A method, system, device and medium for evaluating and optimizing power data accuracy

Abstract

The present invention relates to a method, system, equipment and medium for evaluating and optimizing the accuracy of electric power data. The method comprises the following steps: collecting electric power system topology data and electric power data of each node of the electric power system; constructing an inter-node correlation topology matrix based on the electric power system topology data and the electric power data of each node of the electric power system; calculating the composite weight of each node based on the inter-node correlation topology matrix; constructing an electric power data accuracy evaluation model, evaluating the accuracy of the electric power data of each node based on the composite weight of each node, and obtaining an electric power data accuracy evaluation index for each node; iteratively optimizing the electric power data of nodes whose electric power data accuracy evaluation index is less than a preset threshold, stopping the iteration until an iterative convergence condition is reached, and obtaining optimized electric power data of the current node.

Description

Technical Field

[0001] The present invention relates to a method, system, equipment and medium for evaluating and optimizing power data accuracy, and belongs to the technical field of power data management. Background Art

[0002] With the rapid development of smart grid technology, power system monitoring and dispatching are increasingly dependent on the accuracy and dynamic trend analysis of node data. Traditional power data evaluation methods typically use static error analysis, simple statistical comparison, or basic differential methods. These methods gradually expose obvious shortcomings as the scale of the power grid expands and the complexity of the data increases significantly:

[0003] Most traditional methods are based on single-scale differential or linear trend calculations, which make it difficult to capture multi-scale trend changes in power grid operation, resulting in insufficient ability to evaluate the dynamic correlation of node data in the power grid and an increased risk of misjudgment or missed judgment.

[0004] Existing methods do not fully consider the sensitivity of power data change trends and the characteristics of their dynamic changes, ignore the differences in the speed of data changes at different nodes, make it difficult to achieve a rapid response to emergencies or abnormal fluctuations, and have major timeliness issues.

[0005] Traditional power data analysis methods usually assume that data is completely synchronized, ignoring the problems of asynchronous data sampling and trend time offset between nodes in actual power grid environments, making it difficult to achieve accurate node correlation calculation and accuracy assessment. Summary of the Invention

[0006] In order to solve the above problems existing in the prior art, the present invention proposes a method, system, device and medium for evaluating and optimizing power data accuracy.

[0007] The technical solutions of the present invention are as follows:

[0008] In one aspect, the present invention provides a method for evaluating and optimizing power data accuracy, comprising the following steps:

[0009] Collect power system topology data and power data of each node in the power system;

[0010] Constructing a correlation topology matrix between nodes based on the power system topology data and the power data of each node in the power system;

[0011] Calculate the composite weight of each node based on the association topology matrix between nodes;

[0012] Construct a power data accuracy evaluation model, evaluate the accuracy of power data of each node based on the composite weight of each node, and obtain the power data accuracy evaluation index of each node;

[0013] The power data of the nodes whose power data accuracy evaluation index is less than the preset threshold are iteratively optimized until the iteration convergence condition is reached and the optimized power data of the current node is obtained.

[0014] Preferably, the association topology matrix between nodes includes the topological association degree between each node, and the specific calculation formula of the topological association degree between each node is:

[0015] ;

[0016] in: express Time Node With node The topological correlation between them; express The correlation sensitivity factor of the moment; represents a nonlinear sensitive function; express Time Node With node The difference in changing trends between the corresponding power data; express Time Node With node The difference in changing trends between the corresponding power data; Indicates the total number of nodes.

[0017] Preferably, the composite weight of each node is calculated based on the association topology matrix between the nodes, as shown in the following formula:

[0018] ;

[0019] ;

[0020] in: express Time Node The composite weight of represents the normalization coefficient; The number of dimensions representing power data; express Time Node The power data of The actual value of the dimension; express All node power data at the moment The average value of the dimension; express All node power data at the moment The standard deviation of the dimension.

[0021] Preferably, a power data accuracy assessment model is constructed, as shown in the following formula:

[0022] ;

[0023] in: express Time Node Accuracy of power data; express The composite weight value of the node with the largest composite weight at the moment; express Time Node local deviations; express Time Node Neighborhood bias; express Time Node Global deviation of express Time Node local deviations; express Time Node Neighborhood bias; express Time Node global deviation.

[0024] Preferably, the power data of the nodes whose power data accuracy evaluation index is less than a preset threshold are iteratively optimized, as shown in the following formula:

[0025] ;

[0026] in: Indicates the After iterations Time Node Power data; Indicates the After iterations Time Node Power data; express Time Node Adaptive convergence step size; Indicates the After iterations Time Node Power data; express Time Node The trend impact adjustment factor; express Time Node The error gradient.

[0027] Preferably, the steps for calculating the adaptive convergence step are:

[0028] The initial history window encoding of the current node is extracted through the multi-layer perceptron, as shown in the following formula:

[0029] ;

[0030] in: Representation node The initial history window encoding of ; represents a multilayer perceptron; Indicates the history window size; express to Time Node Time series of electricity data;

[0031] For nodes whose power data accuracy evaluation index is less than the preset threshold, an adjacency matrix is ​​constructed. Based on the adjacency matrix and the initial historical window encoding of the current node, the optimized historical window encoding of the current node is calculated through a multi-layer graph convolutional network, as shown in the following formula:

[0032] ;

[0033] in: Represents the output of the last layer of the multi-layer graph convolutional network, that is, node Optimized history window encoding; Represents the last layer of a multi-layer graph convolutional network; represents the self-loop matrix constructed based on the adjacency matrix, , where represents the adjacency matrix, represents the identity matrix; represents the activation function; Represents a multi-layer graph convolutional network The output of the layer; Represents a multi-layer graph convolutional network The weight matrix of the layer; express degree matrix of ;

[0034] for Each sample in the code is marked with its time position, and the time position mark is constructed as a mark vector. After combining, input into the Transformer network to obtain the node The representation vector corresponding to the power data at each moment;

[0035] Node-based The characterization vector calculation node corresponding to the power data at each moment The adaptive convergence step size at each moment is as follows:

[0036] ;

[0037] in: Represents the weight of the adaptive convergence step; express Time Node The representation vector corresponding to the power data; Indicates the bias for the adaptive convergence step size.

[0038] Preferably, based on the node The characterization vector calculation node corresponding to the power data at each moment The trend impact adjustment factor at each moment is shown in the following formula:

[0039] ;

[0040] in: represents the weight of the trend impact adjustment factor; Represents the bias of the trend effect adjustment factor.

[0041] On the other hand, the present invention also provides a power data accuracy assessment and optimization system, including a data acquisition module, a correlation topology matrix construction module, a composite weight calculation module, a power data accuracy assessment module, and a power data optimization module;

[0042] The data acquisition module is used to collect power system topology data and power data of each node in the power system;

[0043] The association topology matrix construction module is used to construct an association topology matrix between nodes based on the power system topology data and the power data of each node in the power system;

[0044] The composite weight calculation module is used to calculate the composite weight of each node based on the association topology matrix between the nodes;

[0045] The power data accuracy assessment module is used to construct a power data accuracy assessment model, evaluate the accuracy of the power data of each node based on the composite weight of each node, and obtain the power data accuracy assessment index of each node;

[0046] The power data optimization module is used to iteratively optimize the power data of nodes whose power data accuracy evaluation index is less than a preset threshold, and stop the iteration when the iterative convergence condition is reached to obtain the optimized power data of the current node.

[0047] On the other hand, the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method of the present invention when executing the program.

[0048] In another aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the method of the present invention when the program is executed by a processor.

[0049] The present invention has the following beneficial effects:

[0050] 1. The present invention constructs a correlation topology matrix based on the topology and the original power data of the nodes, which can accurately reflect the actual coupling relationship of complex distribution network nodes and avoid the omission of weak coupling anomalies by traditional static or linear correlation.

[0051] 2. The present invention constructs a multi-level accuracy assessment model for evaluation, while constraining short-term fluctuations and long-term drifts, reducing missed reports and false alarms.

[0052] 3. The present invention constructs an electric power adaptive iterative model and calculates the adaptive step size, automatically increasing the correction amplitude for low-accuracy nodes. The iterative step size changes with the real-time state to avoid excessive oscillations or excessive stagnation, thereby ensuring the overall optimization process. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0054] 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 the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0055] It should be understood that the step numbers used herein are only for convenience of description and are not intended to limit the order in which the steps are to be executed.

[0056] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0057] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0058] The term "and / or" refers to and includes any and all possible combinations of one or more of the associated listed items.

[0059] Example 1:

[0060] See also Figure 1 , a method for evaluating and optimizing power data accuracy, comprising the following steps:

[0061] Collect power system topology data and power data from each node in the power system. In this embodiment, the data is collected for the power distribution network system in a certain area. The system includes 10 monitoring nodes, and each node collects three-dimensional power data of voltage, current, and frequency in real time.

[0062] Constructing a correlation topology matrix between nodes based on the power system topology data and the power data of each node in the power system;

[0063] Calculate the composite weight of each node based on the association topology matrix between nodes;

[0064] Construct a power data accuracy evaluation model, evaluate the accuracy of power data of each node based on the composite weight of each node, and obtain the power data accuracy evaluation index of each node;

[0065] The power data of the nodes whose power data accuracy evaluation index is less than the preset threshold are iteratively optimized until the iteration convergence condition is reached and the optimized power data of the current node is obtained.

[0066] As a preferred implementation of this embodiment, the association topology matrix between nodes includes the topological association degree between each node. The specific calculation formula of the topological association degree between each node is:

[0067] ;

[0068] in: express Time Node With node The topological correlation between them; express The correlation sensitivity factor of the moment; represents a nonlinear sensitive function; express Time Node With node The difference in changing trends between the corresponding power data; express Time Node With node The difference in changing trends between the corresponding power data; Indicates the total number of nodes;

[0069] The nonlinear sensitive function is specifically shown in the following formula:

[0070] ;

[0071] in: represents a natural constant; Represents the steepness parameter of the nonlinear curve (controls the drastic change of the nonlinear function); It represents the midpoint threshold of the nonlinear sensitive function. When the data difference is greater than this threshold, it significantly affects the correlation.

[0072] The steps for calculating the difference in change trends between the power data corresponding to the nodes are as follows:

[0073] Defining Nodes From the start of collection to Power data series at each moment for:

[0074] ;

[0075] in: Indicates the length of the acquisition time window of the power data series; express Time Node Power data;

[0076] calculate The multi-scale difference trend series is:

[0077] calculate The short-term difference trend series (first-order difference), as shown below:

[0078] ;

[0079] calculate The medium-term difference trend series (Second-order difference), as shown below:

[0080] ;

[0081] calculate Long-term difference trend series ,in Indicates the width is The sliding average function is as follows:

[0082] ;

[0083] in: Indicates Time Node Power data;

[0084] The dynamic time warping (DTW) algorithm is used to calculate the differences in the multi-scale difference trend series between nodes:

[0085] For nodes and nodes Define the difference as:

[0086] ;

[0087] in: Representation node and nodes Differences in short-term / medium-term / long-term differential trend series; represents dynamic time warping; Representation node Short-term / medium-term / long-term differential trend series; Representation node Short-term / medium-term / long-term differential trend series;

[0088] Node-based and nodes Difference calculation node of short-term / medium-term / long-term differential trend series and nodes The difference in changing trends between power data is shown in the following formula:

[0089] ;

[0090] in: Representation node exist The weight of the moment; Representation node exist The weight of the moment; Representation node and nodes The short-term differential trend series of The difference in moments; Representation node and nodes The medium-term difference trend series is The difference in moments; Representation node and nodes The long-term difference trend series of The difference of moments.

[0091] As a preferred implementation of this embodiment, the composite weight of each node is calculated based on the association topology matrix between nodes, as shown in the following formula:

[0092] ;

[0093] ;

[0094] in: express Time Node The composite weight of represents the normalization coefficient; The number of dimensions representing power data; express Time Node The power data of The actual value of the dimension; express All node power data at the moment The average value of the dimension; express All node power data at the moment The standard deviation of the dimension.

[0095] As a preferred implementation of this embodiment, a power data accuracy assessment model is constructed, as shown in the following formula:

[0096] ;

[0097] in: express Time Node Accuracy of power data; express The composite weight value of the node with the largest composite weight at the moment; express Time Node local deviations; express Time Node Neighborhood bias; express Time Node Global deviation of express Time Node local deviations; express Time Node Neighborhood bias; express Time Node Global deviation of Representation node middle The maximum value of

[0098] The calculation formula of the local deviation is:

[0099] ;

[0100] in: Represents the prediction of historical data trends based on machine learning models Time Node The power data of The actual value of the dimension;

[0101] The calculation formula of the neighborhood deviation is:

[0102] ;

[0103] in: Representation node The set of all nodes in the neighborhood; Representation node The total number of nodes in the neighborhood; express time Middle The power data of each node is The actual value of the dimension;

[0104] The calculation formula of the global deviation is: .

[0105] As a preferred implementation of this embodiment, iterative optimization is performed on the power data of nodes whose power data accuracy evaluation index is less than a preset threshold, as shown in the following formula:

[0106] ;

[0107] in: Indicates the After iterations Time Node Power data; Indicates the After iterations Time Node Power data; express Time Node Adaptive convergence step size; Indicates the After iterations Time Node Power data; express Time Node The trend impact adjustment factor; express Time Node The error gradient of

[0108] The calculation formula of the error gradient is:

[0109] ;

[0110] in: Indicates that in ideal case After iterations Time Node The power data estimation value is determined based on the power data of the previous iteration, such as by calculating the mean of the data in the window of the previous iteration using a window of a preset size as the power data estimation value; In the ideal case, After iterations Time Node Estimated value of electricity data.

[0111] In this embodiment, the above iterative steps are applied to the machine learning model and the loss function is constructed. , as shown in the following formula:

[0112] ;

[0113] in: express Time Node Actual power data; Indicates the final number of iterations; Indicates the Node after iteration Power data; Represents the stable convergence control coefficient, which is used to control whether to encourage stable convergence. The larger the value, the more conservative the optimization (stable convergence, slow speed), and the smaller the value, the more aggressive the optimization (faster, but higher risk). Represents the regularization factor that controls the stability of node iteration. If the node update is too drastic (large oscillation), then It will get bigger. Used to control the intensity of the punishment for this shock; represents the convergence stability index;

[0114] The calculation formula of the convergence stability index is:

[0115] ;

[0116] in: Indicates the After iterations Time Node The power data of The value of the dimension; express Time Node The power data of The actual value of the dimension; Represents a very small constant, used to avoid numerical instability caused by the denominator being zero;

[0117] When the loss function When the preset loss threshold is reached, the convergence condition is determined to have been reached and the iteration is stopped.

[0118] As a preferred implementation of this embodiment, the steps for calculating the adaptive convergence step size are:

[0119] The initial history window encoding of the current node is extracted through the multi-layer perceptron, as shown in the following formula:

[0120] ;

[0121] in: Representation node The initial history window encoding of ; represents a multilayer perceptron; Indicates the history window size; express to Time Node Time series of electricity data;

[0122] For nodes whose power data accuracy evaluation index is less than the preset threshold, an adjacency matrix is ​​constructed. Based on the adjacency matrix and the initial historical window encoding of the current node, the optimized historical window encoding of the current node is calculated through a multi-layer graph convolutional network, as shown in the following formula:

[0123] ;

[0124] in: Represents the output of the last layer of the multi-layer graph convolutional network, that is, node Optimized history window encoding; Represents the last layer of a multi-layer graph convolutional network; represents the self-loop matrix constructed based on the adjacency matrix, , where represents the adjacency matrix, represents the identity matrix; represents the activation function; Represents a multi-layer graph convolutional network The output of the layer; Represents a multi-layer graph convolutional network The weight matrix of the layer; express degree matrix of ;

[0125] for Each sample in the code is marked with its time position, and the time position mark is constructed as a mark vector. After combining, input into the Transformer network to obtain the node The representation vector corresponding to the power data at each moment;

[0126] The Transformer network itself is a disordered modeling structure that pays global attention to each item in the input sequence but does not retain sequential information.

[0127] For applications with strong temporal nature, such as power data, if time point information is not provided, the Transformer will be unable to distinguish the order of "past" and "present", which will seriously affect the effectiveness of trend modeling. Therefore, a "position vector" is added to each time point through a marker vector to help the model understand "time order".

[0128] Node-based The characterization vector calculation node corresponding to the power data at each moment The adaptive convergence step size at each moment is as follows:

[0129] ;

[0130] in: Represents the weight of the adaptive convergence step; express Time Node The representation vector corresponding to the power data; Indicates the bias for the adaptive convergence step size.

[0131] Node-based The characterization vector calculation node corresponding to the power data at each moment The trend impact adjustment factor at each moment is shown in the following formula:

[0132] ;

[0133] in: represents the weight of the trend impact adjustment factor; Represents the bias of the trend effect adjustment factor.

[0134] Example 2:

[0135] A power data accuracy assessment and optimization system includes a data acquisition module, a correlation topology matrix construction module, a composite weight calculation module, a power data accuracy assessment module, and a power data optimization module;

[0136] The data acquisition module is used to collect power system topology data and power data of each node in the power system;

[0137] The association topology matrix construction module is used to construct an association topology matrix between nodes based on the power system topology data and the power data of each node in the power system;

[0138] The composite weight calculation module is used to calculate the composite weight of each node based on the association topology matrix between the nodes;

[0139] The power data accuracy assessment module is used to construct a power data accuracy assessment model, evaluate the accuracy of the power data of each node based on the composite weight of each node, and obtain the power data accuracy assessment index of each node;

[0140] The power data optimization module is used to iteratively optimize the power data of nodes whose power data accuracy evaluation index is less than a preset threshold, and stop the iteration when the iterative convergence condition is reached to obtain the optimized power data of the current node.

[0141] This system is used to implement the method in Example 1 and will not be described in detail here.

[0142] Example 3:

[0143] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method described in any embodiment of the present invention is implemented.

[0144] Example 4:

[0145] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method described in any embodiment of the present invention is implemented.

[0146] In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can represent: a, b, c, a and b, a and c, b and c or a and b and c, where a, b, c can be single or multiple.

[0147] Those skilled in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented using a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0148] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0149] In the several embodiments provided in this application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of this application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory; hereinafter referred to as: ROM), random access memory (Random Access Memory; hereinafter referred to as: RAM), magnetic disk or optical disk, and other media that can store program code.

[0150] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for evaluating and optimizing power data accuracy, characterized in that: The following steps are involved: Collect power system topology data and power data of each node in the power system; Constructing a correlation topology matrix between nodes based on the power system topology data and the power data of each node in the power system; The topological matrix of the association between nodes includes the topological association degree between each node. The specific calculation formula of the topological association degree between each node is: ； in: express Time Node With node The topological correlation between them; express The correlation sensitivity factor of the moment; represents a nonlinear sensitive function; express Time Node With node The difference in changing trends between the corresponding power data; express Time Node With node The difference in changing trends between the corresponding power data; Indicates the total number of nodes; The composite weight of each node is calculated based on the association topology matrix between nodes, as shown in the following formula: ； ； in: express Time Node The composite weight of represents the normalization coefficient; The number of dimensions representing power data; express Time Node The power data of The actual value of the dimension; express All node power data at the moment The average value of the dimension; express All node power data at the moment Standard deviation of the dimension; Construct a power data accuracy evaluation model, evaluate the accuracy of power data of each node based on the composite weight of each node, and obtain the power data accuracy evaluation index of each node; The power data of the nodes whose power data accuracy evaluation index is less than the preset threshold are iteratively optimized until the iteration convergence condition is reached and the optimized power data of the current node is obtained.

2. A method for evaluating and optimizing power data accuracy according to claim 1, characterized in that: Construct a power data accuracy assessment model, as shown in the following formula: ； in: express Time Node Accuracy of power data; express The composite weight value of the node with the largest composite weight at the moment; express Time Node local deviations; express Time Node Neighborhood bias; express Time Node Global deviation of express Time Node local deviations; express Time Node Neighborhood bias; express Time Node global deviation.

3. The method for evaluating and optimizing power data accuracy according to claim 1, wherein: The power data of nodes whose power data accuracy evaluation index is less than the preset threshold is iteratively optimized, as shown in the following formula: ； in: Indicates the After iterations Time Node Power data; Indicates the After iterations Time Node Power data; express Time Node Adaptive convergence step size; Indicates the After iterations Time Node Power data; express Time Node The trend impact adjustment factor; express Time Node The error gradient.

4. A method for evaluating and optimizing power data accuracy according to claim 3, characterized in that: The calculation steps of the adaptive convergence step are: The initial history window encoding of the current node is extracted through the multi-layer perceptron, as shown in the following formula: ； in: Representation node The initial history window encoding of ; represents a multilayer perceptron; Indicates the history window size; express to Time Node Time series of electricity data; For nodes whose power data accuracy evaluation index is less than the preset threshold, an adjacency matrix is ​​constructed. Based on the adjacency matrix and the initial historical window encoding of the current node, the optimized historical window encoding of the current node is calculated through a multi-layer graph convolutional network, as shown in the following formula: ； in: Represents the output of the last layer of the multi-layer graph convolutional network, that is, node Optimized history window encoding; Represents the last layer of a multi-layer graph convolutional network; represents the self-loop matrix constructed based on the adjacency matrix, , where represents the adjacency matrix, represents the identity matrix; represents the activation function; Represents a multi-layer graph convolutional network The output of the layer; Represents a multi-layer graph convolutional network The weight matrix of the layer; express degree matrix of ; for Each sample in the code is marked with its time position, and the time position mark is constructed as a mark vector. After combining, input into the Transformer network to obtain the node The representation vector corresponding to the power data at each moment; Node-based The characterization vector calculation node corresponding to the power data at each moment The adaptive convergence step size at each moment is as follows: ； in: Represents the weight of the adaptive convergence step; express Time Node The representation vector corresponding to the power data; Indicates the bias for the adaptive convergence step size.

5. A method for evaluating and optimizing power data accuracy according to claim 4, characterized in that: Node-based The characterization vector calculation node corresponding to the power data at each moment The trend impact adjustment factor at each moment is shown in the following formula: ； in: represents the weight of the trend impact adjustment factor; represents the bias of the trend effect adjustment factor.

6. A power data accuracy assessment and optimization system, characterized in that: The method according to any one of claims 1 to 5, comprising a data acquisition module, an association topology matrix construction module, a composite weight calculation module, an electric power data accuracy assessment module, and an electric power data optimization module; The data acquisition module is used to collect power system topology data and power data of each node in the power system; The association topology matrix construction module is used to construct an association topology matrix between nodes based on the power system topology data and the power data of each node in the power system; The composite weight calculation module is used to calculate the composite weight of each node based on the association topology matrix between the nodes; The power data accuracy assessment module is used to construct a power data accuracy assessment model, evaluate the accuracy of the power data of each node based on the composite weight of each node, and obtain the power data accuracy assessment index of each node; The power data optimization module is used to iteratively optimize the power data of nodes whose power data accuracy evaluation index is less than a preset threshold, and stop the iteration when the iterative convergence condition is reached to obtain the optimized power data of the current node.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images