as S4, the time required for the relay to trip is recorded as S5, and the time required to return the relay status information and the setting value status information is recorded as S6. The durations of S1, S2, S3, S4, S5, and S6 need to satisfy: S4 is greater than the time when the corresponding actions of S1, S2, S3, and S6 are all completed, and S5 is less than the time when the corresponding actions of S1, S2, S3, and S6 are all completed.

[0091] Optionally, for a digital power system without a relay, during the grid structure adjustment process, the time required for the tripping and closing actions is the same, and the impact of the grid structure adjustment on the power grid is small. The protection action can be not executed, and the setting value modification and the grid structure adjustment actions can be carried out simultaneously.

[0092] Optionally, before performing the step of S105, the smallest stable grid can be incorporated into the current grid structure first to improve the safety of the power grid grid structure adjustment process.

[0093] Thus, in the embodiments of the present application, by calculating the frequency change rate of the power grid grid, the phenomena of power supply and demand imbalance and power system fluctuations can be discovered in time, and the target grid structure and the corresponding target setting value that can improve this phenomenon can be found through the deep learning model. The current grid structure and the current setting value can be adjusted in time, better coping with various fluctuations in the power system, improving the stability of the power grid, and avoiding adverse phenomena such as "large machines with small grids" locally.

[0094] See Figure 2 , this figure is a flowchart of another method for adjusting the power grid grid structure provided by the embodiments of the present application. The method includes:

[0095] S201: Obtain the current system inertia corresponding to the current grid structure.

[0096] S202: Calculate the frequency change rate based on the current system inertia and the difference between the power supply demand and the actual power supply.

[0097] S203: Determine whether the frequency change rate is greater than a preset first threshold. If so, execute the step of S204; if not, execute the step of S201.

[0098] S204: If there is a faulty node, disconnect the faulty node from the current grid structure.

[0099] Specifically, the fault may include but is not limited to a short circuit, a line break, and a ground fault, etc.; the node includes at least one of a power plant, a substation, and a dispatching control center.

[0100] As an example, relay protection devices can be used to monitor parameters such as current and voltage in the power grid to determine whether there are faulty nodes. Or, installed smart sensors can be used to continuously monitor environmental conditions and equipment health, and big data analysis can be used to determine whether there are faulty nodes or predict nodes that may fail within a set period of time.

[0101] After a faulty node is discovered, the faulty node can be removed from the current grid structure by, for example, controlling the opening and closing of a relay, to prevent the faulty node from affecting the stability and reliability of the grid.

[0102] Optionally, the electrical parameters in the power grid can be recorded by a fault recorder so that after the faulty node is found, the changes in electrical parameters before and after the fault occurs can be provided to help analyze the cause of the fault and provide data support for fault recovery.

[0103] S205: Based on the pre-established deep learning model, and according to the matching of power generation and power demand in the current grid structure, determine a target grid structure that satisfies the balance condition between power generation and power demand.

[0104] As an example, if the faulty node is removed through step S204, the matching situation between power generation and power demand in the current grid structure after the faulty node is removed, that is, the source-load matching situation, is calculated; then, the target grid structure that satisfies the balance conditions between power generation and power demand can be obtained through the deep learning model.

[0105] For example, Figure 3 As shown, the power generation change rate of each power generation type grid node in the current grid structure can be obtained. If the power generation of the grid node of the new energy unit changes suddenly, for example, the power generation change rate exceeds the preset change rate threshold, it can be determined whether its power generation increases or decreases suddenly. If it decreases suddenly, then according to the matching of power generation and electricity demand in the current grid structure, it is determined to increase the target grid structure of the power generation type grid on the basis of the current grid structure to provide more power generation for the power grid; if it increases suddenly, according to the matching of power generation and electricity demand in the current grid structure, it is determined to increase the target grid structure of the load type grid on the basis of the current grid structure to provide more load for the power grid and consume the incremental power generation of the new energy unit.

[0106] S206: Based on a pre-established deep learning model and according to the security priority of the current grid structure, a target network structure having a stability higher than a preset limit is determined.

[0107] As an example, security priorities can be set in advance for various power consumption sites, power plants, substations, dispatching and control center nodes. For example, the grid structure that supplies power to medical institutions can be set to the first priority, and the grid structure that supplies power to civil facilities can be set to the second priority, with the first priority being higher than the second priority. The security priority of the current grid structure depends on the highest security priority among the nodes involved. For example, if the current grid structure supplies power to a medical institution with the first priority, then the current network structure is the first priority. A grid structure with a stability higher than a preset limit can be used as a target grid structure, or a grid structure with a stability higher than a preset limit can be incorporated into the current grid structure to form a target grid structure, thereby ensuring that changes in the grid structure will not affect the power consumption of medical institutions.

[0108] Typical grid structures include radial structure, ring structure, petal structure and hybrid structure. The stability of the grid structure is related to the connection method and layout of the power grid. The incorporation of different types of grid structures will have different impacts on the power grid.

[0109] For example, in a radial structure, the power source transmits electricity to the load point through a single path. Once a section of the line fails, it may cause a large-scale power outage and have low stability. Incorporating the radial grid structure into the current grid structure will cause larger fluctuations. In a petal-shaped structure, there is a main power source in the center, surrounded by multiple branches radiating outward, and each branch can branch out further. This structure not only provides a high degree of redundancy, but also allows flexible adjustment of power distribution as needed. It has high stability. Incorporating the petal-shaped grid structure into the current grid structure can enhance the stability of the power grid.

[0110] S207: Based on the operating parameters of all optional grid structures, a pre-established deep learning model is used to calculate the frequency change rates and the difference changes between the power generation and the power demand corresponding to each of the multiple adjustment schemes for switching from the current grid structure to each optional grid structure, and obtain the scheme types to which the multiple adjustment schemes belong respectively; the scheme types include stable and aggressive.

[0111] Specifically, switching from the current grid structure to an optional grid structure is an adjustment scheme. For multiple adjustment schemes, the frequency change rate of the power grid and the difference change between the power generation and the power demand are calculated during the adjustment process. If any one of the frequency change rate and the difference change between the power generation and the power demand is greater than a preset change threshold, the adjustment scheme is considered to be an aggressive adjustment scheme; otherwise, the adjustment scheme is considered to be a stable adjustment scheme.

[0112] As an example, the frequency change rate and the difference change between power generation and electricity demand can be calculated by mathematical methods such as data graph method, vector graph method or Euclidean distance mean method; the change threshold can be a safety threshold obtained by deep learning the historical frequency change rate and the difference change between historical power generation and electricity demand in different scenarios through a deep learning model; wherein the change threshold can be a piecewise function, for example, the piecewise function includes segment A and segment B, if the frequency change rate and the difference change between power generation and electricity demand fall into segment A of the piecewise function, then the corresponding regulation scheme can be considered to be stable; if the frequency change rate and the difference change between power generation and electricity demand fall into segment B of the piecewise function, then the corresponding regulation scheme can be considered to be aggressive.

[0113] Optionally, the adjustment scheme is not limited to the stable type and the radical type. According to actual needs, the adjustment scheme can be classified into more detailed types by setting multiple change thresholds. For example, the adjustment scheme can also include a balanced type whose radicality is between the stable type and the radical type.

[0114] S208: If a stable adjustment solution exists, determine a target grid structure from the optional grid structures corresponding to the stable adjustment solution.

[0115] Specifically, a stable regulation scheme is preferred, which can make the regulation process of the grid structure more stable and keep the power grid stable during the regulation process of the grid structure.

[0116] S209: Based on a pre-established deep learning model and a grid transfer function, obtain the performance impact of multiple adjustment schemes for switching from the current grid structure to various optional grid structures on other grids in the power system; based on the pre-established deep learning model and the performance impact, determine the target grid structure.

[0117] Among them, the grid transfer function may include a backpropagation algorithm, which is the core algorithm for training models in deep learning. The backpropagation algorithm calculates the gradient of the loss function with respect to the network parameters and uses these gradients to update the weights and bias terms in the network to minimize the gap between the predicted results and the actual values. In a multi-layer neural network, since the output of each layer is the input of the next layer, the error propagation is performed layer by layer, from the output layer to the input layer in reverse. As an example, the performance impact between grid structures can be obtained by combining the backpropagation algorithm with one or more error correction methods of the backpropagation algorithm such as weight initialization, activation function, gradient clipping, batch normalization, residual connection, and adaptive learning rate.

[0118] Specifically, for the grid structure composed of three sub-grids A, B, C, and D connected in sequence, the performance of the three sub-grids affects each other. The performance impact between the three sub-grids can be expressed by the following set of equations:

[0119] FA: F(X)=F(Y)+K(F(Z)+F(S));

[0120] FB: F(Y)=F(y)+PF(Z);

[0121] FC: F(Z)=F(z)+HF(S);

[0122] FD: F(S)=F(θ,η,0,N).

[0123] Among them, FA represents the performance impact of subgrid A; FB represents the performance impact of subgrid B; FC represents the performance impact of subgrid C; FD represents the performance impact of production factors on subgrid D of photovoltaic power station; θ represents the illuminance, η represents the photoelectric conversion efficiency, N represents the number of photovoltaic panels; K, P, and H represent the impact coefficients. Therefore, the back propagation algorithm is used to calculate the performance impact between multiple grid structures. For subgrids that are not directly related, the performance impact is analyzed by transferring influence and data characteristics, and the performance impact can be corrected through quantitative tools.

[0124] In the embodiment of the present application, if sub-grid C is disconnected from sub-grid B, the impact on sub-grid B will be F(Y). At this time, sub-grid D, which is directly connected to sub-grid C, is also disconnected from sub-grid A and sub-grid B. The performance impact on sub-grid A will be the result of the superposition of the three coefficients K, P, and H.

[0125] When the performance impact is introduced, a deep learning model can be used to determine a target grid structure whose frequency change rate is less than or equal to the first threshold, thereby more accurately predicting the grid structure adjustment result and obtaining a more stable target grid structure.

[0126] Optionally, steps S205, S206, S207~S208 and S209 can be executed one by one, and one of the multiple target network structures obtained is randomly selected as the adjustment target; or all of them can be executed and any one of the intersections of the target network structures obtained in each step can be taken as the adjustment target; and priorities can also be set for them. For example, the priority of step S206 is 1, the priority of step S205 is 2, the priority of steps S207~S208 is 3, and the priority of step S209 is 4; step S206 can be executed first. If only one target network structure is obtained, the target network structure is used as the adjustment target, and steps S205, S207~S208 and S209 are no longer executed; if multiple target network structures are obtained, step S205 is executed, and the intersection of the target network structures obtained by steps S205 and S206 is taken. If the intersection is empty, one of the multiple target network structures obtained by executing step S206 is selected as the target network structure; if the number of intersections is 1, the target network structure in the intersection is used as the adjustment target, and steps S207~S208 and S209 are no longer executed; if the number of intersections is multiple, steps S207~S208 are continued to be executed, and so on.

[0127] S210: Modify the current set value applied in the current grid structure based on the target set value, and adjust the current grid structure to the target grid structure.

[0128] As an example, data such as target values ​​can be obtained as follows Figure 4 The binary tree structure shown is used for storage, and then a binary tree traversal is performed based on the data stored in the binary tree structure to obtain action information such as the time nodes at which each relay is allowed to act and the sequence in which multiple relays perform actions. Therefore, based on the action information of each relay, the relay is controlled to perform actions such as opening or closing, thereby realizing the adjustment of the grid structure.

[0129] For example, the set value may include data such as the current and duration of the current allowed for each relay to operate. The data indicating the conditions for each relay to operate in the target set value may be stored in each node of the binary tree structure. The nodes may be arranged in order of magnitude, such as the node on the left branch of any node is numerically smaller than the node, and the node on the right branch of any node is numerically larger than the node. The data stored in each node is the current value allowed for the relay to operate. Figure 4In the , the current values ​​stored in each node are arranged from large to small as CFAEBD. Therefore, by storing target constant data such as the current allowed for each relay to act in a binary tree structure, before controlling the relay to perform actions such as opening or closing, based on the goals of minimizing the total action time of the relay or minimizing the economic loss, the time nodes allowed for each relay to act and the order in which multiple relays perform actions can be quickly obtained through parallel calculation, and the relay action plan can be obtained to quickly adjust the grid structure. Figure 5 , this figure is a schematic diagram of a regulating device for a power grid structure provided in an embodiment of the present application, and the device includes: an acquisition module 501, a calculation module 502, a determination module 503 and a regulation module 504.

[0130] The acquisition module 501 is used to acquire the current system inertia corresponding to the current grid structure.

[0131] The calculation module 502 is used to calculate the frequency change rate based on the current system inertia and the difference between the power demand and the actual power supply.

[0132] Determination module 503 is used to determine the target grid structure and the target constant corresponding to the target grid structure based on a pre-established deep learning model when the frequency change rate is greater than a preset first threshold; the frequency change rate of the target grid structure is less than or equal to the first threshold.

[0133] The adjustment module 504 is used to modify the current set value applied in the current grid structure based on the target set value, and adjust the current grid structure to the target grid structure.

[0134] Therefore, in the embodiment of the present application, by calculating the frequency change rate of the power grid, it is possible to promptly discover the imbalance in power supply and demand and the fluctuation in the power system, and through the deep learning model, find the target grid structure and the corresponding target constant that can improve this phenomenon, so that the current grid structure and the current constant can be adjusted in time, so as to better cope with various fluctuations in the power system, improve the stability of the power grid, and avoid local undesirable phenomena such as "large machine and small network".

[0135] Optionally, the determination module 503 includes an inertia determination unit, which is specifically used to: based on a pre-established deep learning model, parallelly calculate the simulation system inertia of all optional grid structures, and determine the optional grid structure with the largest simulation system inertia as the target grid structure.

[0136] Optionally, the determination module 503 includes a scheme determination unit, which is specifically used to: based on the operating parameters of all optional grid structures, calculate the frequency change rate and the difference change amount between the power generation and the power demand corresponding to each of the multiple adjustment schemes switched from the current grid structure to each optional grid structure through a pre-established deep learning model, and obtain the scheme types to which the multiple adjustment schemes belong respectively; the scheme types include stable and aggressive; if there is a stable adjustment scheme, determine the target grid structure from the optional grid structure corresponding to the stable adjustment scheme.

[0137] Optionally, the determination module 503 includes a demand matching unit, which is specifically used to: based on a pre-established deep learning model, determine a target grid structure that satisfies the balance condition between power generation and power demand according to the matching situation between power generation and power demand in the current grid structure.

[0138] Optionally, the demand matching unit is specifically used to: obtain the power generation change rate of each power generation type grid node in the current grid structure; if there is a grid node whose power generation is reduced and the power generation change rate is greater than a preset change rate threshold, then based on a pre-established deep learning model, according to the matching situation of power generation and electricity demand in the current grid structure, determine the target grid structure of adding a power generation type power grid grid on the basis of the current grid structure; if there is a grid node whose power generation is increased and the power generation change rate is greater than a preset change rate threshold, then based on a pre-established deep learning model, according to the matching situation of power generation and electricity demand in the current grid structure, determine the target grid structure of adding a load type power grid grid on the basis of the current grid structure.

[0139] Optionally, the determination module 503 includes a stability determination unit, which is used to determine a target network structure having stability higher than a preset limit based on a pre-established deep learning model and according to the security priority of the current grid structure.

[0140] Optionally, the determination module 503 includes an impact determination unit, which is used to obtain the performance impact of multiple adjustment schemes switching from the current grid structure to each optional grid structure on other grids in the power system based on a pre-established deep learning model and a grid transfer function; and determine the target grid structure based on the pre-established deep learning model and the performance impact.

[0141] Optionally, the determination module 503 includes a constant value determination unit, which is used to determine the target constant value corresponding to the target grid structure based on a typical value library and a target grid structure through a pre-established deep learning model; the typical value library at least includes the correspondence between historical constant value data and the grid structure.

[0142] Optionally, the adjustment module 504 is specifically used to: determine the target action that multiple relays need to perform based on the current grid structure and the target grid structure; the target action is opening or closing; based on the action time information of each relay, control the multiple relays to perform the target action to achieve adjustment from the current grid structure to the target grid structure; the action time information includes the transmission time of the target set value, the time required to write the target set value, the duration of the protection action, the time required to close the relay, the time required to open the relay, and the time required to return the status information.

[0143] Optionally, the adjustment module 504 is specifically used to: perform parallel calculation based on the target set value stored in the binary tree structure to obtain the action information of each of the multiple relays that need to perform the target action in the process of adjusting the current grid structure to the target grid structure; the target action is opening or closing; the action information includes the time node when the relay is allowed to act and the order in which the multiple relays perform the action;

[0144] Based on the action information of each of the multiple relays, the multiple relays are controlled to perform a target action to adjust the current grid structure to the target grid structure.

[0145] Optionally, in another regulating device of a power grid grid structure provided in an embodiment of the present application, a decoupling module is also included, which is used to decoupling the faulty node from the current grid structure when there is a faulty node; the node includes at least one of a power plant, a substation and a dispatching control center.

[0146] See also Figure 6 , this figure is a structural diagram of a regulating device for a power grid structure provided in an embodiment of the present application, and the device includes: a memory 601 and a processor 602.

[0147] Memory 601: used to store program codes and transmit the program codes to the processor.

[0148] Processor 602: used to execute the steps of the above-mentioned power grid structure adjustment method according to the instructions in the program code.

[0149] In addition, the present application also provides a computer-readable storage medium, which stores computer instructions. When the computer instructions are executed on the regulating device of the power grid structure, the regulating device of the power grid structure executes the steps of the above-mentioned power grid structure regulating method.

[0150] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components indicated as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0151] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for adjusting a power grid structure, characterized in that: The method comprises: Get the current system inertia corresponding to the current grid structure; Calculating a frequency change rate based on the current system inertia and a difference between a power demand and an actual power supply; If the frequency change rate is greater than a preset first threshold, a target grid structure and a target constant value corresponding to the target grid structure are determined based on a pre-established deep learning model; the frequency change rate of the target grid structure is less than or equal to the first threshold; modifying the current set value applied in the current grid structure based on the target set value, and adjusting the current grid structure of the power grid to the target grid structure; Wherein, determining the target grid structure based on a pre-established deep learning model includes: Obtaining the power generation change rate of each power generation type grid node in the current grid structure; If there are grid nodes whose power generation is reduced and whose power generation change rate is greater than a preset change rate threshold, then based on the pre-established deep learning model and according to the matching of power generation and power demand in the current grid structure, determine a target grid structure in which a power generation type grid grid is added on the basis of the current grid structure; If there are grid nodes with increased power generation and a power generation change rate greater than a preset change rate threshold, then based on a pre-established deep learning model and according to the matching of power generation and power demand in the current grid structure, a target grid structure for adding a load-type power grid grid is determined based on the current grid structure.

2. The method according to claim 1, characterized in that: Determine the target grid structure based on a pre-established deep learning model, including: Based on the pre-established deep learning model, the simulation system inertia of all optional grid structures is calculated in parallel, and the optional grid structure with the largest simulation system inertia is determined as the target grid structure.

3. The method according to claim 1, characterized in that Determine the target grid structure based on a pre-established deep learning model, including: Based on the operating parameters of all optional grid structures, the frequency change rate and the difference change between the power generation and the power demand corresponding to each of the multiple adjustment schemes switched from the current grid structure to each optional grid structure are calculated through a pre-established deep learning model, and the scheme types to which the multiple adjustment schemes belong are obtained; the scheme types include stable and aggressive; If a stable adjustment scheme exists, the target grid structure is determined from the optional grid structures corresponding to the stable adjustment scheme.

4. The method according to claim 1, characterized in that Before determining the target grid structure and the target value corresponding to the target grid structure based on the pre-established deep learning model, the method further includes: If there is a faulty node, the faulty node is removed from the current grid structure; the node includes at least one of a power plant, a substation and a dispatching control center.

5. The method according to claim 1, characterized in that Determine the target grid structure based on a pre-established deep learning model, including: Based on a pre-established deep learning model, and according to the security priority of the current grid structure, a target network structure having a stability higher than a preset limit is determined.

6. The method according to claim 1, characterized in that Determine the target grid structure based on a pre-established deep learning model, including: Based on the pre-established deep learning model and grid transfer function, the performance impact of multiple regulation schemes switching from the current grid structure to various optional grid structures on other grids in the power system is obtained; Based on the pre-established deep learning model and the performance impact, a target grid structure is determined.

7. The method according to claim 6, characterized in that The grid transfer function includes a back propagation algorithm.

8. The method according to claim 1, characterized in that: Determine the target value corresponding to the target grid structure based on the pre-established deep learning model, including: Through a pre-established deep learning model, based on a typical value library and a target grid structure, a target constant value corresponding to the target grid structure is determined; the typical value library at least includes a correspondence between historical constant value data and a grid structure.

9. The method according to claim 1, characterized in that: The adjusting the current grid structure of the power grid to the target grid structure comprises: Based on the current grid structure and the target grid structure, determining a target action that a plurality of relays need to perform; the target action is opening or closing; Based on the action time information of each relay, multiple relays are controlled to perform the target action to achieve adjustment from the current grid structure to the target grid structure; the action time information includes the transmission time of the target set value, the time required to write the target set value, the duration of the protection action, the time required for the relay to close, the time required for the relay to open, and the time required to return the status information.

10. The method according to claim 1, characterized in that The adjusting the current grid structure of the power grid to the target grid structure comprises: Based on the target constants stored in the binary tree structure, parallel calculation is performed to obtain the action information of multiple relays that need to perform the target action during the process of adjusting the current grid structure to the target grid structure; the target action is opening or closing; the action information includes the time node when the relay is allowed to act and the order in which the multiple relays perform the actions; Based on the action information of each of the plurality of relays, the plurality of relays are controlled to perform a target action, so as to adjust the current grid structure to the target grid structure.

11. A regulating device for a power grid structure, characterized in that: The device comprises: an acquisition module, a calculation module, a determination module and an adjustment module; The acquisition module is used to acquire the current system inertia corresponding to the current grid structure; The calculation module is used to calculate the frequency change rate based on the current system inertia and the difference between the power demand and the actual power supply; The determination module is used to determine the target grid structure and the target constant value corresponding to the target grid structure based on a pre-established deep learning model when the frequency change rate is greater than a preset first threshold; the frequency change rate of the target grid structure is less than or equal to the first threshold; The adjustment module is used to modify the current set value applied in the current grid structure based on the target set value, and adjust the current grid structure to the target grid structure; The determining module includes a demand matching unit, and the demand matching unit is specifically used for: The power generation change rate of each power generation type grid node in the current grid structure is obtained; if there is a grid node whose power generation decreases and whose power generation change rate is greater than a preset change rate threshold, then based on the pre-established deep learning model, according to the matching situation of power generation and electricity demand in the current grid structure, a target grid structure of adding a power generation type power grid grid is determined on the basis of the current grid structure; if there is a grid node whose power generation increases and whose power generation change rate is greater than a preset change rate threshold, then based on the pre-established deep learning model, according to the matching situation of power generation and electricity demand in the current grid structure, a target grid structure of adding a load type power grid grid is determined on the basis of the current grid structure.

12. A regulating device for a power grid structure, characterized in that: The device comprises: a memory and a processor; The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the steps of the method for adjusting the power grid structure according to any one of claims 1-10 according to the program code.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program. When the computer program runs on a regulating device of a power grid structure, the regulating device of the power grid structure executes the steps of the regulating method of the power grid structure according to any one of claims 1 to 10.

Images