CN-122026363-A - Power distribution network regulation and control model construction method and device, electronic equipment and storage medium

CN122026363ACN 122026363 ACN122026363 ACN 122026363ACN-122026363-A

Abstract

The invention relates to the technical field of power control, and provides a power distribution network regulation and control model construction method, a device, electronic equipment and a storage medium. The method comprises the steps of determining change characteristic data of each real topology change based on historical topology change data and historical regulation and control data of each real topology change in a power distribution network, adjusting historical reason change data corresponding to each change reason label based on the historical topology change data of each real topology change, constructing each regulation and control sub-model based on the historical reason change data corresponding to each change reason label after adjustment, the change value of each real topology change and the change threshold value of each change reason label, and determining the power distribution network regulation and control model based on each regulation and control sub-model. The embodiment of the invention can improve the generalization capability and the regulation precision of the model to changeable topological scenes, effectively quantize the operation risks of different induced topological changes, ensure the operation stability of the power distribution network and provide reliable support for real-time accurate regulation.

Inventors

JIN MINHAO
WANG SHAOHUA
WANG LIZHONG
LUO MINGLIANG
FANG XIANG
HUANG JIABIN
TU YONGWEI
XUAN YI
WANG YI
CHEN YIFANG
WANG QIFENG
CHENG WEIDONG
SHAO PENGCHENG
LI NAN
XIE JUN
Xing jiali
LI BIN
WANG JUNJIAN
LIU KEPING
MA JINGYI
YANG MING
CHEN XIAORUI

Assignees

国网浙江省电力有限公司杭州供电公司
国网浙江省电力有限公司杭州市临安区供电公司

Dates

Publication Date: 20260512
Application Date: 20260105

Claims (10)

1. The utility model provides a power distribution network regulation and control model construction method which is characterized by comprising the following steps: determining change characteristic data of each real topology change based on historical topology change data and historical regulation data of each real topology change in the power distribution network; based on the historical topology change data of each real topology change, adjusting the historical reason change data corresponding to each change reason label; calculating the change value of each real topology change in the history reason change data corresponding to the adjustment of each change reason label based on the history reason change data corresponding to the adjustment of each change reason label and the change characteristic data of each real topology change, and calculating the change threshold value of each change reason label; And constructing each regulation sub-model based on the history reason change data, the change value of each real topological change and the change threshold value of each change reason label after corresponding adjustment of each change reason label, and determining a power distribution network regulation model based on each regulation sub-model.
2. The method of claim 1, wherein the history topology change data of the real topology change comprises a topology change label, an initial electrical topology, a reconstructed electrical topology, a set of function level labels of each branch of the reconstructed electrical topology, a function level label of each branch of the reconstructed electrical topology, a change reason label, source load data before topology switching, source load data after topology switching, the topology change label comprises stable and unstable, and the history regulation data of the real topology change comprises a history regulation strategy and a regulation result label.
3. The method of claim 1, wherein determining change characteristic data for each real topology change based on historical topology change data for each real topology change in the power distribution network and historical regulatory data comprises: Comparing an initial electrical topology structure and a reconstructed electrical topology structure in historical topology change data of each real topology change in the power distribution network, and determining a stable topology substructure branch set and a change topology substructure branch set of each real topology change of the power distribution network, wherein the stable topology substructure branch set and the change topology substructure branch set respectively comprise multiple branches; Determining source load change data of each real topology change of the power distribution network based on source load data before topology switching and source load data after topology switching in historical topology change data of each real topology change, wherein the source load change data comprises a plurality of source load characteristics, source load change values of each source load characteristic and source load labels, and the source load labels comprise power supply characteristics and load characteristics; Inputting a change topology sub-structure branch set of each real topology change and the source charge change data into a message transmission neural network, and determining topology-source charge state data of each real topology change, wherein the topology-source charge state data comprises a plurality of topology-source charge indexes, index state values of each topology-source charge index and index weights; And determining change characteristic data of each real topological change based on the stable topological substructure branch set, the change topological substructure branch set, the source charge change data and the topology-source charge state data of each real topological change.
4. The method according to claim 1, wherein the adjusting the historical cause change data corresponding to each change cause tag based on the historical topology change data of each real topology change includes: classifying historical topology change data of each real topology change based on each change reason label to obtain historical reason change data corresponding to each change reason label; And combining the historical topology change data of each simulated topology change, the change value of each simulated topology change, the topology change label and the historical regulation data to the historical reason change data corresponding to each change reason label under the condition that the number of the historical reason change data corresponding to each change reason label is lower than a preset number threshold value so as to obtain the historical reason change data corresponding to each change reason label after adjustment, wherein the historical topology change data of the simulated topology change comprises the topology change label, a generation lot number, a stable topology substructure branch set, a change topology substructure branch set, source charge change data and topology-source charge state data.
5. The method according to claim 1, wherein the calculating, based on the adjusted historical cause change data corresponding to each change cause tag and the change feature data of each real topology change, a change value of each real topology change in the adjusted historical cause change data corresponding to each change cause tag, and calculating a change threshold of each change cause tag includes: Calculating the change value of each real topology change in the history reason change data after corresponding adjustment of each change reason label based on the stable topology substructure, the change topology substructure, the source charge change data and the topology-source charge state data in the change characteristic data of each real topology change; and calculating a change threshold value of each change reason label based on each topology change label, each change value of the simulated topology change or each change value of the real topology change in the history reason change data after the corresponding adjustment of each change reason label.
6. The method of claim 1, wherein the constructing each regulation sub-model based on the respective adjusted historical cause change data for each of the change cause tags, the change value for each of the real topology changes, and the change threshold for each of the change cause tags, and determining a distribution network regulation model based on each of the regulation sub-models, comprises: dividing the history reason change data corresponding to and adjusted by each change reason label based on each topology change time stamp or each generation lot number to obtain history reason training data and history reason test data corresponding to each change reason label; Based on topology change labels in the historical cause training data corresponding to the change reason labels, extracting historical regulation strategies in the historical regulation data of each real topology change and each simulated topology change to obtain a stable regulation strategy set and an unstable regulation strategy set of each change reason label; The stable topology sub-structure branch set, the variable topology sub-structure branch set, the source charge change data, the topology-source charge state data, the change value and the source charge change data, the topology-source charge state data, the stable topology sub-structure branch set, the variable topology sub-structure branch set and the change value in the change characteristic data of each real topology change in the history reason training data corresponding to each change reason label are used as the input of each regulation sub-model, and each real topology change in the history reason training data corresponding to each change reason label and the history regulation data of each simulation topology change are used as the output of each regulation sub-model, so that each regulation sub-model is constructed, wherein one change reason label corresponds to one regulation sub-model; and optimizing each regulation sub-model based on historical reason test data corresponding to each change reason label to obtain the power distribution network regulation model.
7. The method according to claim 6, wherein optimizing each of the regulation sub-models based on the historical reason test data corresponding to each of the change reason tags to obtain the distribution network regulation model comprises: Inputting a stable topological sub-structure branch set, a variable topological sub-structure branch set, source charge change data, topology-source charge state data and a change value of each real topological change or simulated topological change in historical reason test data corresponding to each change reason label into a corresponding regulation sub-model, and judging the change value and the change threshold of each real topological change or simulated topological change through the regulation sub-model to obtain a prediction change label of each real topological change or simulated topological change; If the prediction change label is stable, selecting a prediction regulation strategy of each real topology change or simulated topology change from the stable regulation strategy subset based on a stable topology substructure branch set, a change topology substructure branch set, source charge change data and topology-source charge state data of each real topology change or simulated topology change; if the predictive change label is unstable, selecting a predictive control strategy of each real topology change or simulated topology change from the unstable control strategy subset based on a stable topology substructure branch set, a variable topology substructure branch set, source charge change data and topology-source charge state data of each real topology change or simulated topology change; The method comprises the steps that a prediction regulation strategy and a preset prediction effect label of each real topology change or simulated topology change in historical reason test data corresponding to each change reason label are respectively compared with the historical regulation strategy and the regulation result label; and optimizing each regulation sub-model according to the comparison result, and determining the power distribution network regulation model based on each optimized regulation sub-model.
8. The utility model provides a distribution network regulation and control model construction device which characterized in that includes: The data determining module is used for determining change characteristic data of each real topology change based on historical topology change data and historical regulation and control data of each real topology change in the power distribution network; The adjustment module is used for adjusting the historical reason change data corresponding to each change reason label based on the historical topology change data of each real topology change; The calculation module is used for calculating the change value of each real topological change in the history reason change data after corresponding adjustment of each change reason label based on the history reason change data after corresponding adjustment of each change reason label and the change characteristic data of each real topological change, and calculating the change threshold value of each change reason label; The model construction module is used for constructing each regulation sub-model based on the historical reason change data, the change value of each real topological change and the change threshold value of each change reason label after corresponding adjustment of each change reason label, and determining a power distribution network regulation model based on each regulation sub-model.
9. An electronic device, comprising: And a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.
10. A non-transitory computer readable storage medium storing computer instructions for causing a computer to perform the method of any one of claims 1-7.

Description

Power distribution network regulation and control model construction method and device, electronic equipment and storage medium Technical Field The present invention relates to the field of power control technologies, and in particular, to a method and an apparatus for constructing a power distribution network regulation model, an electronic device, and a storage medium. Background In the operation and regulation process of the existing power distribution network, the topology structure of the power distribution network presents variability and complexity along with the large-scale access of the distributed power supply, frequent change of a load structure and continuous adjustment of an operation mode. Aiming at topology changes occurring in different operation scenes, the existing regulation and control method depends on fixed rules or regulation and control models constructed based on single scene samples. It is difficult to adequately characterize the differences in structural adjustment amplitude, source-to-charge response characteristics, and operational risk caused by different topological changes. In addition, when the topology change type of the power distribution network is switched or a new scene with insufficient coverage appears, the existing model is often insufficient in generalization capability, and the suitability and accuracy of a regulation strategy are easily reduced. Disclosure of Invention The invention provides a power distribution network regulation and control model construction method, a device, electronic equipment and a storage medium, which can solve at least one technical problem. In a first aspect, an embodiment of the present invention provides a method for constructing a regulation model of a power distribution network, including: determining change characteristic data of each real topology change based on historical topology change data and historical regulation data of each real topology change in the power distribution network; based on the historical topology change data of each real topology change, adjusting the historical reason change data corresponding to each change reason label; calculating the change value of each real topology change in the history reason change data corresponding to the adjustment of each change reason label based on the history reason change data corresponding to the adjustment of each change reason label and the change characteristic data of each real topology change, and calculating the change threshold value of each change reason label; And constructing each regulation sub-model based on the history reason change data, the change value of each real topological change and the change threshold value of each change reason label after corresponding adjustment of each change reason label, and determining a power distribution network regulation model based on each regulation sub-model. In a second aspect, an embodiment of the present invention provides a power distribution network regulation model construction apparatus, including: The data determining module is used for determining change characteristic data of each real topology change based on historical topology change data and historical regulation and control data of each real topology change in the power distribution network; The adjustment module is used for adjusting the historical reason change data corresponding to each change reason label based on the historical topology change data of each real topology change; The calculation module is used for calculating the change value of each real topological change in the history reason change data after corresponding adjustment of each change reason label based on the history reason change data after corresponding adjustment of each change reason label and the change characteristic data of each real topological change, and calculating the change threshold value of each change reason label; The model construction module is used for constructing each regulation sub-model based on the historical reason change data, the change value of each real topological change and the change threshold value of each change reason label after corresponding adjustment of each change reason label, and determining a power distribution network regulation model based on each regulation sub-model. In a third aspect, an embodiment of the present invention further provides an electronic device, including at least one processor, and a memory communicatively coupled to the at least one processor, where the memory stores instructions executable by the at least one processor, the instructions being executable by the at least one processor to enable the at least one processor to perform the method according to any one of the embodiments of the present invention. In a fourth aspect, embodiments of the present invention also provide a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method of any one of the e