A distribution network regulation method and system for a new power system

Abstract

This invention discloses a distribution network substation control method for a new type of power system, comprising: acquiring real-time measurement data of distribution substations within a designated area of ​​the new power system; based on the real-time measurement data, using a trained substation power loss calculation model to perform extrapolation and prediction to obtain predicted power loss data for the distribution network substations; combining the predicted power loss data to formulate and implement a flexible control strategy for source-grid-load-storage coordination in the distribution network substations, thereby improving the reliability of power supply in the distribution network; the training of the substation power loss calculation model includes: based on historical measurement data of the distribution network substations, using a distribution network substation aggregated equivalent model to obtain historical power equivalent data; using historical power equivalent data as input and historical power loss data as labels, training a long short-term memory network to obtain the substation power loss calculation model; this invention can achieve accurate and forward-looking prediction of power loss in distribution substations, providing reliable data support for subsequent control.

Description

Technical Field

[0001] This invention relates to a distribution network control method and system for new power systems, belonging to the technical field of flexible control and reliability improvement of distribution networks in new power systems. Background Technology

[0002] The new power system, dominated by new energy sources and featuring a high proportion of large-scale distributed power sources (wind power, photovoltaic), flexible loads, and energy storage devices, exhibits characteristics of bidirectional power flow, multidimensional regulation, and randomized operation in distribution networks. Therefore, the flexible regulation capability of the distribution network and the assurance of power supply reliability have become core and critical issues in the construction of the new power system. Utilizing regional distribution network flexible regulation simulation methods to formulate coordinated regulation strategies involving power generation, grid, load, and storage is an important technical means to ensure the safe and stable operation of the distribution network and improve power supply reliability in the new power system.

[0003] However, in existing new power systems, the large-scale integration of distributed generation, flexible loads, and energy storage devices in distribution substations leads to highly random and fluctuating operational characteristics. Traditional power loss calculation methods rely on fixed parameters, failing to accurately reflect the real-time operating characteristics of substations and hindering advanced power loss prediction. Furthermore, existing distribution network control strategies are mostly passive, lacking source-grid-load-storage coordinated optimization based on loss prediction, resulting in high distribution network power losses, poor voltage stability, and difficulty in effectively improving power supply reliability. In addition, traditional models do not combine substation aggregation and time-series forecasting, leading to complex modeling, low computational efficiency, and insufficient prediction accuracy, failing to meet the practical needs of flexible control and reliability improvement in distribution networks under new power systems. Summary of the Invention

[0004] The technical problem to be solved by this invention is that the power loss in existing distribution substations is difficult to predict accurately due to the diverse access of distributed power sources, flexible loads, and energy storage, and that traditional control methods are lagging behind and lack coordination.

[0005] To solve the above-mentioned technical problems, the present invention is implemented using the following technical solution.

[0006] On the one hand, the present invention provides a distribution network control method for a new type of power system, comprising:

[0007] Acquire real-time measurement data of distribution substations within a designated area of ​​the new power system;

[0008] Based on the real-time measurement data, the predicted power loss data of the distribution network area is obtained by using the trained power loss calculation model of the distribution area.

[0009] Based on the predicted power loss data, coordinated regulation of power generation, grid, load and storage in the distribution network area is carried out.

[0010] The training of the power loss calculation model for the transformer substation includes:

[0011] Based on the historical measurement data of the distribution network area, the historical power equivalent data is obtained by using the distribution network area aggregation equivalent model;

[0012] Using the historical power equivalent data as input and the historical power loss data as labels, the Long Short Time Memory Network is trained to obtain the trained power loss calculation model for the transformer area.

[0013] This invention acquires real-time measurement data of distribution transformer substations and uses a trained substation power loss calculation model to extrapolate and predict power loss. This enables accurate and forward-looking prediction of power loss in distribution transformer substations, providing reliable data support for subsequent regulation and control. It solves the problems of inaccurate traditional loss calculation and delayed prediction.

[0014] The distribution network area aggregated equivalent model includes: the power equivalent algorithm and the internal power loss algorithm of the aggregated equivalent object of the distribution network area;

[0015] The aggregated equivalent objects are adapted to the characteristics of distributed power sources, flexible loads, and multiple access to energy storage devices in new power systems, including five types of objects: wind power, photovoltaic, load, energy storage, and reactive power compensation devices.

[0016] This invention aggregates massive distributed resources in distribution network areas into five core equivalent objects, simplifying the processing of distributed resources and significantly reducing the computational complexity of distribution network control simulation.

[0017] The power equivalence algorithms for the aggregated equivalent objects include five categories: wind power aggregated equivalent algorithm, photovoltaic aggregated equivalent algorithm, load aggregated equivalent algorithm, energy storage aggregated equivalent algorithm, and reactive power aggregated equivalent algorithm. The execution steps of the five categories of algorithms include:

[0018] The wind power aggregation equivalent algorithm is executed on all distributed wind power within the distribution network area to obtain the active and reactive power data of the wind power aggregation equivalent object.

[0019] Execute the photovoltaic aggregation equivalent algorithm on all distributed photovoltaics within the distribution network area to obtain the active and reactive power data of the photovoltaic aggregation equivalent object;

[0020] Perform a load aggregation equivalent algorithm on all flexible user loads within the distribution network area to obtain the active and reactive power data of the load aggregation equivalent object;

[0021] Perform an energy storage aggregation equivalent algorithm on all household and energy storage within the distribution network area to obtain the active and reactive power data of the energy storage aggregation equivalent object;

[0022] Perform reactive power aggregation equivalent algorithm on all reactive power compensation devices within the distribution network area to obtain the active and reactive power data of the reactive power aggregation equivalent object;

[0023] The power data of the above five types of aggregated equivalent objects are summarized to form a complete aggregated equivalent power result for the distribution network area, which is used for subsequent internal power loss calculation.

[0024] The process of determining the power equivalence algorithm for the aggregated equivalent objects includes:

[0025] Aggregate and equate the distributed wind power, distributed photovoltaic, flexible user load, household and distribution area energy storage, and reactive power compensation devices within the distribution network area to obtain the corresponding aggregated and equated objects;

[0026] Based on the active and reactive power data from historical measurement data, power aggregation and equivalence are performed on each aggregated equivalence object to obtain the corresponding power equivalence algorithm.

[0027] The process of determining the internal power loss algorithm of the distribution network area includes:

[0028] Aggregate and equalize each of the aforementioned aggregated equivalent objects and the grid connection points within the distribution network area to obtain internal power loss objects;

[0029] Based on the active power data and reactive power data of the grid connection point in the historical measurement data, and the power equivalence algorithm corresponding to each aggregated equivalence object, the internal power loss object is subjected to power aggregation equivalence to obtain the internal power loss algorithm.

[0030] The formula for the internal power loss algorithm is as follows:

[0031]

[0032] in, For internal power loss objects in Active power data at any given time. For internal power loss objects in Reactive power data at any given time. For grid connection point Active power data at any given time. For grid connection point Reactive power data at any given time. Wind power aggregation equivalent object Active power data at any given time. Wind power aggregation equivalent object Reactive power data at any given time. Photovoltaic polymer equivalent objects Active power data at any given time. Photovoltaic polymer equivalent objects Reactive power data at any given time. Equivalent objects for flexible load aggregation Active power data at any given time. Equivalent objects for flexible load aggregation Reactive power data at any given time. For energy storage aggregation equivalent objects Active power data at any given time. For energy storage aggregation equivalent objects Reactive power data at any given time. Aggregate equivalent objects for reactive power compensation Active power data at any given time. Aggregate equivalent objects for reactive power compensation Reactive power data at any given time.

[0033] The distribution network power loss calculation model based on long short-term memory network enables accurate prediction of future power loss in distribution network areas, solving the problem of missing future operation data.

[0034] The real-time measurement data includes: active power, reactive power, and voltage amplitude of distributed wind power, distributed photovoltaic power, flexible loads, household and substation energy storage, reactive power compensation devices, busbars and grid connection points within the distribution network area.

[0035] The predicted power loss data includes: predicted active power loss data and predicted reactive power loss data of distributed wind power, distributed photovoltaic, flexible loads, energy storage devices and reactive power compensation devices within the distribution network area.

[0036] The coordinated regulation of power generation, grid, load, and storage in the distribution network area includes:

[0037] During peak power loss periods in the transformer substation area, dispatch energy storage discharge to support the voltage of the substation area and guide flexible loads to stagger peak power consumption;

[0038] During periods of low power loss, the system schedules energy storage charging, optimizes the output plan of distributed power sources, and adjusts the output of reactive power compensation devices in real time to achieve reactive power balance in the distribution area and reduce active power loss.

[0039] By employing a flexible control strategy that coordinates power generation, grid, load, and storage, precise control of distribution network load can be achieved, effectively reducing power loss in distribution transformer areas.

[0040] Secondly, the present invention provides a distribution network control system for a new type of power system, comprising:

[0041] The measurement data acquisition module is used to acquire real-time measurement data of the distribution substation within a designated area of ​​the new power system.

[0042] The model training module is used to generate training samples based on historical measurement data and the distribution network area aggregated equivalent model, train the long short time memory network, and obtain the area power loss calculation model.

[0043] The power loss prediction module is used to perform extrapolation and prediction based on the real-time measurement data and the trained power loss calculation model of the distribution network area to obtain the predicted power loss data of the distribution network area.

[0044] The coordinated control module is used to formulate and execute a coordinated control strategy for power generation, grid, load and storage based on the predicted power loss data.

[0045] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0046] This invention acquires real-time measurement data of distribution transformer substations and uses a trained substation power loss calculation model to extrapolate and predict power loss, enabling accurate and forward-looking prediction of power loss in distribution transformer substations and providing reliable data support for subsequent regulation. Attached Figure Description

[0047] Figure 1 This is a schematic flowchart of the distribution network control method for a new type of power system shown in Embodiment 1 of the present invention;

[0048] Figure 2 This is a schematic diagram of the collaborative operation of the distribution network control system modules for a new type of power system, as shown in Embodiment 2 of the present invention. Detailed Implementation

[0049] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0050] The term "and / or" simply describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0051] Example 1

[0052] This embodiment introduces a distribution network control method for a new type of power system, including:

[0053] Acquire real-time measurement data of distribution substations within a designated area of ​​the new power system;

[0054] Based on real-time measurement data, the predicted power loss data of the distribution network area is obtained by extrapolating and predicting using the trained power loss calculation model of the distribution area.

[0055] Based on the predicted power loss data, coordinated regulation of power generation, grid, load and storage in the distribution network area is carried out.

[0056] The training of the power loss calculation model for the transformer substation includes:

[0057] Based on historical measurement data of distribution network areas, historical power equivalent data are obtained using the distribution network area aggregation equivalent model;

[0058] Using historical power equivalent data as input and historical power loss data as labels, a long short-term memory network is trained to obtain a trained power loss calculation model for the transformer area.

[0059] The distribution network area aggregation equivalent model includes: the power equivalent algorithm and the internal power loss algorithm of the aggregation equivalent object of the distribution network area;

[0060] Aggregated equivalent objects are adapted to the characteristics of distributed power sources, flexible loads, and diverse access of energy storage devices in new power systems, including five types of objects: wind power, photovoltaic, load, energy storage, and reactive power compensation devices.

[0061] The power equivalence algorithms for aggregated equivalent objects include five categories: wind power aggregated equivalent algorithm, photovoltaic aggregated equivalent algorithm, load aggregated equivalent algorithm, energy storage aggregated equivalent algorithm, and reactive power aggregated equivalent algorithm. The execution steps of these five types of algorithms include:

[0062] Perform the wind power aggregation equivalent algorithm on all distributed wind power within the distribution network area to obtain the active and reactive power data of the wind power aggregation equivalent object;

[0063] Execute the photovoltaic aggregation equivalent algorithm on all distributed photovoltaics within the distribution network area to obtain the active and reactive power data of the photovoltaic aggregation equivalent object;

[0064] Perform a load aggregation equivalent algorithm on all flexible user loads within the distribution network area to obtain the active and reactive power data of the load aggregation equivalent object;

[0065] Perform an energy storage aggregation equivalent algorithm on all household and energy storage within the distribution network area to obtain the active and reactive power data of the energy storage aggregation equivalent object;

[0066] Perform reactive power aggregation equivalent algorithm on all reactive power compensation devices within the distribution network area to obtain the active and reactive power data of the reactive power aggregation equivalent object;

[0067] The power data of the above five types of aggregated equivalent objects are summarized to form a complete aggregated equivalent power result for the distribution network area, which is used for subsequent internal power loss calculation.

[0068] The process of determining the power equivalence algorithm for aggregated equivalent objects includes:

[0069] Aggregate and equate the distributed wind power, distributed photovoltaic, flexible user load, household and distribution area energy storage, and reactive power compensation devices within the distribution network area to obtain the corresponding aggregated and equated objects;

[0070] Based on the active and reactive power data from historical measurement data, power aggregation and equivalence are performed on each aggregated equivalence object to obtain the corresponding power equivalence algorithm.

[0071] The process of determining the internal power loss algorithm for a distribution network area includes:

[0072] Aggregate and equalize each aggregated equivalent object and the grid connection point within the distribution network area to obtain the internal power loss object;

[0073] Based on the active power data and reactive power data of the grid connection point in the historical measurement data, as well as the power equivalence algorithm corresponding to each aggregated equivalence object, the internal power loss object is aggregated and equivalenced to obtain the internal power loss algorithm.

[0074] The formula for the internal power loss algorithm is as follows:

[0075]

[0076] in, For internal power loss objects in Active power data at any given time. For internal power loss objects in Reactive power data at any given time. For grid connection point Active power data at any given time. For grid connection point Reactive power data at any given time. Wind power aggregation equivalent object Active power data at any given time. Wind power aggregation equivalent object Reactive power data at any given time. Photovoltaic polymer equivalent objects Active power data at any given time. Photovoltaic polymer equivalent objects Reactive power data at any given time. Equivalent objects for flexible load aggregation Active power data at any given time. Equivalent objects for flexible load aggregation Reactive power data at any given time. For energy storage aggregation equivalent objects Active power data at any given time. For energy storage aggregation equivalent objects Reactive power data at any given time. Aggregate equivalent objects for reactive power compensation Active power data at any given time. Aggregate equivalent objects for reactive power compensation Reactive power data at any given time.

[0077] Real-time measurement data includes: active power, reactive power, and voltage amplitude of distributed wind power, distributed photovoltaic power, flexible loads, household and substation energy storage, reactive power compensation devices, busbars and grid connection points within the distribution network area.

[0078] Predicted power loss data includes: predicted active power loss data and predicted reactive power loss data for distributed wind power, distributed photovoltaic, flexible loads, energy storage devices and reactive power compensation devices within the distribution network area.

[0079] Coordinated regulation of power generation, grid, load, and storage in distribution network areas includes:

[0080] During peak power loss periods in the transformer substation area, dispatch energy storage discharge to support the voltage of the substation area and guide flexible loads to stagger peak power consumption;

[0081] During periods of low power loss, the system schedules energy storage charging, optimizes the output plan of distributed power sources, and adjusts the output of reactive power compensation devices in real time to achieve reactive power balance in the distribution area and reduce active power loss.

[0082] Specifically, such as Figure 1 As shown, the distribution network control method for new power systems includes the following steps:

[0083] Step 1: Obtain real-time measurement data of distribution substations within the designated area of ​​the new power system. Process the measurement data for positive and negative values ​​based on the power inflow power. After sorting the data according to the same time series, use at least one of the box plot method, interpolation method, and time series prediction method to remove or reconstruct outliers to obtain the final effective real-time measurement data.

[0084] Step 2: Based on historical measurement data of distribution network areas, construct an aggregated equivalent model of distribution network areas covering five types of aggregated equivalent objects: wind power, photovoltaic, flexible load, energy storage, and reactive power compensation devices. Use this model to calculate historical power equivalent data and historical power loss data. Using historical power equivalent data as input and historical power loss data as output, train a Long Short-Time Memory (LSTM) network to obtain the distribution network power loss calculation model.

[0085] Step 3: Input the effective real-time measurement data obtained in Step 1 into the trained power loss calculation model of the distribution network area, perform extrapolation and prediction, and obtain the predicted active power loss data and predicted reactive power loss data of wind power, photovoltaic, flexible load, energy storage and reactive power compensation devices within the distribution network area.

[0086] Step 4: Based on the predicted power loss data obtained in Step 3, formulate a flexible control strategy for source-grid-load-storage coordination. By adjusting the output of distributed power sources, scheduling energy storage charging and discharging, guiding flexible loads to shift peaks, and controlling the output of reactive power compensation devices, the distribution network load in the designated area can be precisely controlled to improve the reliability of the distribution network power supply.

[0087] Specifically, in step 2, when training the Long Short-Term Memory network, a time window of 4 hours and a time granularity of 5 minutes are set. Historical operating data are continuously sorted out and the model is iteratively optimized until the model prediction accuracy meets the requirements of the new power system distribution network control project.

[0088] Specifically, in step 3, the prediction results of the distribution network power loss calculation model are full-dimensional power loss data of the distribution network distribution area at the time granularity of 4 hours and 5 minutes in the future, realizing the rapid generation of future operating sequence data of the distribution network distribution area.

[0089] Example 2

[0090] Based on the same inventive concept as Embodiment 1, this embodiment introduces a distribution network control system for a new type of power system, comprising:

[0091] The measurement data acquisition module is used to acquire real-time measurement data of the distribution substation within a designated area of ​​the new power system.

[0092] The model training module is used to generate training samples based on historical measurement data and the distribution network area aggregated equivalent model, train the long short time memory network, and obtain the area power loss calculation model.

[0093] The power loss prediction module is used to extrapolate and predict the power loss of the distribution network area based on real-time measurement data and the trained power loss calculation model of the distribution area.

[0094] The coordinated control module is used to formulate and execute coordinated control strategies for power generation, grid, load, and storage by combining predicted power loss data.

[0095] Specifically, such as Figure 2As shown, the distribution network control system for the new power system includes a measurement data acquisition module, a model building and data preprocessing module, a power loss prediction module, and a flexible control and reliability improvement module. These modules work collaboratively to achieve the full-process execution of the aforementioned methods. Specifically, the measurement data acquisition module is responsible for collecting and transmitting multi-source real-time measurement data; the model building and data preprocessing module completes data cleaning and training of the power loss calculation model for distribution areas; the power loss prediction module accurately predicts the future power loss of distribution network areas; and the flexible control and reliability improvement module formulates and executes source-grid-load-storage coordinated control strategies and can feed back control effect data to the model building and data preprocessing module, enabling continuous iterative optimization of the model and forming a closed-loop control system.

[0096] This system achieves accurate acquisition of distribution network measurement data, efficient prediction of power loss, and flexible load control through the collaborative work of its various modules. It adapts to the multi-dimensional interactive operation characteristics of the new power system, which integrates power generation, grid, load, and storage, and provides system support for improving the flexible control capability and reliability of the distribution network.

[0097] Example 3

[0098] This embodiment introduces a distribution network control method for a new type of power system, including the following steps:

[0099] Step 1: Obtain the current measurement data of the distribution substations within the designated area of ​​the new power system;

[0100] Step 2: Based on the current measurement data, use the trained distribution area power loss calculation model to perform extrapolation and prediction to obtain the predicted power loss data of the distribution network distribution area; wherein, the distribution area power loss calculation model is obtained by training the long short time memory network based on historical measurement data and the distribution network distribution area aggregated equivalent model.

[0101] Step 3: Utilize predicted power loss data to formulate a source-grid-load-storage coordinated control strategy, flexibly control the load in the designated area, and improve the reliability of power distribution network supply.

[0102] In step 1 above, the current measurement data is the ledger and operation data of distributed wind power, distributed photovoltaic, flexible user loads (including electric vehicle charging piles), household energy storage, reactive power compensation devices, busbars and other equipment within the designated area; among them, the measurement data types of distributed wind power, distributed photovoltaic, flexible loads, household energy storage and reactive power compensation devices are active and reactive power data, the busbars are voltage amplitude data, and the grid connection points are active and reactive power data.

[0103] Specifically, step 2 includes a data preprocessing step before the extrapolation and prediction:

[0104] Step 2.1: Using the power inflow as a reference, process the positive and negative values ​​of multiple measurement data in the current measurement data to obtain multiple negative measurement data;

[0105] Step 2.2: Sort multiple negative measurement data according to the same time series to obtain the measurement data sorting result;

[0106] Step 2.3: Remove outliers or reconstruct data from the sorted measurement data. Use at least one algorithm from box plot method, interpolation method and time series prediction method to process outliers and obtain the final current measurement data.

[0107] Through the above data preprocessing, data such as active power, reactive power, and voltage amplitude of equipment and grid connection points within the distribution network area can be obtained in the same time series, as well as active and reactive power loss data within the area. This provides a reliable data foundation for the establishment of aggregated equivalent models of distribution network areas and the training of power loss calculation models for distribution areas.

[0108] Specifically, to meet the flexible control requirements of distribution networks in designated areas (exemplarily county-level areas) under the new power system, and to achieve simplified processing of distributed power sources and flexible loads, accurate consideration of power losses within distribution areas, and rapid generation of future operational sequence data, a distribution network distribution area aggregation equivalent model is established, including equivalent objects of wind power, photovoltaics, loads, energy storage, reactive power aggregation connected to the main grid of the new power system, and internal power loss objects. The model structure is as follows: Figure 2 As shown.

[0109] Specifically, the training process of the power loss calculation model for the transformer substation in step 2 includes the following steps:

[0110] Step a: Based on the historical measurement data of the distribution network area, use the distribution network area aggregated equivalent model to obtain the historical power equivalent data;

[0111] Step b: Based on historical power equivalent data, use the distribution network area aggregated equivalent model to obtain historical power loss data;

[0112] Step c: Using historical power equivalent data and historical power loss data, train the Long Short Time Memory (LSTM) network to obtain the power loss calculation model for the transformer area.

[0113] The distribution network area aggregation equivalent model includes power equivalence algorithms for aggregation equivalent objects and internal power loss algorithms. The power equivalence algorithms for aggregation equivalent objects cover wind power, photovoltaic, load, energy storage, and reactive power aggregation equivalent algorithms. Each algorithm achieves aggregation equivalent by summing the active and reactive power data of the corresponding equipment within the distribution area. All algorithms are directly connected to the 10kV bus and are processed by PQ nodes in steady-state power flow calculation.

[0114] The internal power loss algorithm calculates power loss based on the principle of power balance. It subtracts the sum of the active and reactive power data of each aggregated equivalent object at the corresponding time point from the active and reactive power data of the grid-connected transformer area to obtain the internal active and reactive power loss data. The expression is as follows:

[0115]

[0116] in, For internal power loss objects in Active power data at any given time. For internal power loss objects in Reactive power data at any given time. For grid connection point Active power data at any given time. For grid connection point Reactive power data at any given time. Wind power aggregation equivalent object Active power data at any given time. Wind power aggregation equivalent object Reactive power data at any given time. Photovoltaic polymer equivalent objects Active power data at any given time. Photovoltaic polymer equivalent objects Reactive power data at any given time. Equivalent objects for flexible load aggregation Active power data at any given time. Equivalent objects for flexible load aggregation Reactive power data at any given time. For energy storage aggregation equivalent objects Active power data at any given time. For energy storage aggregation equivalent objects Reactive power data at any given time. Aggregate equivalent objects for reactive power compensation Active power data at any given time. Aggregate equivalent objects for reactive power compensation Reactive power data at any given time.

[0117] Specifically, the predicted power loss data in step 2 includes the predicted active and reactive power loss data of distributed wind power, distributed photovoltaic, flexible loads, household energy storage and reactive power compensation devices within the distribution network area; by collecting real-time measurement data of the distribution area or ultra-short-term forecast plan data of internal equipment, and combining the characteristics of each equipment to calculate the intraday future state loss data, after inputting it into the power loss calculation model of the distribution area, the future state operation sequence data that meets the flexible control of the distribution network of the new power system is quickly generated.

[0118] By utilizing the aforementioned predicted power loss data, and through source-grid-load-storage coordinated regulation of distributed power generation output optimization, energy storage charging and discharging scheduling, flexible load peak shifting, and precise reactive power compensation adjustment, flexible regulation of distribution network load can be achieved, reducing distribution network power loss and improving the reliability of power supply in the new power system.

[0119] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0120] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0121] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0122] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0123] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A distribution network area control method for a new type of power system, characterized in that, include: Acquire real-time measurement data of distribution substations within a designated area of ​​the new power system; Based on the real-time measurement data, the predicted power loss data of the distribution network area is obtained by using the trained power loss calculation model of the distribution area. Based on the predicted power loss data, coordinated regulation of power generation, grid, load and storage in the distribution network area is carried out. The training of the power loss calculation model for the transformer substation includes: Based on the historical measurement data of the distribution network area, the historical power equivalent data is obtained by using the distribution network area aggregation equivalent model; Using the historical power equivalent data as input and the historical power loss data as labels, the Long Short Time Memory Network is trained to obtain the trained power loss calculation model for the transformer area.

2. The distribution network control method for new power systems according to claim 1, characterized in that, The distribution network area aggregated equivalent model includes: the power equivalent algorithm and the internal power loss algorithm of the aggregated equivalent object of the distribution network area; The aggregated equivalent objects are adapted to the characteristics of distributed power sources, flexible loads, and multiple access to energy storage devices in new power systems, including five types of objects: wind power, photovoltaic, load, energy storage, and reactive power compensation devices.

3. The distribution network control method for new power systems according to claim 2, characterized in that, The power equivalence algorithms for the aggregated equivalent objects include five categories: wind power aggregated equivalent algorithm, photovoltaic aggregated equivalent algorithm, load aggregated equivalent algorithm, energy storage aggregated equivalent algorithm, and reactive power aggregated equivalent algorithm. The execution steps of the five categories of algorithms include: The wind power aggregation equivalent algorithm is executed on all distributed wind power within the distribution network area to obtain the active and reactive power data of the wind power aggregation equivalent object. Execute the photovoltaic aggregation equivalent algorithm on all distributed photovoltaics within the distribution network area to obtain the active and reactive power data of the photovoltaic aggregation equivalent object; Perform a load aggregation equivalent algorithm on all flexible user loads within the distribution network area to obtain the active and reactive power data of the load aggregation equivalent object; Perform an energy storage aggregation equivalent algorithm on all household and energy storage within the distribution network area to obtain the active and reactive power data of the energy storage aggregation equivalent object; Perform reactive power aggregation equivalent algorithm on all reactive power compensation devices within the distribution network area to obtain the active and reactive power data of the reactive power aggregation equivalent object; The power data of the above five types of aggregated equivalent objects are summarized to form a complete aggregated equivalent power result for the distribution network area, which is used for subsequent internal power loss calculation.

4. The distribution network control method for new power systems according to claim 2, characterized in that, The process of determining the power equivalence algorithm for the aggregated equivalent objects includes: Aggregate and equate the distributed wind power, distributed photovoltaic, flexible user load, household and distribution area energy storage, and reactive power compensation devices within the distribution network area to obtain the corresponding aggregated and equated objects; Based on the active and reactive power data from historical measurement data, power aggregation and equivalence are performed on each aggregated equivalence object to obtain the corresponding power equivalence algorithm.

5. The distribution network control method for new power systems according to claim 2, characterized in that, The process of determining the internal power loss algorithm of the distribution network area includes: Aggregate and equalize each of the aforementioned aggregated equivalent objects and the grid connection points within the distribution network area to obtain internal power loss objects; Based on the active power data and reactive power data of the grid connection point in the historical measurement data, and the power equivalence algorithm corresponding to each aggregated equivalence object, the internal power loss object is subjected to power aggregation equivalence to obtain the internal power loss algorithm.

6. The distribution network control method for new power systems according to claim 5, characterized in that, The formula for the internal power loss algorithm is as follows: in, For internal power loss objects in Active power data at any given time. For internal power loss objects in Reactive power data at any given time. For grid connection point Active power data at any given time. For grid connection point Reactive power data at any given time. Wind power aggregation equivalent object Active power data at any given time. Wind power aggregation equivalent object Reactive power data at any given time. Photovoltaic polymer equivalent objects Active power data at any given time. Photovoltaic polymer equivalent objects Reactive power data at any given time. Equivalent objects for flexible load aggregation Active power data at any given time. Equivalent objects for flexible load aggregation Reactive power data at any given time. For energy storage aggregation equivalent objects Active power data at any given time. For energy storage aggregation equivalent objects Reactive power data at any given time. Aggregate equivalent objects for reactive power compensation Active power data at any given time. Aggregate equivalent objects for reactive power compensation Reactive power data at any given time.

7. The distribution network control method for new power systems according to claim 1, characterized in that, The real-time measurement data includes: active power, reactive power, and voltage amplitude of distributed wind power, distributed photovoltaic power, flexible loads, household and substation energy storage, reactive power compensation devices, busbars and grid connection points within the distribution network area.

8. The distribution network control method for new power systems according to claim 1, characterized in that, The predicted power loss data includes: predicted active power loss data and predicted reactive power loss data of distributed wind power, distributed photovoltaic, flexible loads, energy storage devices and reactive power compensation devices within the distribution network area.

9. The distribution network control method for a new type of power system according to claim 1, characterized in that, The coordinated regulation of power generation, grid, load, and storage in the distribution network area includes: During peak power loss periods in the transformer substation area, dispatch energy storage discharge to support the voltage of the substation area and guide flexible loads to stagger peak power consumption; During periods of low power loss, the system schedules energy storage charging, optimizes the output plan of distributed power sources, and adjusts the output of reactive power compensation devices in real time to achieve reactive power balance in the distribution area and reduce active power loss.

10. A distribution network control system for a new type of power system, characterized in that, include: The measurement data acquisition module is used to acquire real-time measurement data of the distribution substation within a designated area of ​​the new power system. The model training module is used to generate training samples based on historical measurement data and the distribution network area aggregated equivalent model, train the long short time memory network, and obtain the area power loss calculation model. The power loss prediction module is used to perform extrapolation and prediction based on the real-time measurement data and the trained power loss calculation model of the distribution network area to obtain the predicted power loss data of the distribution network area. The coordinated control module is used to formulate and execute a coordinated control strategy for power generation, grid, load and storage based on the predicted power loss data.

Images