A power scheduling method and device for a flexible interconnected power distribution network considering the flexibility of energy storage resource aggregation, a computer device and a storage medium

Abstract

The application relates to a flexible interconnected power distribution network power scheduling method and device considering the aggregation flexibility of a similar energy storage resource, computer equipment and a storage medium. The method comprises the following steps: determining affine transformation parameters of a load to be analyzed according to a target polytope model of the load to be analyzed and a preset polytope model; performing fusion processing on the affine transformation parameters to obtain target affine transformation parameters of a flexible interconnected power distribution network; inputting the target affine transformation parameters of the flexible interconnected power distribution network, power distribution network parameters and cost parameters into a trained power scheduling curve prediction model to obtain a power scheduling curve of the flexible interconnected power distribution network; determining a target power scheduling curve of the load to be analyzed according to the affine transformation parameters, the target affine transformation parameters and the power scheduling curve; and performing corresponding power scheduling processing on the load to be analyzed according to the target power scheduling curve. The method can improve the power scheduling accuracy of the similar energy storage load.

Description

Technical Field

[0001] This application relates to the field of power grid technology, and in particular to a power dispatching method, apparatus, computer equipment, computer-readable storage medium, and computer program product for flexible interconnected distribution networks that takes into account the flexibility of energy storage resource aggregation. Background Technology

[0002] With the integration of new energy sources and the diversification of electricity consumption, it is crucial to accurately schedule the power of energy storage-like loads in flexible interconnected distribution networks in order to ensure the operational stability of such networks.

[0003] In traditional technologies, power dispatching for energy storage-like loads in flexible interconnected distribution networks is generally done manually. However, this manual decision-making method is subjective and prone to errors, resulting in low accuracy of power dispatching for energy storage-like loads. Summary of the Invention

[0004] Based on this, it is necessary to provide a power dispatching method, device, computer equipment, computer-readable storage medium, and computer program product for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation and can improve the accuracy of power dispatching for energy storage-like loads.

[0005] Firstly, this application provides a power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage-like resource aggregation, including:

[0006] Construct a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0007] Obtain the preset multicellular model corresponding to the load to be analyzed, and determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0008] The affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network;

[0009] Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network;

[0010] Based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve, the target power scheduling curve of the load to be analyzed is determined.

[0011] Based on the target power scheduling curve, a power scheduling instruction for the load to be analyzed is generated, and the load to be analyzed is subjected to corresponding power scheduling processing according to the power scheduling instruction.

[0012] In one embodiment, determining the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model includes:

[0013] Based on the target multicellular model and the preset multicellular model, the scaling factor and translation vector corresponding to the load to be analyzed are determined;

[0014] The affine transformation parameters of the load to be analyzed are determined based on the scaling factor and the translation vector.

[0015] In one embodiment, before inputting the target affine transformation parameters, the distribution network parameters, and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network, the method further includes:

[0016] Based on the target affine transformation parameters, the upper and lower power limits of the flexible interconnected distribution network are determined.

[0017] Based on the upper power limit and the lower power limit, the power feasible domain information of the flexible interconnected distribution network is determined;

[0018] The step of inputting the target affine transformation parameters, the distribution network parameters, and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network includes:

[0019] The power feasible domain information, the distribution network parameters, and the cost parameters are input into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0020] In one embodiment, determining the target power scheduling curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve includes:

[0021] Based on the target affine transformation parameters, the power scheduling curve is subjected to inverse affine transformation to obtain the preset trajectory information of the preset multicellular model;

[0022] Based on the affine transformation parameters, the preset trajectory information is subjected to affine transformation processing to obtain the target power scheduling curve of the load to be analyzed.

[0023] In one embodiment, constructing the target multicellular model of the load to be analyzed includes:

[0024] Obtain information on the power regulation range, energy storage capacity, and disturbance parameters of the load to be analyzed;

[0025] Based on the power adjustment range information, the energy storage capacity information, and the disturbance parameters, the constraint information corresponding to the load to be analyzed is determined;

[0026] Based on the constraint information, construct the target multicellular model of the load to be analyzed.

[0027] In one embodiment, the fusion processing of the affine transformation parameters to obtain the target affine transformation parameters of the flexible interconnected distribution network includes:

[0028] Extract the first feature vector of the power regulation range information and the second feature vector of the energy storage capacity information;

[0029] The first feature vector and the second feature vector are fused to obtain a fused feature vector.

[0030] The fused feature vector is input into a pre-built importance prediction model to obtain the predicted importance of the load to be analyzed.

[0031] Based on the weights corresponding to the predicted importance, the initial weights of the affine transformation parameters are determined.

[0032] The initial weights are adjusted based on the perturbation parameters to obtain the target weights of the affine transformation parameters;

[0033] According to the target weights, the affine transformation parameters are fused to obtain fused affine transformation parameters, which are used as the target affine transformation parameters of the flexible interconnected distribution network.

[0034] Secondly, this application also provides a power dispatching device for a flexible interconnected distribution network that considers the flexibility of energy storage-like resource aggregation, comprising:

[0035] The model building module is used to construct a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0036] The parameter determination module is used to obtain the preset multicellular model corresponding to the load to be analyzed, and determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0037] The parameter fusion module is used to fuse the affine transformation parameters to obtain the target affine transformation parameters of the flexible interconnected distribution network.

[0038] The curve prediction module is used to obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0039] The curve determination module is used to determine the target power scheduling curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve.

[0040] The power scheduling module is used to generate a power scheduling instruction for the load to be analyzed based on the target power scheduling curve, and to perform corresponding power scheduling processing on the load to be analyzed according to the power scheduling instruction.

[0041] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0042] Construct a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0043] Obtain the preset multicellular model corresponding to the load to be analyzed, and determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0044] The affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network;

[0045] Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network;

[0046] Based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve, the target power scheduling curve of the load to be analyzed is determined.

[0047] Based on the target power scheduling curve, a power scheduling instruction for the load to be analyzed is generated, and the load to be analyzed is subjected to corresponding power scheduling processing according to the power scheduling instruction.

[0048] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0049] Construct a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0050] Obtain the preset multicellular model corresponding to the load to be analyzed, and determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0051] The affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network;

[0052] Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network;

[0053] Based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve, the target power scheduling curve of the load to be analyzed is determined.

[0054] Based on the target power scheduling curve, a power scheduling instruction for the load to be analyzed is generated, and the load to be analyzed is subjected to corresponding power scheduling processing according to the power scheduling instruction.

[0055] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0056] Construct a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0057] Obtain the preset multicellular model corresponding to the load to be analyzed, and determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0058] The affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network;

[0059] Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network;

[0060] Based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve, the target power scheduling curve of the load to be analyzed is determined.

[0061] Based on the target power scheduling curve, a power scheduling instruction for the load to be analyzed is generated, and the load to be analyzed is subjected to corresponding power scheduling processing according to the power scheduling instruction.

[0062] The aforementioned power dispatching method, device, computer equipment, storage medium, and computer program product for flexible interconnected distribution networks, which considers the flexibility of energy storage-like resource aggregation, first constructs a target multi-cell model of the load to be analyzed. The load to be analyzed is used to represent the energy storage-like load in the flexible interconnected distribution network. Then, a preset multi-cell model corresponding to the load to be analyzed is obtained. Based on the target multi-cell model and the preset multi-cell model, the affine transformation parameters of the load to be analyzed are determined, and the affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network. Next, the distribution network parameters and cost parameters of the flexible interconnected distribution network are obtained. The target affine transformation parameters, distribution network parameters, and cost parameters are input into the trained power dispatching curve prediction model to obtain the power dispatching curve of the flexible interconnected distribution network. Then, based on the affine transformation parameters, the target affine transformation parameters, and the power dispatching curve, the target power dispatching curve of the load to be analyzed is determined. Finally, based on the target power dispatching curve, a power dispatching instruction for the load to be analyzed is generated, and the corresponding power dispatching processing is performed on the load to be analyzed according to the power dispatching instruction. In this way, when performing power dispatch on energy storage-like loads in a flexible interconnected distribution network, a target multi-cell model is constructed for the load to be analyzed. The affine transformation parameters are determined by comparing the model with a preset multi-cell model, ensuring that individual characteristics are accurately quantified. By fusing individual parameters, the target affine transformation parameters at the distribution network level are obtained. Combined with distribution network parameters and cost parameters input into the model, a globally optimal power dispatch curve can be generated. By combining the affine transformation parameters, the target affine transformation parameters, and the power dispatch curve, a target power dispatch curve matching the load to be analyzed can be obtained. Based on the target power dispatch curve, the power dispatch command for the load to be analyzed can be determined more accurately, and the power dispatch of the load to be analyzed can be performed more precisely according to the power dispatch command. This is beneficial to improving the power dispatch accuracy of energy storage-like loads. Moreover, the entire process does not require human intervention, avoiding the subjective factors and errors that can easily occur in manual decision-making, which leads to low power dispatch accuracy for energy storage-like loads. This further improves the power dispatch accuracy of energy storage-like loads. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 This is a flowchart illustrating a power dispatching method for a flexible interconnected distribution network that considers the flexibility of energy storage resource aggregation in one embodiment.

[0065] Figure 2 This is a flowchart illustrating a power dispatching method for a flexible interconnected distribution network that considers the flexibility of energy storage resource aggregation in another embodiment.

[0066] Figure 3 This is a schematic diagram illustrating load aggregation combined with flexible interconnection for coordinated optimization scheduling in one embodiment;

[0067] Figure 4 This is a schematic diagram of a flexible interconnected distribution network system in one embodiment;

[0068] Figure 5 This is a diagram illustrating the comparison of 24-hour operating costs for different cases in one embodiment;

[0069] Figure 6 This is a schematic diagram comparing the amount of wind and solar power curtailment in different cases over 24 hours in one embodiment;

[0070] Figure 7 This is a structural block diagram of a flexible interconnected power dispatching device for a distribution network that considers the flexibility of energy storage resource aggregation in one embodiment.

[0071] Figure 8 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0073] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0074] In one exemplary embodiment, such as Figure 1As shown, a flexible interconnected distribution network power dispatching method considering the flexibility of energy storage resource aggregation is provided. This embodiment illustrates the application of this method to a server; it is understood that this method can also be applied to terminals, and can also be applied to systems including terminals and servers, and is implemented through interaction between terminals and servers. The terminals can be, but are not limited to, various personal computers, laptops, smartphones, and tablets; the servers can be implemented using independent servers or server clusters composed of multiple servers. In this embodiment, the method includes the following steps:

[0075] Step S101: Construct a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0076] Among them, the target multicellular model is used to represent a high-dimensional convex geometric model that matches the boundary conditions of the load operation characteristics of the load to be analyzed.

[0077] Among them, flexible interconnected distribution network is used to represent the distribution network form that achieves coordinated control of different feeders, zones or distributed resources (new energy, energy storage, flexible loads) based on power electronics technology (such as smart soft switches).

[0078] Among them, energy storage loads are used to represent flexible loads that can participate in distribution network dispatch through power regulation.

[0079] For example, the server determines the load operation characteristic boundary conditions of the load to be analyzed; then, in response to the model building instruction for the load to be analyzed, the server constructs a target multi-cell model of the load to be analyzed based on the load operation characteristic boundary conditions of the load to be analyzed; the load to be analyzed is used to represent a storage-like load in a flexible interconnected distribution network.

[0080] Step S102: Obtain the preset multicellular model corresponding to the load to be analyzed, and determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0081] The preset multicellular model is used to represent a standardized, structurally symmetric, and scalable standard multicellular model predefined by the aggregator.

[0082] Among them, the affine transformation parameters are used to represent the geometric transformation relationship between the target polytope model and the preset polytope model.

[0083] For example, the server determines the load type of the load to be analyzed, queries the correspondence between the load type and the polymorphic model, obtains the polymorphic model corresponding to the load type of the load to be analyzed, and uses it as the preset polymorphic model corresponding to the load to be analyzed; then, the server determines the parameters of the geometric transformation relationship between the target polymorphic model and the preset polymorphic model based on the target polymorphic model and the preset polymorphic model, and uses these parameters as the affine transformation parameters of the load to be analyzed.

[0084] Step S103: The affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network.

[0085] The target affine transformation parameters refer to the affine transformation parameters after fusion processing.

[0086] For example, the server determines the load type of the load to be analyzed and queries the correspondence between the load type and the weight to obtain the weight corresponding to the load type of the load to be analyzed, which is used as the preset weight corresponding to the affine transformation parameter; then, the server performs fusion processing on the affine transformation parameter according to the preset weight corresponding to the affine transformation parameter to obtain the fused affine transformation parameter, which is used as the target affine transformation parameter of the flexible interconnected distribution network.

[0087] Step S104: Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, distribution network parameters and cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0088] Among them, the distribution network parameters include the topology parameters (such as node parameters), equipment characteristic parameters (such as power supply parameters), and operating constraint parameters (such as frequency constraints) of the flexible interconnected distribution network.

[0089] Among them, cost parameters are used to represent relevant parameters of economic expenditure during the operation of flexible interconnected distribution networks, including equipment maintenance cost parameters and equipment operating cost parameters.

[0090] Among them, the power dispatch curve prediction model refers to a network model that can obtain the power dispatch curve of the flexible interconnected distribution network by using the target affine transformation parameters, distribution network parameters and cost parameters of the flexible interconnected distribution network, such as the long short-term memory network model.

[0091] Among them, the power dispatch curve refers to the trajectory of power change over time in a flexible interconnected distribution network within a specific dispatch period (such as 24 hours).

[0092] For example, the server retrieves the topology parameters, equipment characteristic parameters, and operational constraint parameters of the flexible interconnected distribution network from the database, which are all used as the distribution network parameters of the flexible interconnected distribution network. The server also retrieves the equipment maintenance cost parameters and equipment operating cost parameters of the flexible interconnected distribution network from the database, which are all used as the cost parameters of the flexible interconnected distribution network. Next, the server inputs the target affine transformation parameters, distribution network parameters, and cost parameters into the trained power dispatch curve prediction model to obtain the first power dispatch curve of the flexible interconnected distribution network output by the trained power dispatch curve prediction model. Then, the server inputs the target affine transformation parameters, distribution network parameters, and cost parameters into the historical power dispatch curve prediction model corresponding to the trained power dispatch curve prediction model to obtain the second power dispatch curve of the flexible interconnected distribution network output by the historical power dispatch curve prediction model. Finally, the server performs a fusion process on the first power dispatch curve and the second power dispatch curve according to the first model weight corresponding to the trained power dispatch curve prediction model and the second model weight corresponding to the historical power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0093] Step S105: Determine the target power scheduling curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve.

[0094] The target power scheduling curve refers to the trajectory of the power of the load to be analyzed over time within a specific scheduling period (such as 24 hours).

[0095] For example, the server performs denoising processing on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve to obtain the denoised affine transformation parameters, the denoised target affine transformation parameters, and the denoised power scheduling curve; then, the server determines the target power scheduling curve of the load to be analyzed based on the denoised affine transformation parameters, the denoised target affine transformation parameters, and the denoised power scheduling curve.

[0096] Step S106: Based on the target power scheduling curve, generate the power scheduling instruction for the load to be analyzed, and perform corresponding power scheduling processing on the load to be analyzed according to the power scheduling instruction.

[0097] Among them, the power scheduling command refers to the command to perform power scheduling processing on the load to be analyzed.

[0098] For example, the server inputs the target power scheduling curve into a feature extraction model for feature extraction processing to obtain the feature vector of the target power scheduling curve; then, the server inputs the feature vector of the target power scheduling curve into a trained power scheduling instruction prediction model to obtain the predicted probability of the load to be analyzed under each preset power scheduling instruction; then, the server selects the preset power scheduling instruction with the highest predicted probability from the preset power scheduling instructions as the power scheduling instruction for the load to be analyzed; next, the server performs state monitoring on the load to be analyzed to obtain the state monitoring results; if the state monitoring results indicate that the load to be analyzed has no abnormal state, the server performs the corresponding power scheduling processing on the load to be analyzed according to the power scheduling instruction.

[0099] In the aforementioned power dispatch method for flexible interconnected distribution networks that considers the flexibility of energy storage-like resource aggregation, the following steps are taken: First, a target multi-cell model of the load to be analyzed is constructed. The load to be analyzed represents the energy storage-like load in the flexible interconnected distribution network. Then, a preset multi-cell model corresponding to the load to be analyzed is obtained. Based on the target multi-cell model and the preset multi-cell model, the affine transformation parameters of the load to be analyzed are determined, and the affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network. Next, the distribution network parameters and cost parameters of the flexible interconnected distribution network are obtained. The target affine transformation parameters, distribution network parameters, and cost parameters are input into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network. Then, based on the affine transformation parameters, the target affine transformation parameters, and the power dispatch curve, the target power dispatch curve of the load to be analyzed is determined. Finally, based on the target power dispatch curve, a power dispatch command for the load to be analyzed is generated, and the corresponding power dispatch processing is performed on the load to be analyzed according to the power dispatch command. In this way, when performing power dispatch on energy storage-like loads in a flexible interconnected distribution network, a target multi-cell model is constructed for the load to be analyzed. The affine transformation parameters are determined by comparing the model with a preset multi-cell model, ensuring that individual characteristics are accurately quantified. By fusing individual parameters, the target affine transformation parameters at the distribution network level are obtained. Combined with distribution network parameters and cost parameters input into the model, a globally optimal power dispatch curve can be generated. By combining the affine transformation parameters, the target affine transformation parameters, and the power dispatch curve, a target power dispatch curve matching the load to be analyzed can be obtained. Based on the target power dispatch curve, the power dispatch command for the load to be analyzed can be determined more accurately, and the power dispatch of the load to be analyzed can be performed more precisely according to the power dispatch command. This is beneficial to improving the power dispatch accuracy of energy storage-like loads. Moreover, the entire process does not require human intervention, avoiding the subjective factors and errors that can easily occur in manual decision-making, which leads to low power dispatch accuracy for energy storage-like loads. This further improves the power dispatch accuracy of energy storage-like loads.

[0100] In an exemplary embodiment, step S102 above, which determines the affine transformation parameters of the load to be analyzed based on the target polymorphic model and the preset polymorphic model, specifically includes the following: determining the scaling factor and translation vector corresponding to the load to be analyzed based on the target polymorphic model and the preset polymorphic model; and determining the affine transformation parameters of the load to be analyzed based on the scaling factor and translation vector.

[0101] The scaling factor is a parameter used to represent the difference in power numerical scale between the target multicell model and the preset multicell model. For example, if the power range of the target multicell model is 200-1000kW and the power range of the preset multicell model is 100-500kW, the corresponding scaling factor is 2.

[0102] The translation vector represents the overall positional offset between the target multicell model and the preset multicell model in the power dimension. For example, if the power range of the target multicell model is 300-700kW and the power range of the preset multicell model is 100-500kW, the corresponding translation vector is 200kW.

[0103] For example, the server extracts the first upper limit and the first lower limit of power interaction corresponding to the target multicell model, as well as the second upper limit and the second lower limit of power interaction corresponding to the preset multicell model. Then, the server subtracts the first upper limit and the first lower limit of power interaction to obtain the first power interaction range value corresponding to the target multicell model, and subtracts the second upper limit and the second lower limit of power interaction to obtain the second power interaction range value corresponding to the preset multicell model. Finally, the server divides the first power interaction range value and the second power interaction range value, and uses the quotient as the value corresponding to the load to be analyzed. The scaling factor is then used. Next, the server multiplies the second power interaction upper limit value and the scaling factor to obtain a processed second power interaction upper limit value. The difference between the first power interaction upper limit value and the processed second power interaction upper limit value is then used as the translation vector corresponding to the load to be analyzed. Alternatively, the server multiplies the second power interaction lower limit value and the scaling factor to obtain a processed second power interaction lower limit value. The difference between the first power interaction lower limit value and the processed second power interaction lower limit value is then used as the translation vector corresponding to the load to be analyzed. Finally, the server uses both the scaling factor and the translation vector as affine transformation parameters for the load to be analyzed.

[0104] In this embodiment, the scaling factor and translation vector are determined by the target multicell model and the preset multicell model, and then the affine transformation parameters are further formed. This allows for the accurate quantification of the scale differences and positional shifts in the power dimension between the individual characteristics of the load to be analyzed and the common benchmark of similar loads, providing unified and reliable parameter support for subsequent data analysis.

[0105] In an exemplary embodiment, step S104, before inputting the target affine transformation parameters, distribution network parameters, and cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network, specifically includes the following: determining the upper and lower power limits of the flexible interconnected distribution network based on the target affine transformation parameters; and determining the power feasible region information of the flexible interconnected distribution network based on the upper and lower power limits.

[0106] Therefore, step S104 above, which inputs the target affine transformation parameters, distribution network parameters, and cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network, specifically includes the following: inputting the power feasible domain information, distribution network parameters, and cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0107] The upper limit of power refers to the maximum power interaction value of the flexible interconnected distribution network.

[0108] The power interaction values ​​include power input values ​​and power output values.

[0109] The lower power limit refers to the minimum power interaction value of a flexible interconnected distribution network.

[0110] Among them, the power interaction range of the power feasible domain information flexible interconnected distribution network.

[0111] For example, the server extracts the target scaling factor and target translation vector of the flexible interconnected distribution network from the target affine transformation parameters. Next, the server obtains the upper limit of power interaction corresponding to a preset polytope model, multiplies the upper limit of power interaction with the target scaling factor to obtain the processed upper limit of power interaction, and adds the processed upper limit of power interaction with the target translation vector to obtain the upper limit of power for the flexible interconnected distribution network. Then, the server obtains the lower limit of power interaction corresponding to the preset polytope model, multiplies the lower limit of power interaction with the target scaling factor to obtain the processed lower limit of power interaction, and adds the processed lower limit of power interaction with the target translation vector to obtain the lower limit of power for the flexible interconnected distribution network. Next, the server uses the upper limit of power as the maximum value and the lower limit of power as the minimum value to construct a power range corresponding to the upper and lower limits of power, as the power feasible region information of the flexible interconnected distribution network. Finally, the server extracts the power feasible region information. The server first obtains the feature vectors of power feasible domain information, distribution network parameters, and cost parameters, and then concatenates these feature vectors to obtain a concatenated feature vector. Next, the server inputs this concatenated feature vector into a trained power dispatch curve prediction model to obtain the first power dispatch curve of the flexible interconnected distribution network output by the trained power dispatch curve prediction model. Then, the server inputs this concatenated feature vector into the historical power dispatch curve prediction model corresponding to the trained power dispatch curve prediction model to obtain the second power dispatch curve of the flexible interconnected distribution network output by the historical power dispatch curve prediction model. Finally, the server fuses the first and second power dispatch curves according to the first model weights corresponding to the trained power dispatch curve prediction model and the second model weights corresponding to the historical power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0112] In this embodiment, the aggregated load regulation characteristics of the flexible interconnected distribution network are accurately quantified by using target affine transformation parameters. This allows for the determination of upper and lower power limits and power feasible domain information that conform to actual operational constraints, ensuring the scientific nature of the scheduling boundary. This achieves a balance between the accuracy, safety, and economy of power scheduling in the flexible interconnected distribution network, thereby adapting to the dynamic scheduling needs of the distribution network in multiple scenarios.

[0113] In an exemplary embodiment, step S105 above, which determines the target power scheduling curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve, specifically includes the following: performing an inverse affine transformation on the power scheduling curve based on the target affine transformation parameters to obtain the preset trajectory information of the preset multicell model; and performing an affine transformation on the preset trajectory information based on the affine transformation parameters to obtain the target power scheduling curve of the load to be analyzed.

[0114] Among them, the preset trajectory information is used to represent the power change trajectory over time of the standard operating law corresponding to the preset multicellular model.

[0115] For example, the server performs denoising processing on the affine transformation parameters, target affine transformation parameters, and power scheduling curve to obtain denoised affine transformation parameters, denoised target affine transformation parameters, and denoised power scheduling curve. Next, the server performs time step alignment processing on the denoised affine transformation parameters, denoised target affine transformation parameters, and denoised power scheduling curve to obtain processed affine transformation parameters, processed target affine transformation parameters, and processed power scheduling curve. Then, the server uses a preset inverse affine transformation processing model and, based on the processed target affine transformation parameters, performs inverse affine transformation processing on the processed power scheduling curve to obtain preset trajectory information of a preset multicell model. Finally, the server uses the preset affine transformation processing model and, based on the processed affine transformation parameters, performs affine transformation processing on the preset trajectory information to obtain the target power scheduling curve of the load to be analyzed.

[0116] In this embodiment, the system-level power scheduling curve is first decoupled into the preset trajectory information of the preset multicellular model by the target affine transformation parameters, ensuring that the preset trajectory information can accurately reflect the standard operation template of the load. Then, based on the individual affine transformation parameters of the load to be analyzed, the preset trajectory information is adapted into the individual target power scheduling curve, which helps to improve the accuracy of determining the target power scheduling curve of the load to be analyzed.

[0117] In an exemplary embodiment, step S101 above, constructing a target multi-cell model of the load to be analyzed, specifically includes the following: obtaining the power regulation range information, energy storage capacity information, and disturbance parameters of the load to be analyzed; determining the constraint information corresponding to the load to be analyzed based on the power regulation range information, energy storage capacity information, and disturbance parameters; and constructing a target multi-cell model of the load to be analyzed based on the constraint information.

[0118] Among them, the power adjustment range information refers to the upper and lower limits of the power interaction of the load to be analyzed.

[0119] Among them, energy storage capacity information refers to the key parameters of the load to be analyzed in terms of energy storage, such as rated capacity and rated charging and discharging power.

[0120] Among them, disturbance parameters are used to represent the uncertainty parameters of the load to be analyzed, including internal disturbance parameters, such as the load's own operating status fluctuation parameters (such as the power fluctuation of air conditioners caused by changes in indoor and outdoor temperature differences, and the instantaneous power impact caused by the start-up and shutdown of production equipment); and also external disturbances, such as small fluctuation parameters of grid-side voltage or frequency.

[0121] Among them, constraint information is used to represent the set of mathematical boundary conditions that restrict the power scheduling of the load to be analyzed.

[0122] For example, the server retrieves the maximum and minimum power interaction values ​​of the load to be analyzed from the database as the power regulation range information of the load to be analyzed, the rated capacity and rated charge / discharge power of the load to be analyzed from the database as the energy storage capacity information of the load to be analyzed, and the internal and external disturbance parameters of the load to be analyzed from the database as the disturbance parameters of the load to be analyzed. Then, based on the power regulation range information, energy storage capacity information, and disturbance parameters, the server determines the static power constraint information and dynamic power constraint information corresponding to the load to be analyzed, both of which are used as the constraint information corresponding to the load to be analyzed. Then, based on the constraint information, the server constructs an initial multi-cell model of the load to be analyzed. Next, the server performs a feasible solution verification process on the initial multi-cell model (i.e., for a given candidate point, checks whether it satisfies all the constraints of the problem) to obtain the verification result of the initial multi-cell model. If the verification result indicates that the initial multi-cell model has passed the verification, the initial multi-cell model is used as the target multi-cell model of the load to be analyzed.

[0123] In this embodiment, the physical characteristic parameters of the load to be analyzed are used as the core input and transformed into mathematical constraint information to ensure that the constraints can accurately cover the static power boundary, dynamic adjustment capability, energy storage collaborative safety and anti-disturbance requirements of the load. Then, a target multi-cell model is constructed based on the constraints, so that the model can completely map the operable state space of the load, making it easy to intuitively quantify the power adjustment potential and operating limitations of the load.

[0124] In an exemplary embodiment, step S103 above, which involves fusing the affine transformation parameters to obtain the target affine transformation parameters for the flexible interconnected distribution network, specifically includes the following: extracting a first feature vector of power regulation range information and a second feature vector of energy storage capacity information; fusing the first and second feature vectors to obtain a fused feature vector; inputting the fused feature vector into a pre-built importance prediction model to obtain the predicted importance of the load to be analyzed; determining the initial weights of the affine transformation parameters based on the weights corresponding to the predicted importance; adjusting the initial weights according to the disturbance parameters to obtain the target weights of the affine transformation parameters; and fusing the affine transformation parameters according to the target weights to obtain the fused affine transformation parameters, which serve as the target affine transformation parameters for the flexible interconnected distribution network.

[0125] The first feature vector is used to represent the representation vector of power adjustment range information.

[0126] The second feature vector is used to represent the representation vector of energy storage capacity information.

[0127] Among them, the fused feature vector refers to the feature vector obtained by fusing the first feature vector and the second feature vector.

[0128] Among them, the importance prediction model refers to the model that can predict the importance of the load to be analyzed, such as the attention model.

[0129] Among them, the prediction importance refers to the predicted value corresponding to the importance of the load to be analyzed.

[0130] The initial weight refers to the weight corresponding to the predicted importance.

[0131] The target weight refers to the initial weight after adjustment.

[0132] Among them, the fused affine transformation parameters refer to the affine transformation parameters after fusion processing.

[0133] For example, the server uses power regulation range information as primary data and energy storage capacity information as secondary data, inputting them into a feature extraction model for feature extraction processing to obtain a first feature vector of power regulation range information and a second feature vector of energy storage capacity information. Next, the server determines a first weight corresponding to the power regulation range information and a second weight corresponding to the energy storage capacity information, and sums the first and second feature vectors according to the first and second weights to obtain a fused feature vector. Then, the server inputs the fused feature vector into a pre-built importance prediction model. The server first obtains the predicted importance of the load to be analyzed. Then, it uses the weight corresponding to the predicted importance as the initial weight of the affine transformation parameters. Next, based on the preset parameter range to which the disturbance parameters belong (this range needs to be pre-defined considering load operating characteristics, historical disturbance data, and engineering requirements), the server queries the pre-defined correspondence between the parameter range and the adjustment coefficient to obtain the adjustment coefficient corresponding to the current parameter range, which serves as the target adjustment coefficient corresponding to the initial weight. Then, the server adjusts the initial weight according to the target adjustment coefficient to obtain the adjusted weight of the affine transformation parameters, which serves as the target weight of the affine transformation parameters. Finally, the server performs a fusion process on the affine transformation parameters according to the target weight to obtain the fused affine transformation parameters, which serve as the target affine transformation parameters for the flexible interconnected distribution network.

[0134] In this embodiment, through the end-to-end design of feature extraction, fusion, importance prediction, weight adjustment, and parameter fusion, the output target affine transformation parameters can not only accurately match the characteristics of the load to be analyzed, but also adapt to the operational requirements under disturbance scenarios. This provides highly reliable parameter support for subsequent determination of the power feasible region and generation of power scheduling curves, and realizes the accurate and dynamic optimization of the target affine transformation parameters of the flexible interconnected distribution network.

[0135] In one exemplary embodiment, such as Figure 2 As shown, another power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation is provided. Taking the application of this method to a server as an example, the specific steps include:

[0136] Step S201: Obtain the power regulation range information, energy storage capacity information and disturbance parameters of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in flexible interconnected distribution networks.

[0137] Step S202: Based on the power regulation range information, energy storage capacity information, and disturbance parameters, determine the constraint information corresponding to the load to be analyzed; based on the constraint information, construct the target multicellular model of the load to be analyzed.

[0138] Step S203: Obtain the preset multicellular model corresponding to the load to be analyzed.

[0139] Step S204: Based on the target multicellular model and the preset multicellular model, determine the scaling factor and translation vector corresponding to the load to be analyzed; based on the scaling factor and translation vector, determine the affine transformation parameters of the load to be analyzed.

[0140] Step S205: Extract the first feature vector of power regulation range information and the second feature vector of energy storage capacity information; perform fusion processing on the first feature vector and the second feature vector to obtain the fused feature vector; input the fused feature vector into the pre-constructed importance prediction model to obtain the predicted importance of the load to be analyzed.

[0141] Step S206: Determine the initial weights of the affine transformation parameters based on the weights corresponding to the predicted importance; adjust the initial weights according to the perturbation parameters to obtain the target weights of the affine transformation parameters.

[0142] Step S207: According to the target weight, the affine transformation parameters are fused to obtain the fused affine transformation parameters, which are used as the target affine transformation parameters of the flexible interconnected distribution network.

[0143] Step S208: Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network.

[0144] Step S209: Determine the upper and lower power limits of the flexible interconnected distribution network based on the target affine transformation parameters; determine the power feasible region information of the flexible interconnected distribution network based on the upper and lower power limits.

[0145] Step S210: Input the power feasible domain information, distribution network parameters and cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0146] Step S211: Based on the target affine transformation parameters, perform inverse affine transformation on the power scheduling curve to obtain the preset trajectory information of the preset multicellular model.

[0147] Step S212: Perform affine transformation processing on the preset trajectory information according to the affine transformation parameters to obtain the target power scheduling curve of the load to be analyzed.

[0148] Step S213: Based on the target power scheduling curve, generate the power scheduling instruction for the load to be analyzed, and perform corresponding power scheduling processing on the load to be analyzed according to the power scheduling instruction.

[0149] In the aforementioned power dispatch method for flexible interconnected distribution networks that considers the flexibility of energy storage-like resource aggregation, when dispatching power to energy storage-like loads in the flexible interconnected distribution network, a target multi-cell model is constructed for the load to be analyzed. The affine transformation parameters are determined by comparing the model with a preset multi-cell model, ensuring that individual characteristics are accurately quantified. By fusing individual parameters, the target affine transformation parameters at the distribution network level are obtained. Combined with distribution network parameters and cost parameters input into the model, a globally optimal power dispatch curve can be generated. By combining the affine transformation parameters, the target affine transformation parameters, and the power dispatch curve, a target power dispatch curve matching the load to be analyzed can be obtained. Based on the target power dispatch curve, the power dispatch command for the load to be analyzed can be determined more accurately, and the power dispatch of the load to be analyzed can be performed more precisely according to the power dispatch command, which is beneficial to improving the power dispatch accuracy of energy storage-like loads. Moreover, the entire process does not require human intervention, avoiding the subjective factors and errors that are prone to occur in manual decision-making, which leads to low power dispatch accuracy of energy storage-like loads, thus further improving the power dispatch accuracy of energy storage-like loads.

[0150] In an exemplary embodiment, to more clearly illustrate the power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation provided in this application, the following specific embodiment will be used to describe the power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation. In one embodiment, as Figure 3As shown, this application also provides a flexible interconnected distribution network scheduling optimization method that considers the flexibility of energy storage-like resource aggregation, specifically including the following:

[0151] Step S1: For each load resource in multiple distributed flexible loads, based on its local physical operational constraints, construct its individual flexibility multi-cell model within a predetermined control time range to obtain the boundary conditions of load operation characteristics. The individual multi-cell model can be uniformly represented as a feasible region composed of the first set of linear inequalities, possessing a high-dimensional convex polyhedral structure. Neither the required data nor the model needs to be uploaded, ensuring user privacy.

[0152] Step S2: A standardized, structurally symmetric, and scalable standard multicellular structure is predefined by the aggregator and used as the baseline model for subsequent geometric approximation of all load individuals. This prototype is defined by a unified set of linear inequalities and does not depend on any individual information.

[0153] Step S3: Each flexible load individual, within its local environment, calculates a locally unique scaling factor and translation vector by solving a maximum inner approximation transformation problem based on its own flexible multicell and the baseline model issued by the system. This ensures that the transformed prototype approximates its actual feasible domain boundary as closely as possible. This optimization process is entirely completed independently by the individual and does not depend on other user information.

[0154] Step S4: All flexible load individuals only need to upload two scalar parameters (scaling factor and translation vector) of their local calculation results to the aggregator. The aggregator then applies these parameters to the prototype baseline multicell to quickly construct a global aggregated flexibility model.

[0155] Step S5: The aggregated set of feasible flexible resources is used as the dispatchable resources of the flexible interconnected distribution network, and a collaborative optimization scheduling model containing flexible interconnected devices is established. The model simultaneously considers power flow and power balance constraints, node voltage and line capacity constraints, active and reactive power conservation and capacity constraints at SOP (Soft Open Point) ports, inverter operation constraints and power factor limitations, etc., with the goal of minimizing the total operating cost. Cost items include the electricity purchase / sale fees of the upstream grid, the cost of energy provided by the aggregator, and the secondary costs of local distributed power sources (such as gas turbines).

[0156] Step S6: Based on an aggregated power scheduling curve that conforms to the aggregated flexibility model, through inverse and forward affine transformations, combined with the total scaling factor, the total translation vector, and the individual scaling factor and individual translation vector of each flexible load, the individual power curve that each flexible load should execute precisely is derived by reverse analysis.

[0157] Furthermore, in step S1, for each flexible load (such as an electric vehicle charger, air conditioning load, industrial flexible load, etc.) among multiple distributed flexible loads, an individual flexible resource multicellular model is constructed within a predetermined control time range. :

[0158] Equation (1)

[0159] in, The power input of the flexible load i, Let i be the power input of the flexible load at time k. A positive value indicates charging. A negative value indicates discharge. The internal energy state of the flexible load i at time k (such as battery state of charge, room temperature, etc.). Where k is the time step size, and k is the discrete time index, also known as the time step number. This refers to the lower limit of power input. This refers to the upper limit of power input. This refers to the lower limit of the internal energy state. This refers to the upper limit of the internal energy state. Formula (1) includes the state transition equation of the flexible load on a discrete-time scale, the upper and lower limits of power input, and the upper and lower limits of the internal energy state. Let be the state transition matrix.

[0160] Furthermore, the above constraints can be uniformly expressed in the form of a standard set of linear inequalities, namely:

[0161] Equation (2)

[0162] in, This represents the linear constraint matrix composed of various constraints in formula (1), where n is the number of variables and m is the matrix. The number of rows, or the number of "faces" of a polyhedron. These are the boundary vectors for each constraint.

[0163] Considering the initial stored energy state of the load and the randomness of energy demand, Using the opportunity constraint technique as the uncertainty parameter, a flexible resource multicellular model considering uncertainty is constructed as follows:

[0164] Equation (3)

[0165] in, Used to represent probability It refers to the minimum probability that the constraint must be satisfied.

[0166] The above opportunity constraint model can be further transformed into:

[0167] Equation (4)

[0168] in, for At a confidence level of Quantiles at that point.

[0169] To describe the overall flexibility of the load population, the aggregate flexibility set is constructed by summing the elements of all individual polycells, and is defined as the Minkowski Sum of the polycells, i.e.:

[0170] Equation (5)

[0171] Among them, set It includes the feasible region of the overall power curve after aggregating all flexible load groups. Let Minkowski sum be the sum of the sums ...

[0172] Furthermore, in step S2, a high-dimensional convex polyhedron defined by the second set of linear inequality constraints is constructed, called the standard polytope. The formal structure may include:

[0173] Equation (6)

[0174] Where v represents the normalized power trajectory, , These are predefined constant matrices and vectors used to define the structure of the standard polytope. They are typically symmetric, zero-mean, or have unit constraint intervals, making them easy to adapt to individual constraints through affine transformations.

[0175] Furthermore, in step S3, for each flexible load, its individual flexible multicell has already been established in step S1. To embed a multicell model that approximates the prototype as closely as possible into this individual multicell, the following affine transformation is constructed:

[0176] Equation (7)

[0177] in, The scaling factor for the flexible load i controls the size of the multicellular model. The translation vector of the flexible load i controls the position of the multicellular model in the power trajectory space.

[0178] To ensure that the prototype model after the affine transformation is completely contained within the individual multicellular body, the following constraints must be satisfied:

[0179] Equation (8)

[0180] Constructing standard multicellular structures An affine transformation model is used as the baseline, and its application in flexible load individual multicellular structures is solved. The largest inner approximation in.

[0181] By solving the above optimization problem, a unique set of affine transformation parameters can be calculated for each flexible load individual, which can be used to embed the unified prototype model into its flexibility adjustment space.

[0182] Furthermore, in step S4, for a group containing N flexible load resources, their affine transformation parameters are respectively... Algebraic summation is performed to construct the total transformation parameters of the aggregation hierarchy, and the unified standard multicellular structure is then applied. Applying the above aggregation parameters, construct the flexible feasible domain after aggregation:

[0183] Equation (9)

[0184] This represents a model of the aggregation flexibility of a flexible load population during the regulation period; the model is still a high-dimensional convex multicellular structure.

[0185] Furthermore, in step S5, considering the collaborative optimization scheduling of the flexible interconnected distribution network and electric vehicle aggregators over the 24 day-ahead periods, the objective of minimizing system operating costs is:

[0186] Equation (10)

[0187] Where T refers to the total number of time steps, and the first term in the objective function represents the power generation cost of distributed generators (such as gas turbines) at each node. Let the generator at node i produce power at time t; and The first term represents the corresponding cost coefficient (commonly a quadratic term structure), characterizing fuel and operating costs. The second term represents the cost of purchasing electricity from the upper-level grid at each time period, determined by the time-of-period electricity price. With power purchase The third item is the energy cost incurred by the aggregator when supplying electrical energy to the system (i.e., discharging it), based on the price quoted to the market. This refers to the unit profit coefficient. This represents the aggregate quotient of the charging and discharging power at node i. A value greater than 0 indicates charging, and a value less than 0 indicates discharging.

[0188] To enhance power flow reconfiguration and voltage regulation capabilities between feeders, a new type of power electronic device, namely intelligent soft switch (SOP), is introduced into the distribution network. An SOP typically consists of a back-to-back voltage source converter and a DC bus, possessing bidirectional active / reactive power regulation capabilities. It can rapidly and continuously transfer active power between its two connected ports and independently provide reactive power support to each connected node.

[0189] Furthermore, in step S6, the central system will derive a power scheduling instruction that conforms to the aggregation flexibility model based on the optimized scheduling results in step S5. ,satisfy To achieve lossless reverse decomposition of scheduling instructions from the aggregation layer to the individual layer, the given aggregation control trajectory is reasonably decomposed into each flexible load individual to ensure that it independently executes the control task without violating the individual operation constraints.

[0190] because It is by using the benchmark multicellular model The trajectory obtained by performing a global affine transformation can be inversely transformed and proportionally distributed to obtain a standardized prototype trajectory:

[0191] Equation (11)

[0192] Then, using the affine transformation parameters obtained in step S3, the trajectory is decomposed into the local transformation space corresponding to each individual load, and the power trajectory to be executed by the individual load is constructed:

[0193] Equation (12)

[0194] in That is, the power regulation trajectory obtained from the power regulation trajectory that flexible load i should perform throughout the entire regulation cycle all satisfy its individual flexibility multi-cell constraint, i.e. This inverse process does not require resolving the optimization problem; it can quickly, accurately, and losslessly map aggregated hierarchical instructions to various flexible workload resources.

[0195] The power distribution system on which this embodiment is based has the following structure: Figure 4As shown. This includes two smart soft-switching SOPs, SOP 1 and SOP 2, WT (Wind Turbine), PV (Photovoltaic), G (Generator), and agg (Aggregator). The rated voltage level is 11kV. The maximum capacity of each branch is 5MVA. Two 1MVA (Mega Volt - Ampere) smart soft-switching SOPs are installed between nodes 12 and 22, and between buses 25 and 29, respectively. Three 1MVA photovoltaic generators (PVs) and two 1MVA wind turbines (WTs) are installed at nodes 10, 13, 16 and 17, 30, respectively. Three energy storage aggregators are connected to nodes 5, 19, and 32. The parameter data for each type of energy storage load is generated as follows: Capacity C ~ (40, 60) kWh, maximum power ~ (6, 9) kW, minimum charging power ~ (-9, -6) kW (negative discharge), expected value of initial energy state ~ (0, 0.4C) kWh, considering uncertainties, the initial energy state follows a normal distribution with a variance of 0.1 times the expected value, and within the capacity range to ensure physical feasibility, 500 flexible loads are aggregated in the experimental scenario.

[0196] To verify the effectiveness of this embodiment, four sets of comparison calculations are set as follows:

[0197] Case 1 (No Aggregation, No Flexible Interconnection): Without considering the aggregation and adjustment capabilities of flexible resources, the distribution network is dispatched according to the AC power flow model, the aggregation power is fixed at 0, and the distribution network is a pure AC network without flexible interconnection equipment.

[0198] Case 2 (Aggregation Only): Based on Case 1, the aggregation feasible domain obtained in step S4 is incorporated as an operational constraint to optimize and determine the charging / discharging power of each node aggregator. The power distribution network is still a pure AC network without flexible interconnection equipment.

[0199] Case 3 (Flexible Interconnection Only): Without considering flexible resource aggregation, the aggregator power is fixed at 0. Two 1000kVA power electronic smart soft switches (SOPs) are installed in the distribution network to participate in dispatch.

[0200] Case 4 (Proposed Energy Storage-like Load Aggregation and Flexible Interconnection Coordinated Optimization): Simultaneously introducing the aggregation feasible domain and two 1000kV smart soft switches, the active and reactive power of distributed power sources and aggregators in the flexible interconnected distribution network are coordinated and optimized under a unified scheduling model.

[0201] in, Figure 5 This is a diagram comparing the 24-hour operating costs of different cases (negative values ​​represent profits). Figure 6 This is a diagram showing the comparison of wind and solar power curtailment in different cases over 24 hours.

[0202] In the above embodiments, when performing power dispatch on energy storage-like loads in a flexible interconnected distribution network, a target multi-cell model is constructed for the load to be analyzed. Affine transformation parameters are determined by comparing the model with a preset multi-cell model, ensuring that individual characteristics are accurately quantified. The target affine transformation parameters at the distribution network level are obtained by fusing individual parameters. Combined with distribution network parameters and cost parameters input into the model, a globally optimal power dispatch curve can be generated. By combining the affine transformation parameters, the target affine transformation parameters, and the power dispatch curve, a target power dispatch curve matching the load to be analyzed can be obtained. Based on the target power dispatch curve, the power dispatch command for the load to be analyzed can be determined more accurately, and power dispatch can be performed more precisely according to the power dispatch command. This improves the accuracy of power dispatch for energy storage-like loads. Furthermore, the entire process requires no manual intervention, avoiding the subjective factors and errors inherent in manual decision-making, which can lead to lower accuracy in power dispatch for energy storage-like loads. This further improves the accuracy of power dispatch for energy storage-like loads. Simultaneously, an approximate aggregation method based on affine transformation is proposed. Through a decentralized architecture, it significantly reduces the computational complexity of massive heterogeneous resource aggregation modeling while effectively protecting user privacy, avoiding the NP-hard problem of Minkowski sum operations, and is suitable for large-scale, real-time demand response scenarios. This efficient aggregation flexible resource model is unified and collaboratively optimized with distribution network scheduling containing flexible interconnected devices (SOPs), deeply integrating decentralized resource regulation capabilities with proactive network power flow control capabilities. This integrated "source-load-grid" framework effectively solves the problems of resource capacity being disconnected from network capacity and scheduling results being difficult to implement in the physical network in existing technologies. It can reduce operating costs, alleviate line congestion, and improve renewable energy consumption at the system-wide level. By utilizing the inverse process of affine transformation, the system-level scheduling instructions are rapidly and losslessly decomposed into individual execution instructions, ensuring the reliability and implementation of regulation. This provides a general technical solution with engineering application prospects for various flexible load resources with convex constraints to participate in grid optimization.

[0203] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0204] Based on the same inventive concept, this application also provides a power dispatching device for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation, for implementing the power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation described above. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the power dispatching device for flexible interconnected distribution networks that consider the flexibility of energy storage resource aggregation provided below can be found in the limitations of the power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage resource aggregation above, and will not be repeated here.

[0205] In one exemplary embodiment, such as Figure 7 As shown, a flexible interconnected distribution network power dispatching device considering the flexibility of energy storage-like resource aggregation is provided, including: a model building module 701, a parameter determination module 702, a parameter fusion module 703, a curve prediction module 704, a curve determination module 705, and a power dispatching module 706, wherein:

[0206] Model building module 701 is used to build a target multicellular model of the load to be analyzed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network.

[0207] The parameter determination module 702 is used to obtain the preset multicellular model corresponding to the load to be analyzed, and to determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model.

[0208] The parameter fusion module 703 is used to fuse the affine transformation parameters to obtain the target affine transformation parameters of the flexible interconnected distribution network.

[0209] The curve prediction module 704 is used to obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network. The target affine transformation parameters, distribution network parameters and cost parameters are input into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

[0210] The curve determination module 705 is used to determine the target power dispatch curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power dispatch curve.

[0211] The power scheduling module 706 is used to generate power scheduling instructions for the load to be analyzed based on the target power scheduling curve, and to perform corresponding power scheduling processing on the load to be analyzed according to the power scheduling instructions.

[0212] In an exemplary embodiment, the parameter determination module 702 is further configured to determine the scaling factor and translation vector corresponding to the load to be analyzed based on the target multicell model and the preset multicell model; and to determine the affine transformation parameters of the load to be analyzed based on the scaling factor and translation vector.

[0213] In an exemplary embodiment, the flexible interconnected distribution network power dispatching device that considers the flexibility of energy storage resource aggregation further includes an information determination module, used to determine the upper and lower power limits of the flexible interconnected distribution network based on the target affine transformation parameters; and to determine the power feasible region information of the flexible interconnected distribution network based on the upper and lower power limits; the curve prediction module 704 is further used to input the power feasible region information, distribution network parameters and cost parameters into the trained power dispatching curve prediction model to obtain the power dispatching curve of the flexible interconnected distribution network.

[0214] In an exemplary embodiment, the curve determination module 705 is further configured to perform inverse affine transformation processing on the power scheduling curve according to the target affine transformation parameters to obtain preset trajectory information of the preset multicell model; and perform affine transformation processing on the preset trajectory information according to the affine transformation parameters to obtain the target power scheduling curve of the load to be analyzed.

[0215] In an exemplary embodiment, the model building module 701 is further configured to acquire power regulation range information, energy storage capacity information, and disturbance parameters of the load to be analyzed; determine the constraint information corresponding to the load to be analyzed based on the power regulation range information, energy storage capacity information, and disturbance parameters; and construct a target multicellular model of the load to be analyzed based on the constraint information.

[0216] In an exemplary embodiment, the parameter fusion module 703 is further configured to extract a first feature vector of power regulation range information and a second feature vector of energy storage capacity information; perform fusion processing on the first feature vector and the second feature vector to obtain a fused feature vector; input the fused feature vector into a pre-constructed importance prediction model to obtain the predicted importance of the load to be analyzed; determine the initial weights of the affine transformation parameters based on the weights corresponding to the predicted importance; adjust the initial weights according to the disturbance parameters to obtain the target weights of the affine transformation parameters; and perform fusion processing on the affine transformation parameters according to the target weights to obtain fused affine transformation parameters, which serve as the target affine transformation parameters for the flexible interconnected distribution network.

[0217] The modules in the aforementioned flexible interconnected distribution network power dispatching device, which considers the flexibility of energy storage resource aggregation, can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0218] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores data such as distribution network parameters and cost parameters. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a flexible interconnected distribution network power dispatching method that considers the flexibility of energy storage resource aggregation.

[0219] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0220] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0221] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above-described method embodiments.

[0222] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0223] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0224] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0225] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A power dispatching method for flexible interconnected distribution networks that considers the flexibility of energy storage-like resource aggregation, characterized in that, The method includes: A target multicellular model of the load to be analyzed is constructed; the load to be analyzed is used to represent energy storage-like loads in a flexible interconnected distribution network; the target multicellular model is used to represent a geometric model that matches the boundary conditions of the load operation characteristics of the load to be analyzed. A preset multicell model corresponding to the load to be analyzed is obtained. Based on the target multicell model and the preset multicell model, the affine transformation parameters of the load to be analyzed are determined. The affine transformation parameters refer to the scaling factor and translation vector corresponding to the load to be analyzed. The scaling factor is used to represent the difference between the target multicell model and the preset multicell model in the power numerical scale. The translation vector is used to represent the overall positional offset between the target multicell model and the preset multicell model in the power dimension. The affine transformation parameters are fused to obtain the target affine transformation parameters of the flexible interconnected distribution network; Obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network; Based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve, the target power scheduling curve of the load to be analyzed is determined. Based on the target power scheduling curve, a power scheduling instruction for the load to be analyzed is generated, and the load to be analyzed is subjected to corresponding power scheduling processing according to the power scheduling instruction.

2. The method according to claim 1, characterized in that, The step of determining the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model includes: Based on the target multicellular model and the preset multicellular model, the scaling factor and the translation vector corresponding to the load to be analyzed are determined; The affine transformation parameters of the load to be analyzed are determined based on the scaling factor and the translation vector.

3. The method according to claim 1, characterized in that, Before inputting the target affine transformation parameters, the distribution network parameters, and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network, the process further includes: Based on the target affine transformation parameters, the upper and lower power limits of the flexible interconnected distribution network are determined. Based on the upper power limit and the lower power limit, the power feasible domain information of the flexible interconnected distribution network is determined; The step of inputting the target affine transformation parameters, the distribution network parameters, and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network includes: The power feasible domain information, the distribution network parameters, and the cost parameters are input into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network.

4. The method according to claim 1, characterized in that, The step of determining the target power scheduling curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve includes: Based on the target affine transformation parameters, the power scheduling curve is subjected to inverse affine transformation to obtain the preset trajectory information of the preset multicellular model; Based on the affine transformation parameters, the preset trajectory information is subjected to affine transformation processing to obtain the target power scheduling curve of the load to be analyzed.

5. The method according to any one of claims 1 to 4, characterized in that, The construction of the target multicellular model of the load to be analyzed includes: Obtain information on the power regulation range, energy storage capacity, and disturbance parameters of the load to be analyzed; Based on the power adjustment range information, the energy storage capacity information, and the disturbance parameters, the constraint information corresponding to the load to be analyzed is determined; Based on the constraint information, construct the target multicellular model of the load to be analyzed.

6. The method according to claim 5, characterized in that, The process of fusing the affine transformation parameters to obtain the target affine transformation parameters of the flexible interconnected distribution network includes: Extract the first feature vector of the power regulation range information and the second feature vector of the energy storage capacity information; The first feature vector and the second feature vector are fused to obtain a fused feature vector. The fused feature vector is input into a pre-built importance prediction model to obtain the predicted importance of the load to be analyzed. Based on the weights corresponding to the predicted importance, the initial weights of the affine transformation parameters are determined. The initial weights are adjusted based on the perturbation parameters to obtain the target weights of the affine transformation parameters; According to the target weights, the affine transformation parameters are fused to obtain fused affine transformation parameters, which are used as the target affine transformation parameters of the flexible interconnected distribution network.

7. A power dispatching device for a flexible interconnected distribution network that considers the flexibility of energy storage-like resource aggregation, characterized in that, The device includes: The model building module is used to construct a target multi-cell model of the load to be analyzed; the load to be analyzed represents an energy storage-like load in a flexible interconnected distribution network; the target multi-cell model is used to represent a geometric model that matches the boundary conditions of the load operation characteristics of the load to be analyzed. The parameter determination module is used to obtain a preset multicellular model corresponding to the load to be analyzed, and to determine the affine transformation parameters of the load to be analyzed based on the target multicellular model and the preset multicellular model. The affine transformation parameters refer to the scaling factor and translation vector corresponding to the load to be analyzed. The scaling factor is used to represent the difference between the target multicellular model and the preset multicellular model in terms of power numerical scale. The translation vector is used to represent the overall positional offset between the target multicellular model and the preset multicellular model in the power dimension. The parameter fusion module is used to fuse the affine transformation parameters to obtain the target affine transformation parameters of the flexible interconnected distribution network. The curve prediction module is used to obtain the distribution network parameters and cost parameters of the flexible interconnected distribution network, and input the target affine transformation parameters, the distribution network parameters and the cost parameters into the trained power dispatch curve prediction model to obtain the power dispatch curve of the flexible interconnected distribution network. The curve determination module is used to determine the target power scheduling curve of the load to be analyzed based on the affine transformation parameters, the target affine transformation parameters, and the power scheduling curve. The power scheduling module is used to generate a power scheduling instruction for the load to be analyzed based on the target power scheduling curve, and to perform corresponding power scheduling processing on the load to be analyzed according to the power scheduling instruction.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images