A regional power dispatching method and system for power equipment

Abstract

The present invention discloses a method and system for regional power dispatching for power equipment. The method includes fusing power parameters and power grid topology data to obtain a spatiotemporal correlation feature dataset; extracting features from the spatiotemporal correlation feature dataset to obtain a device state representation vector; inputting the device state representation vector into a constructed spatiotemporal attention model to obtain dynamic coupling features; constructing a regional power demand model for the target power system based at least on the dynamic coupling features, solving the regional power demand model using a mixed integer linear programming algorithm to obtain several dynamic load distribution strategies; establishing a power grid state assessment model; and evaluating and comparing each of the dynamic load distribution strategies based on the power grid state assessment model, and executing a power dispatching plan for the target power system determined by the evaluation and comparison results. The method provided by the embodiments of the present invention can improve the accuracy of power dispatching.

Description

Technical Field

[0001] The present invention relates to the technical field of power dispatching control, and in particular to a regional power dispatching method and system for power equipment. Background Art

[0002] With the continuous development of social economy, in modern power systems, the transmission and dispatch of electric energy are crucial components of power grid security.

[0003] Existing regional power dispatch methods only consider demand-side power quotas. For example, dispatch is based solely on user-declared power usage or demand response contracts. However, such methods fail to account for the full range of power system demand, resulting in low dispatch accuracy.

[0004] Therefore, how to design a power system electricity dispatching method and improve the accuracy of power dispatching has become a technical problem that needs to be solved urgently by those skilled in the art. Summary of the Invention

[0005] The present invention provides a regional power dispatching method and system for power equipment to solve the technical problem of low accuracy of current dispatching and allocation.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides a regional power dispatching method for power equipment, comprising:

[0007] The acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature dataset;

[0008] Extracting features from the spatiotemporal correlation feature data set to obtain a device state representation vector;

[0009] Inputting the device state representation vector into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has power load attributes and node location attributes;

[0010] constructing a regional electricity demand model of the target power system based at least on the dynamic coupling characteristics, solving the regional electricity demand model using a mixed integer linear programming algorithm to obtain a plurality of dynamic load distribution strategies, wherein the constraints of the regional electricity demand model include at least minimizing power loss;

[0011] Using a graph neural network method to process the grid topology data and the status data of each power device in the target power system, and establishing a grid status assessment model based on the processed data;

[0012] Based on the power grid state assessment model, each of the dynamic load distribution strategies is evaluated and compared, and a power dispatching plan for the target power system determined by the evaluation and comparison results is executed.

[0013] As one preferred solution, before fusing the acquired power parameters and grid topology data of the target power system, the regional power dispatching method for power equipment further includes:

[0014] Acquiring power data of power system equipment in the target power system, performing standardization processing on the power data, and obtaining power parameters of the target power system;

[0015] Based on the acquired location of each of the power system devices and the connection relationship between the power system devices, a power grid topology structure is constructed to obtain the power grid topology structure data.

[0016] As one of the preferred solutions, the acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature dataset. The fusion process includes:

[0017] Based on the timestamp, performing spatiotemporal alignment on the power parameter and the grid topology data;

[0018] The aligned power parameters and grid topology data are fused using a multi-source data fusion method to obtain the spatiotemporal correlation feature data set.

[0019] As one of the preferred solutions, the device state representation vector is input into the constructed spatiotemporal attention model to obtain dynamic coupling features, including:

[0020] Construct a spatiotemporal attention model using a convolutional neural network containing a spatiotemporal attention layer;

[0021] The device state representation vector is processed using the spatiotemporal attention model to obtain a power load feature vector in the time dimension and a node position feature vector in the space dimension;

[0022] The power load characteristic vector and the node position characteristic vector are fused to obtain a dynamic coupling feature with power load attributes and node position attributes.

[0023] As one preferred solution, the regional electricity demand model is solved using a mixed integer linear programming algorithm to obtain several dynamic load distribution strategies. The solution process includes:

[0024] Establishing load balancing constraints based on the power grid topology data and the acquired real-time power demand data;

[0025] Establishing grid operation constraints based on the operating parameters of the power equipment of the target power system;

[0026] Determine the output limit of the generator set according to the maximum output power of the generator set;

[0027] The load balancing constraint, the grid operation constraint, and the output limit of the generator set are integrated, and the integrated result is used as the constraint equation of the mixed integer linear programming. The mixed integer linear programming problem is solved using a mixed integer linear programming solver to obtain a dynamic load distribution strategy.

[0028] As one preferred solution, after solving the constructed regional electricity demand model using mixed integer linear programming to obtain a dynamic load distribution strategy, the regional electricity dispatching method for power equipment further includes:

[0029] Acquiring power operation parameter data corresponding to the target power system in the current dynamic load distribution strategy;

[0030] Acquiring historical scenario data that matches the current dynamic load distribution strategy;

[0031] constructing an error function based on the comparison and analysis results of the historical scenario data and the power operation parameter data, and estimating the error function using the least squares method to obtain an optimal parameter value;

[0032] The dynamic load distribution strategy is optimized using the optimal parameter values.

[0033] As one preferred solution, based on the grid state assessment model, each of the dynamic load distribution strategies is evaluated and compared, and the evaluation and comparison process includes:

[0034] fusing the grid topology and the status data of each power device in the target power system to obtain input constraints;

[0035] Inputting the input constraints into the power grid state assessment model, calculating the power grid state assessment model using a multi-objective optimization algorithm to obtain a non-dominated solution set, wherein each solution in the non-dominated solution set is a power dispatching scheme;

[0036] Each solution in the non-dominated solution set is evaluated to obtain an optimal power scheduling solution.

[0037] As one preferred solution, after executing the power dispatching plan of the target power system determined by the evaluation and comparison results, the regional power dispatching method for power equipment further includes:

[0038] Regularly obtaining the power status of the power equipment corresponding to each node in the target power system;

[0039] Performing a fault assessment on the power status of each of the power devices to obtain a status aging rate;

[0040] Inputting the state aging rate into a power grid state assessment model to obtain a real-time risk assessment result;

[0041] The real-time risk assessment results are analyzed and processed using a natural language model, a processing log is generated, and the log is sent to a processing security terminal.

[0042] As one preferred solution, after executing the power dispatching plan of the target power system determined by the evaluation and comparison results, the regional power dispatching method for power equipment further includes:

[0043] Obtaining the execution result of the power scheduling plan;

[0044] Performing a threshold analysis on each node in the execution result, determining a faulty node according to a preset threshold, and determining a geographical location corresponding to the faulty node according to the power grid topology data;

[0045] The geographical location corresponding to the faulty node is sent to a maintenance terminal.

[0046] Another embodiment of the present invention provides a regional power dispatching system for power equipment, including:

[0047] A fusion module is used to fuse the acquired power parameters and grid topology data of the target power system to obtain a temporal and spatial correlation feature data set;

[0048] An extraction module, configured to extract features from the spatiotemporal correlation feature data set to obtain a device state representation vector;

[0049] a processing module, configured to input the device state representation vector into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has a power load attribute and a node location attribute;

[0050] a solving module, configured to construct a regional electricity demand model of the target power system based at least on the dynamic coupling characteristics, and solve the regional electricity demand model using a mixed integer linear programming algorithm to obtain a plurality of dynamic load distribution strategies, wherein the constraints of the regional electricity demand model include at least minimizing power loss;

[0051] A construction module is used to process the power grid topology data and the status data of each power device in the target power system using a graph neural network method, and establish a power grid status assessment model based on the processed data;

[0052] An evaluation module is used to evaluate and compare each of the dynamic load distribution strategies based on the grid state evaluation model, and execute the power dispatching plan of the target power system determined by the evaluation and comparison results.

[0053] Compared with the prior art, the embodiments of the present invention have the following advantages:

[0054] The acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature data set; feature extraction is performed on the spatiotemporal correlation feature data set to obtain a device state representation vector; the device state representation vector is input into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has power load attributes and node location attributes; a regional power demand model of the target power system is constructed at least based on the dynamic coupling feature, and the regional power demand model is solved using a mixed integer linear programming algorithm to obtain several dynamic load distribution strategies, wherein the constraint conditions of the regional power demand model at least include the goal of minimizing power loss; the grid topology data and the state data of each power device in the target power system are processed using a graph neural network method, and a grid state assessment model is established based on the processed data; based on the grid state assessment model, each of the dynamic load distribution strategies is evaluated and compared, and a power dispatching plan for the target power system determined by the evaluation and comparison results is executed. Compared with the existing technology, the present invention takes all-round demand consideration into account by combining the grid topology data and the status data of the power equipment, with the goal of minimizing power loss, and provides a regional power scheduling method for power equipment to improve the accuracy of power scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flowchart of a method for regional power dispatching for power equipment in one embodiment of the present invention;

[0056] Figure 2 It is a structural diagram of a regional power dispatching system for power equipment in one embodiment of the present invention.

[0057] Reference numerals:

[0058] Among them, 11, fusion module; 12, extraction module; 13, processing module; 14, solution module; 15, construction module; 16, evaluation module. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0060] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meanings as those commonly understood by those skilled in the art. The terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. Those skilled in the art will understand the specific meanings of the above terms in this application in specific circumstances.

[0061] An embodiment of the present invention provides a regional power dispatching method for power equipment. For details, see Figure 1 , Figure 1 The figure shows a flow chart of a method for regional power dispatching for power equipment in one embodiment of the present invention, the method comprising:

[0062] S1: The acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature dataset.

[0063] Specifically, before fusing the acquired power parameters and grid topology data of the target power system, it is necessary to obtain the power data of the power system equipment of the target power system from an acquisition system such as a smart meter or a SCADA distribution monitoring system, standardize the power data, and obtain the power parameters of the target power system.

[0064] Specifically, the longitude and latitude coordinates or logical position of each power system equipment is obtained from Beidou satellite, Galileo or GPS. The logical position refers to the substation number and busbar identification. The electrical connection information between devices is extracted from the SCADA system, design drawings or asset management system. The connection relationship is represented by an adjacency matrix or adjacency table. The power grid topology structure is constructed based on the location of each power system equipment and the connection relationship between each power system equipment to obtain the power grid topology structure data.

[0065] During the fusion processing of the acquired power parameters and grid topology data of the target power system, all data can be sorted chronologically based on timestamps, and the power parameters and grid topology data can be aligned in time and space. IRIG-B codes or PTP protocols can also be used to synchronize the microsecond clocks of SCADA, PMUs, meters, and other equipment. Possible time delays or missing data can be addressed using methods such as interpolation, smoothing, or data padding.

[0066] Key features that can reflect the operating status and change trends of the power system are extracted from power parameter and grid topology data. The extracted features are fused using multi-source data fusion methods to obtain a spatiotemporal correlation feature dataset.

[0067] The multi-source data fusion method is preferably Kalman filtering or Bayesian network.

[0068] In step S1, the power parameters and grid topology data are fused to generate a spatiotemporal correlation feature dataset, which can capture the dynamic relationship between the power parameters and the grid structure in the power system and provide a basis for subsequent analysis.

[0069] S2: Extract features from the spatiotemporal correlation feature dataset to obtain a device state representation vector.

[0070] Specifically, the process of feature extraction for the spatiotemporal correlation feature data set includes: extracting statistical features, time-frequency features, and spatial features, and combining these feature values ​​to form device state representation vectors. These vectors usually have fixed dimensions, and each dimension corresponds to a specific feature.

[0071] It should be noted that the quality of the device state representation vector depends on the selection of feature extraction methods and parameter settings. In practical applications, it is necessary to select a suitable feature extraction method and make appropriate parameter adjustments based on the characteristics of the data and the analysis objectives.

[0072] S3: Input the device state representation vector into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has power load attributes and node location attributes.

[0073] Specifically, a spatiotemporal attention model is constructed by a convolutional neural network including a spatiotemporal attention layer, wherein the convolutional neural network includes at least an input layer, a convolution layer, a pooling layer, a connection layer, a spatiotemporal attention layer and an output layer.

[0074] Preferably, the spatiotemporal attention layer is located after the convolution layer and before the pooling layer. After the convolution layer and before the pooling layer, the spatiotemporal attention layer can directly act on the feature map output by the convolution layer to enhance the capture of important local features.

[0075] Specifically, by inputting the device state representation vector into the spatiotemporal attention model, the power load feature vector in the time dimension and the node position feature vector in the space dimension can be obtained respectively. The power load feature vector and the node position feature vector are fused to obtain a dynamic coupling feature with power load attributes and node position attributes.

[0076] S4: Constructing a regional electricity demand model of the target power system at least based on the dynamic coupling characteristics, solving the regional electricity demand model using a mixed integer linear programming algorithm to obtain several dynamic load distribution strategies, wherein the constraints of the regional electricity demand model at least include the goal of minimizing power loss.

[0077] Specifically, with the goal of minimizing power loss, load balancing constraints are constructed based on the grid topology data and the acquired real-time power demand data; grid operation constraints are constructed based on the operating parameters of the power equipment in the target power system; and the output limits of the generator sets are determined based on the maximum output power of the generator sets. The load balancing constraints, grid operation constraints, and the output limits of the generator sets are integrated, and the integrated result is used as the constraint equation for the mixed integer linear programming.

[0078] Specifically, based on the grid topology data and real-time electricity demand data, a load balance equation is constructed to ensure the balance between electricity supply and demand. The load balance constraint usually involves the power flow equation, which is used to describe the flow and distribution of electricity in the grid.

[0079] Based on the operating parameters of the power equipment in the target power system, grid operation constraints are constructed. These constraints include voltage range limitations, current limitations, line capacity limitations, etc., to ensure that the grid operates under safe and stable conditions.

[0080] According to the maximum output power of the generator set, the output limit of the generator set is determined. These limits involve the start and stop status, output range, etc. of the generator set to ensure that the output of the generator set fluctuates within a reasonable range.

[0081] A reinforcement learning algorithm is used to process dynamic coupling features and construct a regional electricity demand model. This regional electricity demand model is solved using a mixed integer linear programming algorithm, resulting in several dynamic load allocation strategies. A reinforcement learning algorithm is used to process dynamic coupling features and extract key information for constructing the regional electricity demand model. Through a trial-and-error reinforcement learning process, model parameters are continuously adjusted to minimize energy loss and optimize the predictive performance of the regional electricity demand model.

[0082] According to the solution results, several dynamic load distribution strategies are obtained, which minimize the power loss while satisfying all constraints.

[0083] After the dynamic load distribution strategy is obtained, the power operation parameter data corresponding to the target power system in the current dynamic load distribution strategy is obtained; and historical scenario data matching the current dynamic load distribution strategy is obtained.

[0084] An error function is constructed based on the comparison and analysis results of the historical scenario data and the power operation parameter data, the error function is estimated using the least squares method to obtain optimal parameter values, and the dynamic load distribution strategy is optimized using the optimal parameter values.

[0085] S5: Using a graph neural network method to process the grid topology data and the status data of each power device in the target power system, and establishing a grid status assessment model based on the processed data.

[0086] The graph neural network method can process power grid topology data and power equipment status data, capture the complex relationships between nodes and edges in the power grid, and establish a power grid status assessment model based on the processed data to evaluate the operating status and potential risks of the power grid.

[0087] At the same time, in this step, the power loss of different types of power equipment, that is, the status data, is combined. According to the combination of the power loss of different types of power equipment, the accuracy of scheduling allocation can be greatly improved.

[0088] S6: Based on the grid state evaluation model, each of the dynamic load distribution strategies is evaluated and compared, and a power dispatching plan for the target power system determined by the evaluation and comparison results is executed.

[0089] The evaluation and comparison process includes: fusing the grid topology and the target power equipment state to obtain input constraints; inputting the input constraints into the grid state evaluation model, and calculating the grid state evaluation model using a multi-objective optimization algorithm to obtain a non-dominated solution set, wherein each solution in the non-dominated solution set is a power dispatching plan.

[0090] The solutions in the non-dominated solution set are evaluated and analyzed to assess their performance on various objectives. Based on actual needs and the results of the solution set analysis, an optimal solution is selected as the final power dispatching plan.

[0091] After executing the power dispatching plan of the target power system determined by the evaluation and comparison results, the regional power dispatching method for power equipment further includes:

[0092] Regularly obtain real-time or recent power status data for power equipment at each node in the target power system from sensors and monitoring systems. Leveraging existing fault assessment models or algorithms, analyze the collected power equipment status data to identify any abnormalities or signs of faults. Based on the fault assessment results, calculate the aging rate of each device. The aging rate is a quantitative indicator used to measure the degree of performance degradation or aging of equipment due to factors such as long-term operation, environmental factors, or load fluctuations.

[0093] The state aging rate is input into the power grid state assessment model to obtain real-time risk assessment results; the power grid risk assessment results include multiple dimensions such as failure probability, impact range, and potential losses.

[0094] Real-time risk assessment results are analyzed and processed using natural language models to generate a processing log. This log is then sent to the security processing terminal via appropriate communication channels, ensuring that relevant personnel have timely access to risk assessment results and recommended actions. The processing log contains information such as a risk assessment overview, key findings, and recommended actions.

[0095] Preferably, after executing the power dispatching plan of the target power system determined by the evaluation and comparison results, the regional power dispatching method for power equipment further includes:

[0096] Obtain the execution results of the power dispatch plan, including key parameters such as current, voltage, power factor, and load distribution of each node; set reasonable thresholds for each key parameter based on the operating standards and experience of the power system. These thresholds are used to determine whether the node is in normal operation. Compare the execution result data with the set thresholds to identify nodes that exceed the thresholds, i.e., faulty nodes.

[0097] Utilizing grid topology data, including node connectivity and geographic location information, the specific location of the faulty node within the grid is determined. Using a geographic information system (GIS), the faulty node is mapped to its actual geographic location, making it easier for maintenance personnel to quickly locate the fault.

[0098] The geographical location corresponding to the faulty node is sent to a maintenance terminal, and a maintenance terminal is determined to receive the geographical location information, which may be a mobile device of a maintenance person, a monitoring system of a power grid maintenance center, or the like.

[0099] An embodiment of the present invention provides a regional power dispatching system for power equipment. For details, see Figure 2 , Figure 2 FIG. 1 is a schematic diagram showing the structure of a regional power dispatching system for power equipment in one embodiment of the present invention. The system includes:

[0100] A fusion module 11 is used to fuse the acquired power parameters and grid topology data of the target power system to obtain a spatiotemporal correlation feature data set;

[0101] An extraction module 12 is used to extract features from the spatiotemporal correlation feature data set to obtain a device state representation vector;

[0102] A processing module 13 is configured to input the device state representation vector into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has power load attributes and node location attributes;

[0103] a solving module 14, configured to construct a regional electricity demand model of the target power system based at least on the dynamic coupling characteristics, and solve the regional electricity demand model using a mixed integer linear programming algorithm to obtain a plurality of dynamic load distribution strategies, wherein the constraints of the regional electricity demand model include at least minimizing power loss;

[0104] A construction module 15 is configured to process the power grid topology data and the status data of each power device in the target power system using a graph neural network method, and establish a power grid status assessment model based on the processed data;

[0105] The evaluation module 16 is configured to evaluate and compare each of the dynamic load distribution strategies based on the grid state evaluation model, and execute a power dispatching plan for the target power system determined by the evaluation and comparison results.

[0106] Compared with the prior art, the embodiments of the present invention have the following advantages:

[0107] The acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature data set; feature extraction is performed on the spatiotemporal correlation feature data set to obtain a device state representation vector; the device state representation vector is input into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has power load attributes and node location attributes; a regional power demand model of the target power system is constructed at least based on the dynamic coupling feature, and the regional power demand model is solved using a mixed integer linear programming algorithm to obtain several dynamic load distribution strategies, wherein the constraint conditions of the regional power demand model at least include the goal of minimizing power loss; the grid topology data and the state data of each power device in the target power system are processed using a graph neural network method, and a grid state assessment model is established based on the processed data; based on the grid state assessment model, each of the dynamic load distribution strategies is evaluated and compared, and a power dispatching plan for the target power system determined by the evaluation and comparison results is executed. Compared with the existing technology, the present invention takes all-round demand consideration into account by combining the grid topology data and the status data of the power equipment, with the goal of minimizing power loss, and provides a regional power scheduling method for power equipment to improve the accuracy of power scheduling.

[0108] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A regional power dispatching method for power equipment, characterized in that: include: The acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature dataset; Extracting features from the spatiotemporal correlation feature data set to obtain a device state representation vector; Inputting the device state representation vector into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has power load attributes and node location attributes; constructing a regional electricity demand model of the target power system based at least on the dynamic coupling characteristics, solving the regional electricity demand model using a mixed integer linear programming algorithm to obtain a plurality of dynamic load distribution strategies, wherein the constraints of the regional electricity demand model include at least minimizing power loss; Using a graph neural network method to process the grid topology data and the status data of each power device in the target power system, and establishing a grid status assessment model based on the processed data; Based on the grid state assessment model, each of the dynamic load distribution strategies is evaluated and compared, and a power dispatching plan for the target power system determined by the evaluation and comparison results is executed.

2. The regional power dispatching method for power equipment according to claim 1, characterized in that: Before fusing the acquired power parameters and grid topology data of the target power system, the regional power dispatching method for power equipment further includes: Acquiring power data of power system equipment in the target power system, performing standardization processing on the power data, and obtaining power parameters of the target power system; Based on the acquired location of each of the power system devices and the connection relationship between the power system devices, a power grid topology structure is constructed to obtain the power grid topology structure data.

3. The regional power dispatching method for power equipment according to claim 1, characterized in that: The acquired power parameters and grid topology data of the target power system are fused to obtain a spatiotemporal correlation feature dataset, and the fusion process includes: Based on the timestamp, performing spatiotemporal alignment on the power parameter and the grid topology data; The aligned power parameters and grid topology data are fused using a multi-source data fusion method to obtain the spatiotemporal correlation feature data set.

4. The regional power dispatching method for power equipment according to claim 1, characterized in that: The device state representation vector is input into the constructed spatiotemporal attention model to obtain dynamic coupling features, including: Construct a spatiotemporal attention model using a convolutional neural network containing a spatiotemporal attention layer; The device state representation vector is processed using the spatiotemporal attention model to obtain a power load feature vector in the time dimension and a node position feature vector in the space dimension; The power load characteristic vector and the node position characteristic vector are fused to obtain a dynamic coupling feature with power load attributes and node position attributes.

5. The regional power dispatching method for power equipment according to claim 1, characterized in that: The mixed integer linear programming algorithm is used to solve the regional electricity demand model to obtain several dynamic load distribution strategies. The solution process includes: Establishing load balancing constraints based on the power grid topology data and the acquired real-time power demand data; Establishing grid operation constraints based on the operating parameters of the power equipment of the target power system; Determine the output limit of the generator set according to the maximum output power of the generator set; The load balancing constraint, the grid operation constraint, and the output limit of the generator set are integrated, and the integrated result is used as the constraint equation of the mixed integer linear programming. The mixed integer linear programming problem is solved using a mixed integer linear programming solver to obtain a dynamic load distribution strategy.

6. The regional power dispatching method for power equipment according to claim 1, characterized in that: After solving the constructed regional electricity demand model using mixed integer linear programming to obtain a dynamic load distribution strategy, the regional electricity dispatching method for power equipment further includes: Acquiring power operation parameter data corresponding to the target power system in the current dynamic load distribution strategy; Acquiring historical scenario data that matches the current dynamic load distribution strategy; constructing an error function based on the comparison and analysis results of the historical scenario data and the power operation parameter data, and estimating the error function using the least squares method to obtain an optimal parameter value; The dynamic load distribution strategy is optimized using the optimal parameter values.

7. The regional power dispatching method for power equipment according to claim 1, characterized in that: The evaluation and comparison of each of the dynamic load distribution strategies is performed based on the grid state evaluation model, wherein the evaluation and comparison process includes: fusing the grid topology and the status data of each power device in the target power system to obtain input constraints; Inputting the input constraints into the power grid state assessment model, calculating the power grid state assessment model using a multi-objective optimization algorithm to obtain a non-dominated solution set, wherein each solution in the non-dominated solution set is a power dispatching scheme; Each solution in the non-dominated solution set is evaluated to obtain an optimal power scheduling solution.

8. The regional power dispatching method for power equipment according to claim 1, characterized in that: After executing the power dispatching plan of the target power system determined by the evaluation and comparison results, the regional power dispatching method for power equipment further includes: Regularly obtaining the power status of the power equipment corresponding to each node in the target power system; Performing a fault assessment on the power status of each of the power devices to obtain a status aging rate; Inputting the state aging rate into a power grid state assessment model to obtain a real-time risk assessment result; The real-time risk assessment results are analyzed and processed using a natural language model, a processing log is generated, and the log is sent to a processing security terminal.

9. The regional power dispatching method for power equipment according to claim 1, characterized in that: After executing the power dispatching plan of the target power system determined by the evaluation and comparison results, the regional power dispatching method for power equipment further includes: Obtaining the execution result of the power scheduling plan; Performing a threshold analysis on each node in the execution result, determining a faulty node according to a preset threshold, and determining a geographical location corresponding to the faulty node according to the power grid topology data; The geographical location corresponding to the faulty node is sent to a maintenance terminal.

10. A regional power dispatching system for power equipment, characterized in that: include: A fusion module is used to fuse the acquired power parameters and grid topology data of the target power system to obtain a temporal and spatial correlation feature data set; An extraction module, configured to extract features from the spatiotemporal correlation feature data set to obtain a device state representation vector; a processing module, configured to input the device state representation vector into the constructed spatiotemporal attention model to obtain a dynamic coupling feature, wherein the dynamic coupling feature has a power load attribute and a node location attribute; a solving module, configured to construct a regional electricity demand model of the target power system based at least on the dynamic coupling characteristics, and solve the regional electricity demand model using a mixed integer linear programming algorithm to obtain a plurality of dynamic load distribution strategies, wherein the constraints of the regional electricity demand model include at least minimizing power loss; A construction module is used to process the power grid topology data and the status data of each power device in the target power system using a graph neural network method, and establish a power grid status assessment model based on the processed data; An evaluation module is used to evaluate and compare each of the dynamic load distribution strategies based on the grid state evaluation model, and execute the power dispatching plan of the target power system determined by the evaluation and comparison results.

Images