Peak-valley coordination method and device based on twin power distribution

Abstract

The present invention discloses a peak-valley coordination method and device based on twin distribution, which belongs to the field of power system optimization and intelligent control, wherein the method includes: interactively pre-managing the digital distribution twin system of the area and performing coding and labeling, executing system functional decomposition and label transfer; building a coordination decision model based on the sub-order twin system; combining the consumption prediction branch to predict the consumption trend on the demand side, the task identifier has a time-phased peak-valley span, and the prediction branch and the sub-order twin system establish a communication connection; transmitting the peak-valley coordination task to the coordination decision model, and performing bilateral decision fitting on the demand side and the distribution side; based on the label transfer relationship, based on the digital distribution twin system simulation verification target coordination strategy, and responding to the distribution system for peak-valley coordination management. This application solves the technical problem in the prior art that the power system is difficult to efficiently manage and optimize scheduling during peak and valley load periods, resulting in supply and demand imbalance and energy waste.

Description

Technical Field

[0001] The present invention relates to the field of power system optimization and intelligent control, and in particular to a peak-valley coordination method and device based on twin power distribution. Background Art

[0002] With the continuous increase in global energy demand and the rapid development of renewable energy, the stability and efficiency of the power system are facing new challenges. Traditional power systems mainly rely on centralized power generation and one-way transmission, which is difficult to adapt to the access and use of modern diversified and distributed energy. Especially during the peak and valley periods of power load, the supply and demand balance of the power grid is more complicated, and a more flexible and intelligent power dispatching and management method is urgently needed to improve the stability, reliability and economy of the power grid.

[0003] At present, the distribution system is gradually developing in the direction of digitalization and intelligence, and the application of digital twin technology in the distribution system is becoming more and more extensive. Digital twin technology realizes real-time monitoring, simulation analysis and optimization control of the power grid by building a virtual model of the physical power grid. However, in practical applications, digital twin technology still faces problems such as data islands, poor information exchange, and insufficient prediction and analysis accuracy. Especially in the coordination of power load peaks and valleys, the existing technology is difficult to effectively integrate the data and resources on the demand side and the power supply side, resulting in inaccurate and inflexible scheduling decisions, and unable to fully utilize the potential of distributed energy and energy storage facilities. Summary of the invention

[0004] The present application provides a peak-valley coordination method and device based on twin distribution, aiming to solve the technical problem in the prior art that the power system is difficult to efficiently manage and optimize during peak and valley load periods, thereby leading to supply and demand imbalance and energy waste.

[0005] In view of the above problems, the present application provides a peak-valley coordination method and device based on twin power distribution.

[0006] The first aspect disclosed in the present application provides a peak-valley coordination method based on twin distribution, which includes interacting with a digital distribution twin system of a pre-managed area and performing coding and labeling, executing system functional decomposition and label transfer, and determining a sub-order twin system, wherein the function is oriented towards peak-valley coordination; based on the sub-order twin system, building a coordination decision model, wherein the coordination decision model includes parallel demand side branches and distribution side branches, and the branches include three layers of fully connected identification and analysis layers, greedy decision layers and integrated output layers; combining with the consumption prediction branch, predicting the consumption trend of the demand side, and determining the peak-valley coordination task, wherein the task identifier has a time-phased peak-valley span, and the prediction branch establishes a communication connection with the sub-order twin system; transmitting the peak-valley coordination task to the coordination decision model, performing bilateral decision fitting on the demand side and the distribution side, and determining the target coordination strategy; based on the label transfer relationship, the target coordination strategy is simulated and verified based on the digital distribution twin system, and the distribution system is responded to for peak-valley coordination management.

[0007] Another aspect disclosed in the present application provides a peak-valley coordination device based on twin power distribution, which includes a sub-order twin system determination module: a digital power distribution twin system for interactive pre-management area and coding and labeling, system functional decomposition and label transfer, and determination of the sub-order twin system, wherein the function is oriented to peak-valley coordination; a coordination decision model building module: a coordination decision model for building a coordination decision model based on the sub-order twin system, wherein the coordination decision model includes parallel demand-side branches and distribution-side branches, and the branches include a three-layer fully connected identification and analysis layer, a greedy decision layer, and an integrated output layer; peak-valley coordination Task determination module: used to combine with the consumption prediction branch to predict the consumption trend on the demand side and determine the peak-valley coordination task, wherein the task identifier has a time-phased peak-valley span, and the prediction branch establishes a communication connection with the sub-order twin system; target coordination strategy determination module: used to transmit the peak-valley coordination task to the coordination decision model, perform bilateral decision fitting on the demand side and the distribution side, and determine the target coordination strategy; target coordination strategy verification module: used to verify the target coordination strategy based on the label transfer relationship and the digital distribution twin system, and respond to the distribution system for peak-valley coordination management.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The adoption of a digital distribution twin system for interactive pre-management, coding and labeling, system functional decomposition and label transfer, the establishment of a parallel demand-side and distribution-side coordination decision-making model, and the technical solution combined with the consumption prediction branch solves the technical problem of the existing technology that the power system is difficult to efficiently manage and optimize during peak and valley load periods, resulting in an imbalance between supply and demand and energy waste. Through this solution, dynamic monitoring and optimization control of the power system are achieved, ensuring effective supply and demand adjustments during peak and valley load periods, improving the stability and efficiency of power supply, reducing energy waste, and significantly improving the utilization rate of power resources, ultimately achieving the technical effect of optimizing power dispatch, improving system operation reliability, and reducing costs.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A flow chart of a peak-valley coordination method based on twin power distribution is provided for an embodiment of the present application.

[0012] Figure 2 A structural schematic diagram of a peak-valley coordination device based on twin power distribution is provided for an embodiment of the present application.

[0013] Explanation of the accompanying drawings: sub-order twin system determination module 11, coordination decision model construction module 12, peak-valley coordination task determination module 13, target coordination strategy determination module 14, target coordination strategy verification module 15. DETAILED DESCRIPTION

[0014] The overall idea of ​​the technical solution provided by this application is as follows:

[0015] The embodiments of the present application provide a peak-valley coordination method and device based on twin distribution. By constructing a digital distribution twin system, the distribution system is monitored and optimized in real time. First, the system is coded and labeled, functionally decomposed, and labels are transferred to form a sub-order twin system. Then, a coordination decision model including demand-side and distribution-side branches is built, and the demand-side power consumption trend is predicted in combination with the consumption prediction branch to determine the peak-valley coordination task. Through bilateral decision fitting, the target coordination strategy is determined, and a simulation verification is performed based on the label transfer relationship, and finally, peak-valley coordination management is performed in the actual distribution system to ensure supply and demand balance and efficient operation of the system.

[0016] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0017] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides a peak-valley coordination method based on twin power distribution, the method comprising:

[0018] Step S100: Interact with the digital distribution twin system of the pre-management area and perform coding and labeling, perform system functional decomposition and label transfer, and determine the sub-level twin system, where the function is guided by peak-valley coordination.

[0019] Specifically, the pre-management area refers to a specific area in the power distribution system that is managed and controlled in advance through digital means. All equipment and resources in this area are included in the digital management system. System functional decomposition refers to the decomposition of complex power distribution system functions into several sub-functional modules for more refined and targeted management and optimization. Label transfer refers to the transfer of assigned labels from one functional module or node to another to achieve coordination and unified management between different functional modules.

[0020] Specifically, we first establish a digital distribution twin system for interactive pre-management areas. This system acquires and analyzes data from the distribution system in real time to build a virtual model that can truly reflect the operating status of the physical distribution system. To facilitate management and optimization, all devices and nodes in the system are coded and labeled, that is, uniformly coded and labeled to achieve accurate identification and tracking.

[0021] Next, the system functional decomposition is performed to decompose the complex functions of the power distribution system into several sub-functional modules, such as monitoring, analysis, and control. Through this decomposition, each functional module can be managed more meticulously to improve the overall efficiency of the system. At the same time, label transfer is performed to transfer the assigned labels from one functional module or node to another module to achieve coordination and unified management between different functional modules.

[0022] Finally, determine the sub-level twin systems. These sub-level twin systems are divided according to different functions or levels, and each sub-level system focuses on a specific function. For example, peak-valley coordination is an important sub-level system, through which the dynamic adjustment and optimization of power loads are achieved. This step uses methods and tools such as digital twin technology, real-time data analysis, coding and tag management to ensure efficient and stable operation of the distribution system and achieve the best match between supply and demand.

[0023] Step S200: Based on the sub-order twin system, a coordinated decision model is constructed, wherein the coordinated decision model includes parallel demand-side branches and distribution-side branches, and the branches include three layers of fully connected identification and analysis layers, greedy decision layers, and integrated output layers.

[0024] Specifically, the demand-side branch is part of the coordinated decision-making model, focusing on analyzing and optimizing the behavior and needs of power demanders. The distribution-side branch is another part of the coordinated decision-making model, focusing on analyzing and optimizing the power supply behavior and strategies of the distribution system. The identification and analysis layer is the first layer in the coordinated decision-making model, responsible for identifying and analyzing data and information on the demand side and the distribution side. The greedy decision layer is the second layer in the coordinated decision-making model, which uses a greedy algorithm to make local optimal decisions to gradually approach the global optimal solution. The integrated output layer is the third layer in the coordinated decision-making model, which integrates the results of the identification and analysis layer and the greedy decision layer and outputs the final decision and strategy.

[0025] Specifically, firstly, a complete decision model, namely a coordinated decision model, is constructed by using the data and functional modules of the sub-order twin system. The coordinated decision model is divided into a demand side branch and a distribution side branch, and the two run in parallel.

[0026] The demand-side branch focuses on analyzing and optimizing the behavior and needs of power demanders. It includes three layers: identification and analysis layer, greedy decision layer, and integrated output layer. The identification and analysis layer is responsible for obtaining and identifying relevant data from the demand side, such as the user's power consumption pattern and load demand. The greedy decision layer uses a greedy algorithm to make local optimal decisions based on these data, and gradually optimizes the power consumption strategy on the demand side. The integrated output layer integrates the analysis and decision results and outputs the final demand-side optimization strategy.

[0027] The distribution side branch focuses on analyzing and optimizing the power supply behavior and strategy of the distribution system. It also includes the identification analysis layer, greedy decision layer and integrated output layer. The identification analysis layer obtains and identifies relevant data from the distribution system, such as grid load, distribution network structure and power supply capacity. The greedy decision layer uses a greedy algorithm to make local optimal decisions based on these data and gradually optimizes the power supply strategy of the distribution system. The integrated output layer integrates the analysis and decision results and outputs the final distribution side optimization strategy.

[0028] Through the operation of these two parallel branches, the coordinated decision model can analyze and optimize both the demand side and the distribution side. In the identification and analysis layer, each branch obtains and analyzes data from its own side to identify key issues and potential optimization points. Then, in the greedy decision layer, each branch gradually optimizes to find the local optimal solution and approach the global optimal solution. Finally, in the integration output layer, each branch integrates its results and outputs a comprehensive optimization decision strategy.

[0029] This step ensures that the optimization of the demand side and the distribution side can be carried out in coordination, and through parallel operation and a three-layer fully connected structure, more accurate and efficient coordinated management of power load peaks and valleys can be achieved. Using the real-time data and functional modules in the sub-order twin system, the coordinated decision model can dynamically adapt to changes in the power system and provide efficient and reliable optimization decisions.

[0030] Step S300: Combined with the consumption prediction branch, the consumption trend on the demand side is predicted, and the peak-valley coordination task is determined, wherein the task identifier has a time-phased peak-valley span, and the prediction branch establishes a communication connection with the sub-order twin system.

[0031] Specifically, the consumption forecasting branch is a functional branch used to predict the power consumption trend on the demand side, helping to determine the changes in power load in different time periods. The peak-valley coordination task refers to determining the coordination tasks that need to be performed during peak and valley periods based on the demand-side consumption trend to optimize power supply and demand. The phase refers to the time characteristics of power consumption, that is, the peak and valley of power consumption in different time periods.

[0032] Specifically, the consumption forecasting branch can identify the consumption patterns and trends by acquiring and analyzing the user's power consumption data in real time. These data include daily power consumption, seasonal changes, emergencies, etc. These data are analyzed using statistical analysis and machine learning algorithms to identify power consumption patterns and trends. Through these analyses, the consumption forecasting branch predicts the changes in power consumption in different time periods in the future and clarifies the peak and trough periods and spans.

[0033] Based on the forecast of demand-side consumption trends, peak-valley coordination tasks are determined. These tasks are identified according to peak and valley periods of electricity consumption, and corresponding coordination strategies are formulated. For example, during peak periods, it is necessary to increase power supply or regulate demand to prevent grid overload; during valley periods, it is necessary to reduce power supply or use energy storage facilities.

[0034] The prediction branch establishes a communication connection with the sub-order twin system to ensure that the prediction data can be updated and transmitted to the coordination decision model in real time. Through this communication connection, the prediction branch can adjust the coordination tasks according to the latest consumption trends and transmit these tasks to the coordination decision model for further processing.

[0035] This step, by combining the consumption prediction branch, can effectively predict the power consumption trend on the demand side, determine the time-phase peak-valley coordination tasks, and achieve efficient management and optimization of power loads through communication connection with the sub-order twin system.

[0036] Step S400: transmitting the peak-valley coordination task to the coordination decision model, performing bilateral decision fitting on the demand side and the distribution side, and determining a target coordination strategy.

[0037] Specifically, after the peak-valley coordination task is determined, it is transmitted to the coordination decision model. The transmission process begins by passing the detailed information of the peak-valley coordination task, including the time phase of the task and the peak-valley span, to each branch of the coordination decision model. After receiving this information, the demand-side branch identifies specific measures for load reduction and increase based on the user's power consumption pattern and demand forecast. At the same time, the distribution-side branch analyzes how to adjust the power generation and distribution strategies in different time periods to meet the changes on the demand side according to the power supply capacity and dispatching strategy of the power grid.

[0038] Next, we enter the bilateral decision fitting stage. The demand-side branch and the distribution-side branch obtain their respective data through the identification and analysis layer, and make local optimal decisions in the greedy decision layer. The demand side optimizes its load management strategy, such as reducing non-essential power loads during peak hours and using energy storage equipment during off-peak hours. The distribution side optimizes its power supply strategy, such as dispatching generators and adjusting the distribution network structure.

[0039] Finally, at the integration output layer, the decision results of the two branches are integrated to form a comprehensive target coordination strategy. This strategy ensures that the demand side and the distribution side can operate in a coordinated manner during peak and trough periods of electricity demand to achieve supply and demand balance and efficient operation of the power system.

[0040] This step ensures the accuracy and timeliness of decision-making. In this way, the coordinated decision-making model can dynamically adjust and optimize the operation strategy of the power system to meet the actual changes in power demand and achieve efficient peak-valley coordinated management.

[0041] Step S500: Based on the label transfer relationship, the target coordination strategy is verified based on the digital distribution twin system simulation, and the distribution system is responded to for peak-valley coordination management.

[0042] Specifically, the label transfer relationship refers to the association and label information transfer relationship between nodes and functional modules in the digital distribution twin system, which is used to track and manage the status and changes of the system.

[0043] First, the target coordination strategy is mapped to the corresponding functional modules and nodes in the digital distribution twin system by using the label transfer relationship. The label transfer relationship ensures that each functional module and node can correctly receive the strategy information related to it and perform corresponding simulation operations in the virtual model.

[0044] Next, a simulation verification is performed in the digital power distribution twin system. This step verifies the effectiveness and feasibility of the target coordination strategy by combining virtual simulation with actual data. Virtual simulation can simulate the power demand and supply in different time periods, detect the performance of the target coordination strategy in various scenarios, and ensure that the strategy can effectively respond to various changes in actual conditions.

[0045] After verification through simulation, the verification results are fed back to the actual power distribution system. According to the verification results, the target coordination strategy is optimized and adjusted to ensure its efficiency and reliability in actual application. Then, the optimized strategy is applied to the power distribution system to implement specific peak-valley coordination management measures.

[0046] During the peak-valley coordination management process, the system monitors power demand and supply in real time and dynamically adjusts strategies based on actual conditions. For example, during peak power demand periods, the system can reduce unnecessary loads and increase the discharge of energy storage equipment; during valley periods, it can increase loads and perform equipment maintenance and other operations. The entire process ensures that the implementation of strategies can efficiently and accurately optimize and balance the power system through the collaborative work of the digital distribution twin system and the label transfer relationship.

[0047] In this step, through the simulated verification and response mechanism based on the label transfer relationship and the digital distribution twin system, the distribution system can achieve dynamic and intelligent peak-valley coordination management and improve the stability and operation efficiency of the power system.

[0048] Furthermore, the coding and labeling includes: traversing the digital distribution twin system to determine the distribution of distribution nodes; traversing the distribution of distribution nodes, and setting a label format based on the union of the island systems running the nodes, wherein the label format has uniformity and node differentiation; based on the label format, coding and labeling the distribution of distribution nodes.

[0049] Specifically, the digital distribution twin system is first traversed to determine the distribution of distribution nodes. By collecting and analyzing the location and related information of each node in the distribution system, the specific location and role of each node in the entire distribution network are clarified.

[0050] Next, for the distribution of distribution nodes, consider the island systems running on each node. These island systems include multiple synergistic system levels such as the device layer and the application layer. For example, the island system of a substation includes transformers (device layer), monitoring software (application layer), and control systems (control layer). The implementation and operation of each distribution node requires the joint action of multiple systems, so when setting the label format, it needs to be determined based on the combination of these systems.

[0051] When setting the label format, it is necessary to ensure the uniformity of the label and the differentiation of the nodes. Uniformity refers to the standardization of the encoding and combination of all labels to ensure consistency and identifiability in the entire distribution system. Node differentiation refers to setting differentiated label formats according to the specific needs and characteristics of different nodes to facilitate tracking and analysis of data from different nodes. For example, a substation labeled "001-TR-01" represents the first-level information of the transformer equipment in the first site, while "001-SC-02" represents the second-level information of the monitoring system in the same site.

[0052] After the label format is set, the distribution nodes are coded and labeled based on these formats. That is, each node is assigned a unique label code and these labels are attached to the corresponding nodes. In this way, all distribution nodes can be accurately identified and managed in the digital distribution twin system.

[0053] This step not only makes the management of the power distribution system more efficient, but also provides a basis for subsequent data tracking and analysis. By tracking the tag information of different nodes, the operating status of each node can be monitored in real time, potential problems can be discovered and solved in a timely manner, and the optimization and efficient operation of the power distribution system can be achieved.

[0054] Furthermore, the functional decomposition and label transfer of the execution system include: determining a coordination information pool based on the peak-valley coordination requirements of the power grid; based on the digital distribution twin system, performing system dimensionality reduction decomposition based on the coordination information pool to determine a first twin system; and performing label transfer on the first twin system to determine the sub-order twin system.

[0055] Specifically, the peak-valley coordination demand of the power grid refers to the coordination measures required by the power grid during the peak and trough periods of electricity demand to ensure the balance between supply and demand. The coordination information pool is a database used to store various information and data related to the peak-valley coordination of the power grid. System dimensionality reduction decomposition refers to decomposing a complex system model into smaller and more manageable parts by simplifying its dimensions. The first twin system is a preliminary virtual model obtained after the system dimensionality reduction decomposition. The sub-order twin system is a more detailed subsystem determined on the basis of the first twin system through operations such as label transfer.

[0056] First, according to the peak-valley coordination needs of the power grid, the coordination information pool is determined. The coordination information pool is a database that stores various information and data related to the peak-valley coordination of the power grid. Including the historical operation data of the power grid, real-time monitoring data, forecast data, etc., it can provide rich data support for subsequent system decomposition and optimization.

[0057] Next, based on the digital distribution twin system and the coordination information pool, the system is decomposed by dimensionality reduction. The purpose of dimensionality reduction is to decompose the complex system model into smaller and more manageable parts by simplifying its dimensions. By analyzing the data in the coordination information pool, the main functional modules and sub-functional modules in the system can be identified. For example, a typical digital distribution twin system includes a power monitoring module, a load forecasting module, a scheduling optimization module, etc. These modules can be further decomposed to form the first twin system.

[0058] After the first twin system is determined, the next step is label transfer. Label transfer is to transfer the assigned labels from one functional module or node to another to achieve coordination and unified management between different functional modules. Specifically, it is to transfer the label information in the initial twin system to the sub-level twin system to ensure that each functional module and node can be accurately identified and managed in the new architecture. For example, the label information of the power monitoring module is transferred from the initial twin system to the load forecasting module and the scheduling optimization module to form a more refined and focused sub-level twin system.

[0059] This step can achieve supply and demand balance in the power grid during peak and trough periods of electricity demand through precise management and coordination of each sub-functional module, thereby improving the efficiency and reliability of power grid operation.

[0060] For example, if the peak-valley coordination requirements of the power grid include load reduction during peak periods and load increase during valley periods, then after determining the coordination information pool, the relevant load management modules and dispatch optimization modules can be identified by analyzing the data. After dimensionality reduction and decomposition, the power monitoring and load forecasting functions in the initial twin system are transferred to these modules to form a sub-order twin system focusing on load management and dispatch optimization. In this way, it can be ensured that each functional module can be accurately managed and coordinated in the new architecture, ultimately achieving efficient peak-valley coordination management.

[0061] Furthermore, constructing a consumption prediction branch includes: reading the power distribution records of a preset time zone, performing periodic mining based on the power dispatch characteristics and the consumption characteristics, and determining the power distribution trend curve; combining the sub-order twin system with the power distribution trend curve to determine the branch architecture; integrating the power distribution records, performing data-driven training on the branch architecture, and generating the consumption prediction branch.

[0062] Specifically, the preset time zone refers to the time range or period pre-set for analysis and prediction. The power dispatch characteristics describe the characteristic data of power distribution and dispatch in the power system, such as load distribution, power generation and power consumption. The consumption characteristics describe the characteristic data of power consumption on the power demand side, such as user power consumption patterns and load curves.

[0063] First, read the power distribution records for the preset time zone. This includes the power system's power dispatch and consumption data for a specific time period. You can use a database management system (such as MySQL or MongoDB) to store and retrieve this data. For example, read the daily power distribution records for the past year, including hourly power generation, power consumption, and load distribution.

[0064] Next, based on the characteristics of power dispatch and consumption, these data are periodically mined to determine the power distribution trend curve. Periodic mining is implemented using data mining tools (such as Pandas and NumPy libraries in Python) to analyze the periodic changes in loads, user power consumption patterns, etc. For example, by analyzing the data to find the peak and valley periods of power consumption every day, a typical daily load curve is formed.

[0065] Combine the sub-order twin system with the distribution trend curve to determine the branch architecture of the consumption prediction branch. The sub-order twin system provides real-time operating status and behavior data of different subsystems in the power system, while the distribution trend curve provides trend information of historical data. Combining the two, the overall architecture of the consumption prediction branch can be determined, that is, the input, output and internal structure of the model. For example, the real-time load data in the sub-order twin system is used as input, and the distribution trend curve is used as a reference to establish a prediction model architecture containing a multi-layer neural network.

[0066] Integrate distribution records and perform data-driven training on the branch architecture. Data-driven training uses machine learning algorithms (such as neural networks, random forests, etc.) to train models with a large amount of historical data to improve the accuracy and reliability of predictions. The training process can be implemented using a machine learning framework (such as TensorFlow or PyTorch), inputting distribution records as training data into the model for repeated iterative optimization. For example, divide the distribution records of the past year into training sets and validation sets, use neural networks for training, and improve prediction accuracy by adjusting model parameters.

[0067] Through this step, the consumption forecast branch is finally generated. The power consumption trend on the demand side can be predicted in real time, providing accurate data support for peak-valley coordination decisions. For example, the power demand changes in the next week can be predicted, and the peak and valley periods can be identified, so as to help formulate corresponding load adjustment and power dispatching strategies and realize efficient management and optimization of the power system.

[0068] Furthermore, the determination of the target coordination strategy includes: identifying the peak-valley coordination task, performing scale transformation and granularity segmentation on the sub-order twin system, and determining multiple sub-areas; decomposing the peak-valley coordination task and determining multiple greedy targets; mapping the multiple sub-areas and the multiple greedy targets, performing distribution coordination to determine multiple coordination strategies; and combining the multiple coordination strategies to determine the target coordination strategy.

[0069] Specifically, the peak-valley coordination task refers to the power supply and demand adjustment task formulated according to the peak and valley periods of power demand. Scale transformation and granularity segmentation refer to dividing the sub-order twin system into multiple sub-areas, each of which corresponds to a specific analysis and optimization task. Scale transformation refers to adjusting the observation and analysis scale of the system, and granularity segmentation refers to refining the system to a smaller granularity level. The greedy goal is a local optimal goal determined based on the greedy algorithm, and the overall optimization is achieved through the combination of multiple greedy goals.

[0070] First, identify the peak-valley coordination tasks. This includes specific measures for load adjustment and power dispatch during peak and valley periods of power demand. Based on the peak-valley coordination tasks, the sub-order twin system is scaled and granularly segmented to divide the system into multiple sub-areas. Scale transformation means adjusting the scope and level of analysis, while granularity segmentation means breaking the system down into smaller units for local optimization analysis. For example, a city's distribution system can be divided into multiple areas, each of which contains several substations and end users.

[0071] Next, the peak-valley coordination task is decomposed to determine multiple greedy targets. The greedy target is a local optimal target determined based on the greedy algorithm, which achieves overall optimization by gradually optimizing the power supply and demand of each sub-region. For example, in a certain area, the greedy target can be to minimize the power load during peak hours or to maximize the energy storage utilization during valley hours.

[0072] Then, multiple sub-areas are mapped to multiple greedy targets, distribution coordination is performed, and multiple coordination strategies are determined. This step involves matching and optimizing the greedy targets of each sub-area with specific power dispatch strategies. For example, substation A in a sub-area needs to reduce non-essential loads during peak hours, while charging of energy storage devices can be increased during off-peak hours.

[0073] Finally, multiple coordination strategies are combined to determine the target coordination strategy. The local optimal strategies of each sub-region are combined to form an overall optimization plan to ensure efficient peak-valley coordination management in the entire distribution system. For example, the load reduction strategy and energy storage utilization strategy of each sub-region are integrated to form a unified dispatching plan to reduce the total load during peak hours and make full use of energy storage equipment during off-peak hours to achieve supply and demand balance and stable operation of the power system.

[0074] Through this step, the process of determining the target coordination strategy ensures the efficient management and optimization of the power system during peak and valley periods, achieving the best match between supply and demand and the stable operation of the power system.

[0075] Furthermore, the distribution coordination determines multiple coordination strategies, including: determining the first coordination strategy based on grid-connected coordination based on real-time power generation as the first coordination method, using energy storage as the second coordination method, and using the grid dispatch distance as the coordination constraint; determining the second coordination strategy driven by demand-side policies; and determining the first coordination strategy for the first sub-area by fitting the first coordination strategy and the second coordination strategy.

[0076] Specifically, grid coordination refers to adjusting the power generation and supply strategies in the power system according to the real-time power generation to ensure the balance of power supply and demand. Demand-side policy drive refers to adjusting power demand according to user needs and policy guidance (such as electricity prices, power consumption restrictions, etc.).

[0077] First, grid coordination is performed based on real-time power generation, and the first coordination mode is determined. Real-time power generation data can be obtained in real time through monitoring equipment in power plants and distribution systems. The purpose of grid coordination is to adjust power generation and supply strategies according to the current power generation situation to ensure a balance between power supply and demand. For example, in a solar power generation system, real-time solar power generation data can be used to adjust the output of other power generation sources to adapt to the volatility of solar power generation.

[0078] Secondly, energy storage is used as the second coordination method. Energy storage equipment (such as battery energy storage systems) can store excess electricity when electricity demand is low and release electricity during peak hours to balance the power load. It can effectively alleviate the fluctuation of power load and provide a stable power supply. For example, when electricity demand is low, wind or solar power is used to charge the battery; during peak hours, the battery is discharged to support the grid demand.

[0079] In addition, the first coordination strategy is determined with the grid dispatch distance as the coordination constraint. The shorter the distance from the power source to the load point, the smaller the transmission loss. Therefore, when coordinating power distribution, it is necessary to consider the grid dispatch distance and optimize the power transmission path and dispatch scheme. For example, power stations that are closer to the load point are given priority to reduce power transmission losses and improve system efficiency.

[0080] Next, the second coordination strategy is determined based on the demand-side policy drive. The power demand is adjusted according to user needs and policy guidelines (such as electricity prices, power consumption restrictions, etc.). For example, during peak hours when electricity prices are high, users are encouraged to reduce electricity consumption or use energy storage devices; during off-peak hours when electricity prices are low, users are encouraged to increase electricity consumption or charge energy storage devices.

[0081] Finally, the first coordination strategy and the second coordination strategy are fitted to determine the first coordination strategy for the first sub-region. The coordination strategy fitting is to comprehensively consider multiple coordination strategies to form a unified coordination plan to ensure the best balance of power supply and demand in the first sub-region. For example, in a certain sub-region, by combining the grid-connected coordination of real-time power generation, energy storage call and demand-side policy drive, a comprehensive power dispatch and load management plan is formulated to ensure efficient operation and stable power supply of the power system.

[0082] This step ensures efficient management and optimization of the power system during peak and off-peak periods by coordinating distribution and determining multiple coordination strategies, achieving the best match between supply and demand and stable operation of the power system.

[0083] Furthermore, the peak-valley coordination management includes: executing and monitoring the target coordination strategy to determine the strategy response data; traversing the strategy response data, combining the simulated verification data to perform deviation measurement of distribution coordination, and determining the response deviation characteristics, wherein the deviation measurement dimensions include response vector deviation and response trend deviation; if the response deviation characteristics do not satisfy the coordination freedom, a feedback adjustment instruction is generated.

[0084] Specifically, deviation metrics refer to indicators that evaluate the difference between actual execution results and expected strategies, including response vector deviation and response trend bias. Response vector deviation refers to the specific difference in numerical values ​​between the actual response and the expected response. For example, if the expected load reduction during peak hours is 500MW, but the actual reduction is only 450MW, then the response vector deviation is 50MW. Response trend bias refers to the deviation direction of the actual response from the expected response in terms of trend. For example, the load during the expected trough period should be gradually increased to utilize excess power generation, but the actual situation is that the load increases too fast or too slow. Coordination freedom refers to the deviation range or flexibility allowed during the strategy adjustment process.

[0085] First, the target coordination strategy is executed and monitored. During the execution process, the system collects strategy response data in real time, including actual operating conditions such as power load, power generation, and energy storage status. For example, the power load of each node and the output data of the power station are monitored in real time through the power monitoring system (such as the SCADA system).

[0086] Next, these strategy response data are traversed and combined with simulation verification data to measure the deviation of distribution coordination. Simulation verification data is data that simulates and verifies the effectiveness and feasibility of the target coordination strategy through the digital twin system. For example, the operation of the power system is simulated using a digital twin platform (such as GE's Predix and Siemens' MindSphere) to verify the performance of the target coordination strategy under various conditions. Deviation measurement refers to the difference between the strategy execution result and the expected result, including response vector deviation and response trend deviation.

[0087] By analyzing the response deviation characteristics, if it is found that these characteristics do not meet the coordination freedom, that is, the flexibility and adaptability of strategy adjustment and optimization are insufficient, then the system needs to generate feedback adjustment instructions. For example, if the actual response deviation is too large, it means that the current strategy cannot effectively balance supply and demand. The system will generate feedback adjustment instructions to adjust the load distribution or power generation through the control system (such as PLC, DCS) to optimize the power dispatch strategy.

[0088] For example, in a city's power distribution system, by real-time monitoring of the power load conditions of each node and the output data of the power station, the system found that the load reduction during a peak period did not reach the expected target. By analyzing the response deviation characteristics, it was found that the response vector deviation was 50MW, and the response trend bias showed that the load reduction speed was too slow. The system generates feedback adjustment instructions, adjusts the discharge rate and load distribution strategy of the energy storage equipment, optimizes power dispatch, and ensures the efficient operation of the system during peak and valley periods.

[0089] The process of peak-valley coordinated management through this step ensures efficient management and optimization of the power system in actual operation, achieves the best match between supply and demand and stable operation of the power system.

[0090] In summary, the peak-valley coordination method based on twin power distribution provided in the embodiment of the present application has the following technical effects:

[0091] 1. Dynamic monitoring and optimization control of the power system are achieved by establishing a digital distribution twin system and conducting peak-valley coordinated management. This method ensures effective supply and demand adjustments during peak and valley periods of power load through real-time data collection and analysis. As a result, the power system operates more stably, power supply efficiency is improved, and the utilization rate of power resources is significantly improved, while reducing energy waste and power dispatch costs.

[0092] 2. The implementation of system functional decomposition and label transfer makes the management of the distribution system more modular and refined. By determining the coordination information pool and performing system dimensionality reduction decomposition, key issues in the system can be more effectively identified and handled. At the same time, the label transfer process ensures coordination and unified management between functional modules, improves the flexibility and adaptability of the system, and helps to achieve more efficient power dispatching and load management.

[0093] 3. Constructing the consumption forecast branch Through periodic mining and analysis of historical and real-time data, accurate forecasting of demand-side power consumption trends is achieved. The data-driven training method is used to improve the accuracy and reliability of the forecasting model. This forecasting capability helps the power system to formulate load adjustment strategies in advance and optimize power dispatching plans, thereby achieving a dynamic balance between supply and demand and efficient operation of the system.

[0094] Embodiment 2 is based on the same inventive concept as the peak-valley coordination method based on twin power distribution in the above embodiment. Figure 2 As shown, an embodiment of the present application provides a peak-valley coordination device based on twin power distribution, the device comprising:

[0095] Sub-level twin system determination module 11: used for interactive pre-management area digital distribution twin system and code labeling, perform system functional decomposition and label transfer, determine the sub-level twin system, where the function is guided by peak-valley coordination;

[0096] Coordination decision model building module 12: used to build a coordination decision model based on the sub-order twin system, wherein the coordination decision model includes parallel demand side branches and distribution side branches, and the branches include three layers of fully connected identification and analysis layers, greedy decision layers, and integrated output layers;

[0097] The peak-valley coordination task determination module 13 is used to combine the consumption prediction branch, predict the consumption trend of the demand side, and determine the peak-valley coordination task, wherein the task identifier has a peak-valley span with a time phase, and the prediction branch establishes a communication connection with the sub-order twin system;

[0098] The target coordination strategy determination module 14 is used to transmit the peak-valley coordination task to the coordination decision model, perform bilateral decision fitting on the demand side and the distribution side, and determine the target coordination strategy;

[0099] The target coordination strategy verification module 15 is used to verify the target coordination strategy based on the label transfer relationship and the digital distribution twin system simulation, and respond to the distribution system to perform peak-valley coordination management.

[0100] Furthermore, the device is also used to perform the following steps:

[0101] Traversing the digital power distribution twin system to determine the distribution of power distribution nodes;

[0102] Traversing the distribution of the power distribution nodes, setting a label format based on a union of island systems running on the nodes, wherein the label format has uniformity and node differentiation;

[0103] Based on the label format, the distribution nodes are coded and labeled.

[0104] Furthermore, the device is also used to perform the following steps:

[0105] Determine the coordination information pool based on the peak-valley coordination needs of the power grid;

[0106] Based on the digital power distribution twin system, a system dimensionality reduction decomposition is performed based on the coordination information pool to determine a first twin system;

[0107] Perform label transfer on the first twin system to determine the sub-order twin system.

[0108] Furthermore, the device is also used to perform the following steps:

[0109] Read the power distribution records of the preset time zone, conduct periodic mining based on the power dispatch and consumption characteristics, and determine the power distribution trend curve;

[0110] Determine a branch architecture by combining the sub-order twin system with the power distribution trend curve;

[0111] The power distribution records are integrated, and data-driven training is performed on the branch architecture to generate the consumption prediction branch.

[0112] Furthermore, the device is also used to perform the following steps:

[0113] Identify the peak-valley coordination task, perform scale transformation and granularity segmentation on the sub-order twin system, and determine multiple sub-areas;

[0114] Decomposing the peak-valley coordination task and determining multiple greedy goals;

[0115] Mapping the multiple sub-areas and the multiple greedy targets, performing power distribution coordination and determining multiple coordination strategies;

[0116] The multiple coordination strategies are combined to determine the target coordination strategy.

[0117] Furthermore, the device is also used to perform the following steps:

[0118] The first coordination strategy is determined based on the grid connection coordination based on the real-time power generation as the first coordination mode, the energy storage call as the second coordination mode, and the grid dispatch distance as the coordination constraint;

[0119] Determine the second coordination strategy based on demand-side policy drive;

[0120] The first coordination strategy and the second coordination strategy are fitted to determine a first coordination strategy for a first sub-region.

[0121] Furthermore, the device is also used to perform the following steps:

[0122] Execute and monitor the target coordination strategy and determine the strategy response data;

[0123] Traversing the strategy response data, combining the simulated verification data to measure the deviation of the power distribution coordination, and determining the response deviation characteristics, wherein the deviation measurement dimensions include the response vector deviation and the response trend deviation;

[0124] If the response deviation characteristic does not satisfy the coordination degree of freedom, a feedback adjustment instruction is generated.

[0125] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0126] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to multiple elements that can be selected individually or in full. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. The peak-valley coordination method based on twin power distribution is characterized by: The method comprises: The digital distribution twin system of the interactive pre-management area is coded and labeled, the system functional decomposition and label transfer are performed, and the sub-level twin system is determined, where the function is guided by peak-valley coordination; the pre-management area refers to a specific area in the distribution system that is managed and controlled in advance by digital means, and all equipment and resources in the area are included in the digital management system; system functional decomposition refers to the decomposition of complex distribution system functions into several sub-functional modules; label transfer refers to the transfer of already allocated labels from one functional module or node to another; Based on the sub-order twin system, a coordinated decision model is built, which includes parallel demand-side branches and distribution-side branches, and the branches include three fully connected identification and analysis layers, greedy decision layers, and integrated output layers; Combined with the consumption prediction branch, the consumption trend on the demand side is predicted, and the peak-valley coordination task is determined, wherein the task identifier has a peak-valley span with a time phase, and the prediction branch establishes a communication connection with the sub-order twin system, and the time phase refers to the time characteristics of power consumption, that is, the peak and valley of electricity consumption in different time periods; The peak-valley coordination task is transmitted to the coordination decision model, and the demand side and the distribution side are used for bilateral decision fitting to determine the target coordination strategy; Based on the label transfer relationship, the target coordination strategy is simulated and verified based on the digital distribution twin system, and the distribution system is responded to for peak-valley coordination management; The execution system functional decomposition and label transfer includes: Determine the coordination information pool based on the peak-valley coordination needs of the power grid; Based on the digital power distribution twin system, a system dimensionality reduction decomposition is performed based on the coordination information pool to determine a first twin system; Perform label transfer on the first twin system to determine the sub-order twin system.

2. The peak-valley coordination method based on twin power distribution according to claim 1, characterized in that: The coding and labeling comprises: Traversing the digital power distribution twin system to determine the distribution of power distribution nodes; Traversing the distribution of the power distribution nodes, setting a label format based on a union of island systems running on the nodes, wherein the label format has uniformity and node differentiation; Based on the label format, the distribution nodes are coded and labeled.

3. The peak-valley coordination method based on twin power distribution according to claim 1, characterized in that: Construct the consumption prediction branch, including: Read the power distribution records of the preset time zone, conduct periodic mining based on the power dispatch and consumption characteristics, and determine the power distribution trend curve; Determine a branch architecture by combining the sub-order twin system with the power distribution trend curve; The power distribution records are integrated, and data-driven training is performed on the branch architecture to generate the consumption prediction branch.

4. The peak-valley coordination method based on twin power distribution according to claim 1, characterized in that: The target coordination strategy is determined, including: Identify the peak-valley coordination task, perform scale transformation and granularity segmentation on the sub-order twin system, and determine multiple sub-areas; Decomposing the peak-valley coordination task and determining multiple greedy goals; Mapping the multiple sub-areas and the multiple greedy targets, performing power distribution coordination and determining multiple coordination strategies; The multiple coordination strategies are combined to determine the target coordination strategy.

5. The peak-valley coordination method based on twin power distribution according to claim 4, characterized in that: The power distribution coordination determines multiple coordination strategies, including: The first coordination strategy is determined based on the grid connection coordination based on the real-time power generation as the first coordination mode, the energy storage call as the second coordination mode, and the grid dispatch distance as the coordination constraint; Determine the second coordination strategy based on demand-side policy drive; The first coordination strategy and the second coordination strategy are fitted to determine a first coordination strategy for a first sub-region.

6. The peak-valley coordination method based on twin power distribution according to claim 1, characterized in that: The peak-valley coordination management includes: Execute and monitor the target coordination strategy and determine the strategy response data; Traversing the strategy response data, combining the simulated verification data to measure the deviation of the power distribution coordination, and determining the response deviation characteristics, wherein the deviation measurement dimensions include the response vector deviation and the response trend deviation; If the response deviation characteristic does not satisfy the coordination degree of freedom, a feedback adjustment instruction is generated.

7. The peak-valley coordination device based on twin power distribution is characterized by: For executing the method according to any one of claims 1 to 6, the device comprises: Sub-level twin system determination module: used to interact with the digital distribution twin system of the pre-management area and perform coding and labeling, perform system functional decomposition and label transfer, and determine the sub-level twin system, where the function is guided by peak-valley coordination; the pre-management area refers to a specific area in the distribution system that is managed and controlled in advance by digital means, and all equipment and resources in the area are included in the digital management system; system functional decomposition refers to the decomposition of complex distribution system functions into several sub-functional modules; label transfer refers to the transfer of assigned labels from one functional module or node to another; Coordination decision model building module: used to build a coordination decision model based on the sub-order twin system, the coordination decision model includes parallel demand side branches and distribution side branches, and the branches include three layers of fully connected identification and analysis layer, greedy decision layer and integrated output layer; Peak-valley coordination task determination module: used to combine the consumption prediction branch, predict the consumption trend on the demand side, and determine the peak-valley coordination task, wherein the task identifier has a peak-valley span with a time phase, and the prediction branch establishes a communication connection with the sub-order twin system. The time phase refers to the time characteristics of power consumption, that is, the peak and valley of electricity consumption in different time periods; Target coordination strategy determination module: used to transmit the peak-valley coordination task to the coordination decision model, perform bilateral decision fitting on the demand side and the distribution side, and determine the target coordination strategy; Target coordination strategy verification module: used to verify the target coordination strategy based on the label transfer relationship and the digital distribution twin system simulation, and respond to the distribution system to perform peak-valley coordination management.

Images