An abnormal power meteorological data correction method and device for power planning

Abstract

The application discloses an abnormal power meteorological data correction method and device for power planning, relates to the power technology field, and mainly aims to improve the accuracy of power system operation and planning. The main technical scheme of the application is as follows: obtaining initial power meteorological data sequences corresponding to each meteorological parameter in each transformer substation; constructing a power grid topology graph according to the initial power meteorological data sequences, the topology structure of a power transmission network, the geographical relationship between each transformer substation and the admittance matrix relationship; determining the space-time correlation characteristics between each node in the power grid topology graph based on a preset construction method; identifying abnormal data in the initial power meteorological data sequences corresponding to each meteorological parameter in each transformer substation based on the power grid topology graph and the space-time correlation characteristics between each node; and correcting the abnormal data according to the space-time correlation characteristics between each node and a preset data correction rule to obtain optimized power meteorological data, so that power planning is performed according to the optimized power meteorological data.

Description

Technical Field

[0001] This invention relates to the field of power technology, and in particular to a method and apparatus for correcting abnormal power meteorological data for power planning. Background Technology

[0002] Power meteorological data refers to meteorological information closely related to the operation of the power system, covering key elements such as temperature, humidity, wind speed, rainfall, and air pressure. However, in the actual data collection process, due to potential equipment malfunctions, human errors, or temporary impacts of extreme weather conditions on equipment, power meteorological data often suffers from data quality issues such as missing values, outliers, and jumps. If these problems are not addressed promptly, they will seriously affect the accuracy and reliability of power operation planning.

[0003] Currently, the standard practice for correcting power meteorological data is as follows: First, obtain the power meteorological data sequence for each substation within a specific time period. Then, use fixed threshold or sliding window statistical methods to detect anomalies in the meteorological data for each station. After identifying abnormal data, interpolation is typically used to fill in missing data, data is corrected based on statistics within the sliding window, and smoothing methods are applied to reduce data jumps.

[0004] However, this method of processing time-series data from a single substation has significant limitations. Because it focuses only on data changes at a single station, it overlooks natural phenomena that should be caused by changes at surrounding stations, which are then misidentified as anomalous data. This misidentification not only leads to incorrect corrections of normal data but may also introduce new data errors, thereby reducing the accuracy of power operation planning. Summary of the Invention

[0005] In view of the above problems, the present invention provides a method and apparatus for correcting abnormal power meteorological data for power planning, the main purpose of which is to improve the accuracy of power system operation and planning.

[0006] To solve the above-mentioned technical problems, the present invention proposes the following solution:

[0007] In a first aspect, the present invention provides a method for correcting abnormal power meteorological data for power planning, the method comprising:

[0008] Obtain the initial power meteorological data sequence corresponding to each meteorological parameter in each substation;

[0009] Based on the initial power meteorological data sequence, the topology of the transmission network, the geographical relationships between substations and the admittance matrix relationships, a power grid topology map is constructed. The power grid topology map uses substations as nodes and transmission lines as edges to characterize the connection relationships between nodes and their spatial distribution characteristics.

[0010] Based on a preset construction method, the spatiotemporal correlation characteristics between nodes in the power grid topology are determined. These spatiotemporal correlation characteristics are used to characterize the temporal series influence of historical power meteorological data of each node on the current power meteorological data value and the spatial adjacency weight between adjacent substation nodes.

[0011] Based on the power grid topology and the spatiotemporal correlation characteristics between nodes, abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation are identified.

[0012] The abnormal data is corrected based on the spatiotemporal correlation characteristics between the nodes and the preset data correction rules to obtain optimized power meteorological data, which is then used for power planning.

[0013] Secondly, the present invention provides an aberrant power meteorological data correction device for power planning, the device comprising:

[0014] The acquisition unit is used to acquire the initial power meteorological data sequence corresponding to each meteorological parameter in each substation.

[0015] The topology construction unit is used to construct a power grid topology map based on the initial power meteorological data sequence, the topology of the transmission network, the geographical relationships between substations and the admittance matrix relationships obtained by the acquisition unit. The power grid topology map uses substations as nodes and transmission lines as edges to characterize the connection relationships between nodes and their spatial distribution characteristics.

[0016] The feature determination unit is used to determine the spatiotemporal correlation features between nodes in the power grid topology diagram constructed by the topology construction unit based on a preset construction method. The spatiotemporal correlation features are used to characterize the temporal series influence strength of historical power meteorological data of each node on the current power meteorological data value and the spatial adjacency weight between adjacent substation nodes.

[0017] An anomaly identification unit is used to identify abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation based on the power grid topology map and the spatiotemporal correlation characteristics between each node determined by the feature determination unit.

[0018] An anomaly correction unit is used to correct the abnormal data identified by the anomaly identification unit according to the spatiotemporal correlation characteristics between the nodes and the preset data correction rules, so as to obtain optimized power meteorological data for power planning.

[0019] To achieve the above objectives, according to a third aspect of the present invention, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to perform the above-described method for correcting abnormal power and meteorological data for power planning.

[0020] To achieve the above objectives, according to a fourth aspect of the present invention, a processor is provided for running a program, wherein the program executes the abnormal power meteorological data correction method for power planning described in the first aspect.

[0021] By employing the above technical solution, this invention provides a method and apparatus for correcting abnormal power meteorological data in power planning. First, initial power meteorological data sequences corresponding to each meteorological parameter in each substation are obtained. Then, based on the initial power meteorological data sequences, the topology of the transmission network, the geographical relationships between substations, and the admittance matrix relationships, a power grid topology map is constructed. This topology map uses substations as nodes and transmission lines as edges to characterize the connection relationships and spatial distribution characteristics between nodes, thus comprehensively considering the spatial correlation between power grid nodes. Based on this, a spatiotemporal autoregressive model is used to analyze the spatiotemporal correlation characteristics between nodes in the power grid topology map. These spatiotemporal correlation characteristics are used to characterize the temporal series influence of historical power meteorological data on current data, as well as the spatial adjacency weights between adjacent substation nodes. Through this model, the dynamic propagation law of meteorological parameters in the power grid can be modeled from both temporal and spatial dimensions, revealing the intrinsic connections between meteorological data of each node, thereby providing theoretical support for subsequent anomaly identification and data correction. Subsequently, by combining the constructed power grid topology and the spatiotemporal correlation characteristics between nodes, abnormal data in the initial power meteorological data sequences corresponding to each meteorological parameter in each substation are identified. Compared with traditional methods that rely solely on the data change trends of individual nodes for anomaly identification, this invention further introduces the historical behavior and spatial correlation information of adjacent nodes, effectively identifying group anomalies caused by spatial propagation and significantly improving the accuracy of anomaly data identification. Finally, based on the spatiotemporal correlation characteristics between nodes and preset data correction rules, the identified abnormal data is corrected to obtain optimized power meteorological data for subsequent power operation planning. During the data correction process, adjustments are made based on spatiotemporal correlation characteristics, making the correction results more consistent with the temporal evolution and spatial distribution characteristics. This effectively handles data fluctuations caused by natural phenomena resulting from changes in surrounding sites, reduces the possibility of misjudging anomalies and incorrectly correcting normal data, thereby improving the quality of power meteorological data and enhancing the reliability and accuracy of power system operation and planning.

[0022] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0023] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0024] Figure 1 This invention provides a flowchart of a method for correcting abnormal power meteorological data for power planning, according to an embodiment of the present invention.

[0025] Figure 2 This invention provides a flowchart of another method for correcting abnormal power meteorological data for power planning.

[0026] Figure 3 This diagram illustrates a block diagram of an abnormal power meteorological data correction device for power planning, provided by an embodiment of the present invention.

[0027] Figure 4 This invention provides a block diagram of another abnormal power meteorological data correction device for power planning. Detailed Implementation

[0028] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0029] Addressing the limitations of existing methods for correcting abnormal power meteorological data, the inventors, after in-depth research, decided to move beyond traditional approaches and instead pioneer a completely new method. This method aims to enhance the reliability and accuracy of power system operation and planning by improving the quality of power meteorological data.

[0030] In power planning, optimized power meteorological data is used to predict the impact of extreme weather conditions on the power grid, such as strong winds and heavy rains that could lead to line outages. Furthermore, it can assess the impact of climate factors on electricity demand, such as trends in electricity consumption during periods of extreme heat or cold. Through comprehensive analysis of this information, power generation plans can be formulated more scientifically, maintenance work can be scheduled, and load allocation strategies can be adjusted, thereby improving the stability and efficiency of the entire power system and ensuring the safe and reliable supply of electricity.

[0031] Next, combined Figure 1 This invention provides a method for correcting abnormal power meteorological data for power planning, and its specific execution steps are as follows: Figure 1 As shown, it includes:

[0032] 101. Obtain the initial power meteorological data sequence corresponding to each meteorological parameter in each substation.

[0033] 102. Construct a power grid topology map based on the initial power meteorological data sequence, the topology of the transmission network, the geographical relationships between substations, and the admittance matrix relationships.

[0034] The power grid topology diagram uses substations as nodes and transmission lines as edges to represent the connection relationships between nodes and their spatial distribution characteristics.

[0035] In step 101, the key meteorological parameters that need to be monitored are first determined based on the actual needs of power grid operation and planning. Then, power meteorological data corresponding to these meteorological parameters are obtained from multiple sources, including meteorological departments, satellite observation systems, and ground meteorological stations. These meteorological parameters may include typical meteorological elements such as temperature, humidity, wind speed, rainfall, and air pressure.

[0036] After acquiring the raw data, it is integrated and processed according to a unified format and standard. Specifically, the collected data is first preliminarily screened to remove obviously erroneous or abnormal data, such as values ​​exceeding the physically reasonable range or trends that violate basic physical laws (e.g., temperatures below absolute zero or far above the highest possible surface temperature). Subsequently, the data is formatted and standardized, for example, temperature data from different sources are unified to degrees Celsius, and wind speed data is unified to meters per second, to ensure data consistency and usability, resulting in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation.

[0037] Subsequently, in step 102, a topology diagram of the power grid can be collected. This diagram should include the location coordinates, electrical connections, and technical parameters of each transmission line, such as length, resistance, and reactance, of all substations in the power grid (or substations whose power meteorological data needs to be analyzed in detail).

[0038] Based on the transmission line parameters and electrical connections in the aforementioned power grid topology, the admittance matrix between each substation can be further established.

[0039] The admittance matrix is ​​used to describe the electrical connection characteristics between nodes in the power grid.

[0040] At the same time, by using the location coordinates of substations in geographic information systems or power grid topology maps, the geographical relationships between substations can be analyzed. The geographical relationships between substations include geographical information such as the spatial distance and terrain features between the two stations.

[0041] Furthermore, a power grid topology map can be constructed using graphical modeling tools or programming languages ​​based on the power grid topology, the geographical relationships between substations, and the admittance matrix relationships. In this topology map, each substation is treated as a node, and transmission lines are treated as edges, representing the connection relationships between nodes and their spatial distribution characteristics.

[0042] Based on this, the initial power meteorological data sequence obtained in step 101 is combined with the power grid topology map, and corresponding meteorological parameter time series data are added to each substation node. The resulting composite data structure not only reflects the electrical connection relationship of the power grid, but also integrates meteorological environmental information at each node, providing solid data support for subsequent analysis based on the spatiotemporal autoregressive model.

[0043] In addition, each node in the constructed power grid topology map should have attribute information, including but not limited to basic information such as the geographical coordinates of the substation, its region, and installed capacity; each edge should also contain corresponding attribute information, such as electrical parameters such as the length, transmission capacity, and impedance of the transmission line, so as to achieve deep integration of power grid structure and power meteorological data.

[0044] 103. Based on the preset construction method, determine the spatiotemporal correlation characteristics between nodes in the power grid topology diagram.

[0045] 104. Based on the power grid topology and the spatiotemporal correlation characteristics between nodes, identify abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation.

[0046] 105. Based on the spatiotemporal correlation characteristics between nodes and the preset data correction rules, abnormal data are corrected to obtain optimized power meteorological data, which is then used for power planning.

[0047] Among them, the spatiotemporal correlation feature is used to characterize the temporal series influence of historical power meteorological data of each node on the current power meteorological data value and the spatial adjacency weight between adjacent substation nodes.

[0048] In this step, various methods can be used to determine the spatiotemporal correlation characteristics between nodes:

[0049] The first approach involves time-series modeling of meteorological parameters for each substation node. Specifically, an autoregressive (AR) model is used to analyze its temporal evolution characteristics. For example, an AR(p) model is established to identify the time lag order p of meteorological elements such as temperature and precipitation, thereby quantifying the influence of historical data at a certain moment on the current power meteorological data value of that node.

[0050] Simultaneously, a vector autoregressive (VAR) model is constructed, using selected adjacent substation nodes as endogenous variables to form a multi-equation system. This model can quantitatively analyze the potential joint influence of historical values ​​from 3 to 5 neighboring substations on the air pressure changes at a given station, thereby capturing the spatial interactions between nodes.

[0051] Furthermore, by combining the AR model with the VAR model, a spatiotemporal autoregressive model (STAR) is constructed. This model incorporates a spatial weight matrix and time lag terms during the modeling process, enabling a comprehensive description of the dynamic propagation of meteorological parameters within the power grid. The model parameters can be solved using maximum likelihood estimation or Bayesian methods, thereby obtaining the correlation indicators between nodes in both time and space dimensions.

[0052] The second method first establishes a time correlation model for the meteorological parameters at each substation node based on its corresponding initial power meteorological data sequence. This model characterizes the variation of meteorological data within a single node over time. Simultaneously, a spatial weight matrix is ​​constructed based on the transmission network topology, admittance matrix relationships, and geographical distance information between substations. This matrix characterizes the spatial adjacency weights between nodes, reflecting the strength of mutual influence between different stations.

[0053] Subsequently, the aforementioned time series model is combined with a spatial weight matrix to construct a spatiotemporal autoregressive model among nodes. This model not only considers the impact of historical data on the current state in the time dimension but also fully reflects the coupling relationship between adjacent nodes in the spatial dimension. Finally, the parameters of the spatiotemporal autoregressive model are solved using the maximum likelihood estimation method to obtain the autoregressive coefficients of each node in the time dimension and the adjacency weights in the spatial dimension, thereby achieving a quantitative characterization of the spatiotemporal correlation characteristics among nodes.

[0054] It should be noted that although both methods can effectively characterize the spatiotemporal correlation between power grid nodes, they differ significantly in terms of modeling approach, applicable scenarios, and interpretability.

[0055] The first method, based on autoregressive (AR) and vector autoregressive (VAR) models, focuses on avoiding explicit construction of spatial weight matrices. Instead, it treats the time series data of multiple nodes as endogenous variables of the system, capturing their dynamic relationships by estimating the regression coefficients between the variables. The spatial correlations in this method are automatically learned from the data and do not rely on prior structures such as power grid topology or geographical information. For example, in a VAR model, if the historical temperature changes at station A can significantly predict the current temperature changes at station B, then a statistically significant "influence" is considered to exist between the two. However, this influence is more of a statistical correlation between variables and does not necessarily have a clear physical adjacency meaning.

[0056] The second method constructs an explicit spatial weight matrix. This weight matrix is ​​typically built a priori based on physical information such as the power grid topology, admittance matrix, and geographical distances between substations, thus giving the spatial correlation clear engineering and geographical significance. For example, if substations A and B are geographically close, the corresponding weight wAB is large; while if A and C are far apart, wAC is close to 0, indicating a weak spatial influence between them.

[0057] In summary, the first method focuses on data-driven statistical modeling and is suitable for exploring potential interactions between variables; while the second method emphasizes the integration of structured modeling and physical mechanisms, making it more suitable for the analysis of power grid systems with clear spatial layout and electrical connections, and possessing stronger interpretability and engineering applicability.

[0058] After obtaining the spatiotemporal correlation characteristics between each node, the process proceeds to step 105, where abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation is identified based on the power grid topology map and the aforementioned spatiotemporal correlation characteristics. Subsequently, in step 106, the identified abnormal data is corrected according to the spatiotemporal correlation characteristics between each node and preset data correction rules to obtain optimized power meteorological data, thereby providing high-quality data support for subsequent power system operation analysis and power grid planning.

[0059] It should be noted that different types of abnormal data have different manifestations and causes, thus requiring differentiated identification strategies. Furthermore, this embodiment does not limit the specific method for identifying abnormal data, as long as it falls within the scope of "identifying abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation based on the power grid topology map and the spatiotemporal correlation characteristics between nodes." For example, for identifying missing data, different strategies such as direct detection or spatially assisted verification can be used.

[0060] Among them, the direct detection method refers to directly determining that if a certain meteorological parameter of a substation has no valid record or is marked as null at a specific time point in the initial power meteorological data sequence, the data is missing. The spatial auxiliary verification method analyzes the data change trends of adjacent nodes within the same time period. If the target node shows no response during this period, while its neighboring nodes show significant changes, it can help determine that the node's data is abnormally missing. By employing multiple identification methods, the comprehensiveness and accuracy of abnormal data identification can be improved from different perspectives, enhancing the system's adaptability to different types of anomalies.

[0061] Based on the above Figure 1 As can be seen from the implementation method, the present invention provides a method for correcting abnormal power meteorological data for power planning. First, initial power meteorological data sequences corresponding to each meteorological parameter in each substation are obtained. Then, based on the initial power meteorological data sequences, the topology of the transmission network, the geographical relationships between substations, and the admittance matrix relationships, a power grid topology map is constructed. This topology map uses substations as nodes and transmission lines as edges to characterize the connection relationships and spatial distribution characteristics between nodes, thus comprehensively considering the spatial correlation between power grid nodes. Based on this, a spatiotemporal autoregressive model is used to analyze the spatiotemporal correlation characteristics between nodes in the power grid topology map. These spatiotemporal correlation characteristics are used to characterize the temporal series influence of historical power meteorological data on current data, as well as the spatial adjacency weights between adjacent substation nodes. Through this model, the dynamic propagation law of meteorological parameters in the power grid can be modeled from both temporal and spatial dimensions, revealing the intrinsic connections between meteorological data of each node, thereby providing theoretical support for subsequent anomaly identification and data correction. Subsequently, by combining the constructed power grid topology and the spatiotemporal correlation characteristics between nodes, abnormal data in the initial power meteorological data sequences corresponding to each meteorological parameter in each substation are identified. Compared with traditional methods that rely solely on the data change trends of individual nodes for anomaly identification, this invention further introduces the historical behavior and spatial correlation information of adjacent nodes, effectively identifying group anomalies caused by spatial propagation and significantly improving the accuracy of anomaly data identification. Finally, based on the spatiotemporal correlation characteristics between nodes and preset data correction rules, the identified abnormal data is corrected to obtain optimized power meteorological data for subsequent power operation planning. During the data correction process, adjustments are made based on spatiotemporal correlation characteristics, making the correction results more consistent with the temporal evolution and spatial distribution characteristics. This effectively handles data fluctuations caused by natural phenomena resulting from changes in surrounding sites, reduces the possibility of misjudging anomalies and incorrectly correcting normal data, thereby improving the quality of power meteorological data and enhancing the reliability and accuracy of power system operation and planning.

[0062] Furthermore, as a response to Figure 1Further refinement and extension of the illustrated embodiment, this invention also provides another method for correcting abnormal power meteorological data for power planning, such as... Figure 2 As shown, the specific steps are as follows:

[0063] 201. Obtain the initial power meteorological data sequence corresponding to each meteorological parameter in each substation.

[0064] 202. Construct a power grid topology map based on the initial power meteorological data sequence, the topology of the transmission network, the geographical relationships between substations, and the admittance matrix relationships.

[0065] The implementation methods of steps 201-202 are the same as those of steps 101-102, and can achieve the same technical effect and solve the same technical problem, so they will not be repeated here.

[0066] 203. Based on the preset construction method, determine the spatiotemporal correlation characteristics between nodes in the power grid topology diagram.

[0067] In this embodiment, the second implementation method described in step 103 is a preferred implementation method in the present invention.

[0068] 204. Based on the power grid topology and the spatiotemporal correlation characteristics between nodes, identify abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation.

[0069] In this embodiment, the identification of outliers can be achieved using the following steps:

[0070] 1. For each node in the power grid topology diagram, for each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in that node, based on the historical power meteorological data within the preset time window before the data acquisition time, the time series prediction value at the data acquisition time is calculated using a time series model.

[0071] 2. For each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in each node, determine the initial power meteorological data values ​​of the adjacent nodes at the same data acquisition time based on the spatial weight matrix.

[0072] 3. The initial power meteorological data values ​​of adjacent nodes at the same data acquisition time are weighted and calculated to obtain the spatial weighted prediction value of the node at the data acquisition time.

[0073] 4. For each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in this node, the corresponding time series prediction value and spatial weighted prediction value are fused to obtain the corresponding spatiotemporal fusion prediction value.

[0074] 5. For each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in this node, calculate the deviation between the initial power meteorological data value and the corresponding spatiotemporal fusion prediction value.

[0075] 6. If the deviation exceeds the preset threshold, the initial power meteorological data value is determined to be an outlier in the abnormal data.

[0076] In the time dimension, this method uses a time series model to model the historical variation trends of meteorological parameters at each node, thereby predicting the theoretical values ​​at each time point. In the spatial dimension, it uses a spatial weight matrix combined with observation data from neighboring nodes to calculate the corresponding spatially weighted predicted values. By fusing the time prediction results with the spatial prediction results, a spatiotemporal fusion predicted value is obtained, which can more comprehensively and accurately reflect the variation patterns of meteorological parameters under the actual physical environment, thereby effectively improving the accuracy of anomaly identification.

[0077] This anomaly identification process not only relies on the time-series data of a single substation, but also comprehensively considers the behavioral characteristics of neighboring nodes that are related to it in terms of power grid topology and geospatial, further enhancing the robustness and reliability of the identification results.

[0078] Furthermore, for identifying missing data in abnormal data, this embodiment preferably uses a spatial auxiliary verification method. This involves comparing and analyzing the data change trends of adjacent nodes within the same time period. If a node shows no response during this period while adjacent nodes exhibit significant changes, it can help determine if the data is abnormally missing. By introducing data from adjacent nodes as a reference and performing spatial consistency checks, interference from accidental fluctuations or temporary signal loss can be effectively eliminated, improving the reliability of the identification results.

[0079] Similarly, spatial consistency testing can be used to identify abrupt changes in anomalous data. This method compares the data changes of a node with those of its neighboring nodes at the same point in time. If a target node exhibits a drastic change while its neighboring nodes do not show a similar trend at the same time, the change can be further confirmed as an anomalous behavior.

[0080] Of course, the identification of abrupt data can also be achieved by combining multiple methods to improve the comprehensiveness and accuracy of the identification. For example, spatial consistency verification can be introduced on the basis of differential detection. This means that only when both methods are used to verify the data and both determine that the data attribute has an abrupt change, will the data be considered to have an abrupt change.

[0081] Among them, the differential detection method calculates the difference (i.e., first-order difference) between two adjacent time points for a certain meteorological parameter at the same node. If the absolute value exceeds a preset threshold (such as three times the standard deviation of the historical fluctuation range), it is initially identified as a jump-type anomaly. Through this multi-method fusion identification strategy, not only can the temporal abrupt change characteristics within a single node be captured, but the spatial correlation between nodes in the power grid topology can also be used to enhance the robustness of the identification results, significantly improving the system's identification accuracy and adaptability to jump-type anomalies.

[0082] It is important to note that the anomaly identification and correction involved in this invention does not simply treat all outliers, jump values, or missing values ​​as anomalous. Instead, it involves a comprehensive analysis of the spatiotemporal evolution of the data, combined with the spatiotemporal correlation characteristics of the power grid topology and nodes, to identify those data points that truly deviate from the normal trend and possess anomalous properties.

[0083] 205. Based on the spatiotemporal correlation characteristics between each node and the preset data correction rules, abnormal data are corrected to obtain optimized power meteorological data, which is then used for power planning.

[0084] In this step, outliers in the identified abnormal data are replaced with corresponding spatiotemporal fusion prediction values ​​to obtain an optimized power meteorological data sequence.

[0085] To correct abrupt changes in identified anomalous data, the following steps can be taken:

[0086] 1. Determine the target data acquisition time corresponding to the jump data;

[0087] 2. In the initial power meteorological data sequence of the meteorological parameters corresponding to the target node where the jump data is located, extract the first initial power meteorological data sequence before the target data collection time and the second initial power meteorological data sequence after the target data collection time. The first initial power meteorological data sequence and the second initial power meteorological data sequence are normal data sequences after removing other abnormal data (that is, in the embodiment of the present invention, it is preferred to correct the jump data last).

[0088] 3. Linear regression was used to fit the first initial power meteorological data sequence and the second initial power meteorological data sequence to obtain the trend line before the jump and the trend line after the jump.

[0089] 4. Based on the trend lines before and after the jump, correct the jump data in the abnormal data to obtain optimized power meteorological data:

[0090] 4.1 Substitute the target data acquisition time into the mathematical expressions corresponding to the trend line before the jump and the trend line after the jump, respectively, to calculate the first theoretical power meteorological data value of the trend line before the jump at the target data acquisition time, and the second theoretical power meteorological data value of the trend line after the jump at the target data acquisition time.

[0091] 4.2. Based on the lengths of the first initial power meteorological data sequence and the second initial power meteorological data sequence, determine the weighting coefficients corresponding to the first theoretical power meteorological data value and the second theoretical power meteorological data value;

[0092] 4.3. Based on the weighting coefficients, the first theoretical power meteorological data value and the second theoretical power meteorological data value are weighted and fused to obtain the theoretical correction value corresponding to the target data acquisition time.

[0093] 4.4 Replace the jump data at the target data acquisition time with the theoretical correction value to obtain optimized power meteorological data.

[0094] To correct missing values ​​in abnormal data, a time series model can be used to model the historical data of nodes to predict the expected power and meteorological data values ​​at the time of missing data collection. These power and meteorological data values ​​can then be used to fill in the missing values. This approach can improve the consistency between the filling results and the overall trend, and reduce the risk of sudden changes.

[0095] 206. After replacing the jump data at the target data acquisition time with the theoretical correction value, verify the theoretical correction value.

[0096] In this embodiment, for jump data, the target neighbor node with the largest spatial adjacency weight of the target node where the jump data is located can be obtained according to the spatial weight matrix. Then, the target neighbor power meteorological data value at the target data acquisition time is determined in the initial power meteorological data sequence of the meteorological parameters corresponding to the target neighbor node.

[0097] Subsequently, it can be determined whether the difference between the theoretical correction value and the target adjacent power meteorological data value exceeds a preset threshold. If so, it indicates that the theoretical correction value may deviate from the overall meteorological change trend of the region and lacks physical rationality, requiring secondary adjustment. In this case, the product value between the difference and the preset adjustment coefficient can be determined, and the sum of the product value and the theoretical correction value can be determined as the final theoretical correction value corresponding to the target data acquisition time. Then, the jump data at the target data acquisition time is replaced with the final theoretical correction value to obtain optimized power meteorological data. Conversely, if the difference does not exceed the preset threshold, it is considered that the theoretical correction value reasonably reflects the regional meteorological characteristics, and no additional adjustment is required; the jump data can be directly replaced with this value.

[0098] The preset adjustment coefficient is determined based on the spatial adjacency weight between the target node and its adjacent nodes.

[0099] By using the above methods, not only are the historical data trends of the nodes themselves utilized, but also the observation information of their most relevant neighboring nodes in the power grid topology is combined to ensure that the correction results have good consistency and physical reliability in both time and space dimensions, thereby improving the overall quality of power meteorological data and providing reliable data support for subsequent power operation planning.

[0100] Furthermore, as a response to the above Figure 1 In addition to the implementation of the method shown, this embodiment of the invention also provides an abnormal power meteorological data correction device for power planning, used to correct the above-mentioned abnormal power meteorological data. Figure 1 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 3 As shown, the device includes:

[0101] Acquisition unit 301 is used to acquire the initial power meteorological data sequence corresponding to each meteorological parameter in each substation;

[0102] The topology construction unit 302 is used to construct a power grid topology map based on the initial power meteorological data sequence, the topology of the transmission network, the geographical relationships between substations and the admittance matrix relationships obtained by the acquisition unit 301. The power grid topology map uses substations as nodes and transmission lines as edges to characterize the connection relationships between nodes and their spatial distribution characteristics.

[0103] The feature determination unit 303 is used to determine the spatiotemporal correlation features between nodes in the power grid topology diagram constructed by the topology construction unit 302 based on a preset construction method. The spatiotemporal correlation features are used to characterize the temporal series influence strength of the historical power meteorological data of each node on the current power meteorological data value and the spatial adjacency weight between adjacent substation nodes.

[0104] Anomaly identification unit 304 is used to identify abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation based on the power grid topology map and the spatiotemporal correlation characteristics between each node determined by the feature determination unit 303.

[0105] Anomaly correction unit 305 is used to correct the abnormal data identified by anomaly identification unit 304 according to the spatiotemporal correlation characteristics between the nodes and preset data correction rules, so as to obtain optimized power meteorological data for power planning.

[0106] Furthermore, as a response to the above Figure 2 In addition to the implementation of the method shown, this embodiment of the invention also provides another abnormal power meteorological data correction device for power planning, used to correct the above-mentioned abnormal power meteorological data. Figure 2 The method shown is implemented accordingly. This device embodiment corresponds to the foregoing method embodiment. For ease of reading, this device embodiment will not repeat the details of the foregoing method embodiment, but it should be clear that the device in this embodiment can implement all the contents of the foregoing method embodiment. Figure 4 As shown, the device includes:

[0107] Acquisition unit 301 is used to acquire the initial power meteorological data sequence corresponding to each meteorological parameter in each substation;

[0108] The topology construction unit 302 is used to construct a power grid topology map based on the initial power meteorological data sequence, the topology of the transmission network, the geographical relationships between substations and the admittance matrix relationships obtained by the acquisition unit 301. The power grid topology map uses substations as nodes and transmission lines as edges to characterize the connection relationships between nodes and their spatial distribution characteristics.

[0109] The feature determination unit 303 is used to determine the spatiotemporal correlation features between nodes in the power grid topology diagram constructed by the topology construction unit 302 based on a preset construction method. The spatiotemporal correlation features are used to characterize the temporal series influence strength of the historical power meteorological data of each node on the current power meteorological data value and the spatial adjacency weight between adjacent substation nodes.

[0110] Anomaly identification unit 304 is used to identify abnormal data in the initial power meteorological data sequence corresponding to each meteorological parameter in each substation based on the power grid topology map and the spatiotemporal correlation characteristics between each node determined by the feature determination unit 303.

[0111] Anomaly correction unit 305 is used to correct the abnormal data identified by anomaly identification unit 304 according to the spatiotemporal correlation characteristics between the nodes and preset data correction rules, so as to obtain optimized power meteorological data for power planning.

[0112] In one optional implementation, the feature determination unit 303 is specifically used for:

[0113] For each meteorological parameter in each node, a time correlation model is established based on the initial power meteorological data sequence corresponding to each meteorological parameter;

[0114] Based on the topology of the power transmission network, the admittance matrix relationship, and the geographical distance information, a spatial weight matrix is ​​constructed, wherein the spatial weight matrix is ​​used to characterize the spatial adjacency weight between each node.

[0115] By combining the time series model with the spatial weight matrix, a spatiotemporal autoregressive model among the nodes is constructed.

[0116] The parameters of the spatiotemporal autoregressive model are solved using the maximum likelihood estimation method to obtain the autoregressive coefficients of each node in the time dimension and the adjacency weights in the spatial dimension, thereby quantifying the spatiotemporal correlation characteristics between each node.

[0117] In one optional implementation, the anomaly detection unit 304 is specifically used for:

[0118] For each node in the power grid topology diagram, and for each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in that node, the time series prediction value at the data acquisition time is calculated using a time series model based on the historical power meteorological data within a preset time window before the data acquisition time.

[0119] For each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in each node, the initial power meteorological data values ​​of the adjacent nodes of the node at the same data acquisition time are determined according to the spatial weight matrix.

[0120] The initial power meteorological data values ​​of the adjacent nodes at the same data acquisition time are weighted and calculated to obtain the spatial weighted prediction value of the node at the data acquisition time.

[0121] For each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in this node, the corresponding time series prediction value and spatial weighted prediction value are fused to obtain the corresponding spatiotemporal fusion prediction value.

[0122] For each data acquisition time in the initial power meteorological data sequence corresponding to each meteorological parameter in this node, calculate the deviation between the initial power meteorological data value and the corresponding spatiotemporal fusion prediction value.

[0123] If the deviation exceeds a preset threshold, the initial power meteorological data value is determined to be an outlier in the abnormal data.

[0124] In one optional implementation, the anomaly correction unit 305 is specifically used for:

[0125] For outliers in the identified abnormal data, the outliers are replaced with the corresponding spatiotemporal fusion prediction values ​​to obtain an optimized power meteorological data sequence.

[0126] In another optional implementation, the anomaly correction unit 305 is specifically used for:

[0127] For the abrupt changes in the identified abnormal data, determine the target data acquisition time corresponding to the abrupt change.

[0128] In the initial power meteorological data sequence of the meteorological parameters corresponding to the target node where the jump data is located, the first initial power meteorological data sequence before the target data acquisition time and the second initial power meteorological data sequence after the target data acquisition time are extracted, wherein the first initial power meteorological data sequence and the second initial power meteorological data sequence are normal data sequences after removing other abnormal data.

[0129] Linear regression was used to fit the first initial power meteorological data sequence and the second initial power meteorological data sequence to obtain the trend line before the jump and the trend line after the jump.

[0130] Based on the trend line before the jump and the trend line after the jump, the jump data in the abnormal data is corrected to obtain optimized power meteorological data.

[0131] In another optional implementation, when the anomaly correction unit 305 corrects the jump data in the anomaly data according to the trend line before the jump and the trend line after the jump to obtain optimized power meteorological data, it is specifically used for:

[0132] Substitute the target data acquisition time into the mathematical expressions corresponding to the trend line before the jump and the trend line after the jump, respectively, to calculate the first theoretical power meteorological data value of the trend line before the jump at the target data acquisition time, and the second theoretical power meteorological data value of the trend line after the jump at the target data acquisition time.

[0133] Based on the lengths of the first initial power meteorological data sequence and the second initial power meteorological data sequence, the weighting coefficients corresponding to the first theoretical power meteorological data value and the second theoretical power meteorological data value are determined.

[0134] Based on the weighting coefficient, the first theoretical power meteorological data value and the second theoretical power meteorological data value are weighted and fused to obtain the theoretical correction value corresponding to the target data collection time.

[0135] Replace the jump data at the target data acquisition time with the theoretical correction value to obtain optimized power meteorological data.

[0136] In an optional implementation, after the anomaly correction unit 305 replaces the jump data at the target data acquisition time according to the theoretical correction value, the device further includes a theoretical value verification unit 306, which is specifically used for:

[0137] Based on the spatial weight matrix, obtain the target neighbor node with the largest spatial adjacency weight of the target node where the jump data is located;

[0138] In the initial power meteorological data sequence corresponding to the meteorological parameters of the target neighboring nodes, the target neighboring power meteorological data values ​​at the target data acquisition time are determined;

[0139] Determine whether the difference between the theoretical correction value and the target adjacent power meteorological data value exceeds a preset threshold;

[0140] If so, the product of the difference and the preset adjustment coefficient is determined, and the sum of the product and the theoretical correction value is determined as the final theoretical correction value corresponding to the target data acquisition time. The preset adjustment coefficient is determined based on the spatial adjacency weight between the target node and the target adjacent node.

[0141] Replace the jump data at the target data acquisition time with the final theoretical correction value to obtain optimized power meteorological data.

[0142] Furthermore, embodiments of the present invention also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1-2 The method for correcting abnormal power meteorological data for power planning described in [the document].

[0143] Furthermore, embodiments of the present invention also provide a processor for running a program, wherein the program executes the above-described... Figure 1-2 The method for correcting abnormal power meteorological data for power planning described in [the document].

[0144] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0145] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.

[0146] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0147] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0148] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0149] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0150] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0151] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0152] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0153] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0154] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0155] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0156] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0157] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0158] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An abnormal power weather data correction method for power planning, characterized by, The method comprises: obtaining initial power meteorological data sequences corresponding to each meteorological parameter in each transformer substation; constructing a power grid topology graph according to the initial power meteorological data sequences, a topology structure of a power transmission network, geographical relationships between the transformer substations, and admittance matrix relationships, wherein the power grid topology graph takes the transformer substations as nodes and power transmission lines as edges, and is used to represent connection relationships between the nodes and spatial distribution characteristics thereof; determining spatiotemporal correlation characteristics between the nodes in the power grid topology graph based on a preset construction method, wherein the spatiotemporal correlation characteristics are used to represent time series influence intensities of historical power meteorological data of the nodes on current power meteorological data values and spatial adjacency weights between adjacent transformer substation nodes; identifying abnormal data in the initial power meteorological data sequences of the transformer substations based on the power grid topology graph and the spatiotemporal correlation characteristics between the nodes; correcting the abnormal data according to the spatiotemporal correlation characteristics between the nodes and a preset data correction rule to obtain optimized power meteorological data, so that power planning is performed according to the optimized power meteorological data; identifying abnormal data in the initial power meteorological data sequences of the transformer substations based on the power grid topology graph and the spatiotemporal correlation characteristics between the nodes, comprising: for each node in the power grid topology graph, for each data collection time in initial power meteorological data sequences corresponding to each meteorological parameter in the node, calculating a time series prediction value of the data collection time based on historical power meteorological data in a preset time window before the data collection time by using a time series model, wherein the time series model is established according to the corresponding initial power meteorological data sequence; for each data collection time in the initial power meteorological data sequences corresponding to each meteorological parameter in each node, determining initial power meteorological data values of adjacent nodes of the node at the same data collection time according to a spatial weight matrix, wherein the spatial weight matrix is constructed based on a topology structure of a power transmission network, admittance matrix relationships, and geographical distance information between the transformer substations; performing weighted calculation on the initial power meteorological data values of the adjacent nodes at the same data collection time to obtain a spatial weighted prediction value of the node at the data collection time; for each data collection time in the initial power meteorological data sequences corresponding to each meteorological parameter in the node, fusing the corresponding time series prediction value and the spatial weighted prediction value to obtain a corresponding spatiotemporal fusion prediction value; for each data collection time in the initial power meteorological data sequences corresponding to each meteorological parameter in the node, calculating a deviation between the initial power meteorological data value and the corresponding spatiotemporal fusion prediction value; if the deviation exceeds a preset threshold, determining that the initial power meteorological data value is an outlier in the abnormal data.

2. The method of claim 1, wherein, determining the spatiotemporal correlation characteristics between the nodes in the power grid topology graph based on a preset construction method, comprising: for each meteorological parameter in each node, establishing a time series model according to an initial power meteorological data sequence corresponding to the meteorological parameter; According to the topological structure of the power transmission network, the admittance matrix relationship and the geographical distance information, a spatial weight matrix is constructed, wherein the spatial weight matrix is used to represent the spatial adjacency weight between nodes; The time series model is combined with the spatial weight matrix to construct a spatio-temporal autoregressive model between nodes; The parameters of the spatio-temporal autoregressive model are solved by using the maximum likelihood estimation method, and the autoregressive coefficients of each node in the time dimension and the adjacency weight in the space dimension are obtained, thereby quantifying the spatio-temporal correlation characteristics between nodes.

3. The method of claim 1, wherein, According to the spatio-temporal correlation characteristics between nodes and a preset data correction rule, the abnormal data is corrected to obtain optimized power meteorological data, including: For the outlier in the identified abnormal data, the outlier is replaced by the corresponding spatio-temporal fusion prediction value to obtain an optimized power meteorological data sequence.

4. The method of claim 1, wherein, According to the spatio-temporal correlation characteristics between nodes and a preset data correction rule, the abnormal data is corrected to obtain optimized power meteorological data, including: For the jump data in the identified abnormal data, a target data acquisition time corresponding to the jump data is determined; In the initial power meteorological data sequence of the corresponding meteorological parameter of the target node where the jump data is located, a first initial power meteorological data sequence before the target data acquisition time and a second initial power meteorological data sequence after the target data acquisition time are extracted, wherein the first initial power meteorological data sequence and the second initial power meteorological data sequence are both normal data sequences after excluding other abnormal data; A linear regression method is used to fit the first initial power meteorological data sequence and the second initial power meteorological data sequence respectively to obtain a pre-jump trend line and a post-jump trend line; According to the pre-jump trend line and the post-jump trend line, the jump data in the abnormal data is corrected to obtain optimized power meteorological data.

5. The method of claim 4, wherein, According to the pre-jump trend line and the post-jump trend line, the jump data in the abnormal data is corrected to obtain optimized power meteorological data, including: The target data acquisition time is substituted into the mathematical expressions corresponding to the pre-jump trend line and the post-jump trend line respectively to calculate a first theoretical power meteorological data value of the pre-jump trend line at the target data acquisition time and a second theoretical power meteorological data value of the post-jump trend line at the target data acquisition time; Based on the lengths of the first initial power meteorological data sequence and the second initial power meteorological data sequence, the weight coefficients corresponding to the first theoretical power meteorological data value and the second theoretical power meteorological data value are determined; According to the weight coefficients, the first theoretical power meteorological data value and the second theoretical power meteorological data value are weighted and fused to obtain a theoretical correction value corresponding to the target data acquisition time; According to the theoretical correction value, the jump data at the target data acquisition time is replaced to obtain optimized power meteorological data.

6. The method of claim 5, wherein, After replacing the jump data at the target data acquisition time according to the theoretical correction value, the method further includes: According to the spatial weight matrix, a target adjacent node with a maximum spatial adjacency weight of a target node where the jump data is located is obtained; An adjacent power meteorological data value of the target adjacent node at the target data acquisition moment in an initial power meteorological data sequence of a meteorological parameter corresponding to the target adjacent node is determined; It is judged whether a difference between the theoretical correction value and the adjacent power meteorological data value exceeds a preset threshold value; If yes, a product value between the difference and a preset adjustment coefficient is determined, and a sum value between the product value and the theoretical correction value is determined as a final theoretical correction value corresponding to the target data acquisition moment, wherein the preset adjustment coefficient is determined according to a spatial adjacency weight between the target node and the target adjacent node; The jump data at the target data acquisition moment is replaced according to the final theoretical correction value, so as to obtain optimized power meteorological data.

7. An abnormal power weather data correction device for power planning, characterized by, The device comprises: An acquisition unit is configured to acquire initial power meteorological data sequences corresponding to meteorological parameters in each transformer substation; A topology construction unit is configured to construct a power grid topology graph according to the initial power meteorological data sequences acquired by the acquisition unit, a topology structure of a power transmission network, geographical relationships between the transformer substations, and admittance matrix relationships, wherein the power grid topology graph takes the transformer substations as nodes and power transmission lines as edges, and is used to represent connection relationships between the nodes and spatial distribution characteristics thereof; A feature determination unit is configured to determine spatiotemporal correlation features between the nodes in the power grid topology graph constructed by the topology construction unit based on a preset construction method, wherein the spatiotemporal correlation features are used to represent time sequence influence intensities of historical power meteorological data of the nodes on current power meteorological data values and spatial adjacency weights between adjacent transformer substation nodes; An anomaly identification unit is configured to identify abnormal data in the initial power meteorological data sequences corresponding to the meteorological parameters in each transformer substation based on the power grid topology graph and the spatiotemporal correlation features between the nodes determined by the feature determination unit; An anomaly correction unit is configured to correct the abnormal data identified by the anomaly identification unit according to the spatiotemporal correlation features between the nodes and a preset data correction rule, to obtain optimized power meteorological data, and to perform power planning according to the optimized power meteorological data; The anomaly identification unit is specifically configured to: For each node in the power grid topology graph, for each data acquisition moment in initial power meteorological data sequences corresponding to meteorological parameters in the node, a time sequence prediction value at the data acquisition moment is calculated based on historical power meteorological data in a preset time window before the data acquisition moment, wherein the time sequence prediction value is calculated by using a time sequence model established according to the corresponding initial power meteorological data sequence; For each data acquisition moment in the initial power meteorological data sequences corresponding to the meteorological parameters in each node, an initial power meteorological data value of a neighboring node of the node at the same data acquisition moment is determined according to a spatial weight matrix, wherein the spatial weight matrix is constructed based on the topology structure of the power transmission network, the admittance matrix relationships, and geographical distance information between the transformer substations. The initial power meteorological data values of the adjacent nodes at the same data collection time are weighted to obtain a spatial weighted prediction value of the node at the data collection time; For each data collection time in the initial power meteorological data sequence corresponding to each meteorological parameter in the node, the corresponding time series prediction value and the spatial weighted prediction value are fused to obtain a corresponding spatio-temporal fusion prediction value; For each data collection time in the initial power meteorological data sequence corresponding to each meteorological parameter in the node, the deviation between the initial power meteorological data value and the corresponding spatio-temporal fusion prediction value is calculated. If the deviation exceeds a preset threshold, the initial power meteorological data value is determined as an outlier in the abnormal data.

8. A storage medium, characterized by The storage medium includes a stored program, wherein the program controls the device in which the storage medium is located to execute the abnormal power meteorological data correction method for power planning according to any one of claims 1 to 6 when the program is running.

9. A processor, comprising: The processor is configured to execute the program, wherein the program executes the abnormal power meteorological data correction method for power planning according to any one of claims 1 to 6 when the program is running.

Images