CN110084448A

CN110084448A - A kind of analysis method of electric system, device and its equipment

Info

Publication number: CN110084448A
Application number: CN201810067602.7A
Authority: CN
Inventors: 梁新乐; 李锐; 邱剑; 黄振; 吴云崇; 杨熙亮; 杨斌; 闵万里
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Group Holding Ltd
Priority date: 2018-01-24
Filing date: 2018-01-24
Publication date: 2019-08-02

Abstract

The application provides analysis method, device and its equipment of a kind of electric system, this method comprises: obtaining the first power project；The first grid nodes layout is obtained according to first power project；According to the label value of first grid nodes layout and the first grid nodes layout, the corresponding relationship of training power grid feature and characteristic value；Wherein, the corresponding relationship is used to analyze the state of electric system.By the technical solution of the application, quick Contingency Analysis of Power Systems can be carried out, supports power planning and scheduling.The time that safety analysis can be saved saves the computing resource of safety analysis.

Description

Analysis method, device and equipment of power system

Technical Field

The present application relates to the field of power, and in particular, to an analysis method, an analysis device, and an analysis apparatus for a power system.

Background

In a power system, static safety analysis is generally required to evaluate the safety and stability of the power system. The N-1 safety criterion is an important means of static safety analysis, and the realization principle is as follows: when any node in the N nodes of the power system is disconnected due to faults, the power system can stably operate and normally supply power.

In order to implement static security analysis by using the N-1 security criterion, each node of the power system may be sequentially disconnected by using a brute force enumeration method, and the power system may be subjected to security analysis, which may consume a lot of time and computational resources in the whole process. For example, when 3000 nodes exist in the power system, each of the 3000 nodes needs to be disconnected in turn, and the power system needs to be analyzed for safety, that is, at least 3000 analysis processes are required, which consumes a lot of time and computing resources, for example, at least 70 minutes may be required.

Disclosure of Invention

The application provides an analysis method of a power system, which comprises the following steps:

acquiring a first power plan;

acquiring a first power grid node layout according to the first power plan;

training the corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout and the label values of the first power grid node layout; wherein the corresponding relation is used for analyzing the state of the power system.

acquiring a second power plan of a state needing to be analyzed;

obtaining a third power grid node layout according to the second power plan;

inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values;

and analyzing the state of the second power plan according to the obtained characteristic value.

acquiring a third power plan of a state to be analyzed;

querying whether a normal power plan matched with the third power plan exists in a normal sample library;

if yes, analyzing that the state of the third power plan is normal;

if not, analyzing that the state of the third power plan is abnormal.

The application provides an analytical equipment of electric power system, the device includes:

an acquisition module for acquiring a first power plan; acquiring a first power grid node layout according to the first power plan;

the training module is used for training the corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout and the label values of the first power grid node layout; the corresponding relation is used for analyzing the state of the power system.

the acquisition module is used for acquiring a second power plan of a state needing to be analyzed; obtaining a third power grid node layout according to the second power plan; inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values;

and the analysis module is used for analyzing the state of the second power plan according to the obtained characteristic value.

the acquisition module is used for acquiring a third power plan of a state needing to be analyzed;

the query module is used for querying whether a normal power plan matched with the third power plan exists in a normal sample library or not;

the analysis module is used for analyzing the state of the third power plan to be normal when the query result is yes; and when the query result is negative, analyzing that the state of the third power plan is abnormal.

The present application provides a power management apparatus, the power management apparatus including:

a processor for obtaining a first power plan; acquiring a first power grid node layout according to the first power plan; training a corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout and the label values of the first power grid node layout; wherein the corresponding relation is used for analyzing the state of the power system.

The present application provides a power management apparatus, the power management apparatus including: a processor for obtaining a second power plan requiring analysis of a state; obtaining a third power grid node layout according to the second power plan; inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values; and analyzing the state of the second power plan according to the obtained characteristic value.

a processor for obtaining a third power plan requiring analysis of a state; querying whether a normal power plan matched with the third power plan exists in a normal sample library; if yes, analyzing that the state of the third power plan is normal; if not, analyzing that the state of the third power plan is abnormal.

Based on the technical scheme, in the embodiment of the application, the corresponding relation between the power grid characteristics and the characteristic values can be trained, the state of the power system is analyzed according to the corresponding relation, the power grid planning, the operation scheduling and the maintenance power failure plan are guided based on the analysis result, the static safety analysis of the power system can be rapidly carried out, and the power planning and the scheduling are supported. Each node of the power system is sequentially disconnected without adopting a violent enumeration mode, so that the time of safety analysis can be saved, and the computing resources of the safety analysis can be saved. Safety analysis can be carried out in a short time, so that power operators can make scheduling decisions quickly, and large-scale power faults are avoided.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments of the present application or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art according to the drawings of the embodiments of the present application.

FIGS. 1A and 1B are schematic diagrams of an application scenario in an embodiment of the present application;

FIG. 2A is a flow chart of a method of analyzing a power system in one embodiment of the present application;

FIG. 2B is a schematic diagram of a power grid topology in one embodiment of the present application;

FIG. 2C is a schematic diagram of a first intermediate node layout in one embodiment of the present application;

FIG. 2D is a flow chart of a method of analyzing a power system in one embodiment of the present application;

FIG. 2E is a flow chart of a method of analyzing a power system in another embodiment of the present application;

FIG. 3 is a flow chart of a method of analyzing a power system in another embodiment of the present application;

fig. 4 is a block diagram of an analysis device of a power system according to an embodiment of the present application;

fig. 5 is a block diagram of an analysis device of a power system according to another embodiment of the present application;

fig. 6 is a block diagram of an analysis device of a power system according to another embodiment of the present application.

Detailed Description

The terminology used in the embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein is meant to encompass any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in the embodiments of the present application to describe various information, the information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present application. Depending on the context, moreover, the word "if" as used may be interpreted as "at … …" or "when … …" or "in response to a determination".

The embodiment of the application provides an analysis method of a power system, which can be applied to a power management device used for analyzing the state (for example, safety) of the power system, such as static safety analysis. The power management device may be a PC (personal computer), a notebook computer, a mobile terminal, a server, or the like, and the type of the power management device is not limited in detail.

In an example, the corresponding relationship between the grid characteristic and the characteristic value may be trained, for example, the corresponding relationship between the grid characteristic and the characteristic value may be trained through a machine learning network (e.g., a Neural network such as CNN (Convolutional Neural network), LSTM (Long Short-term memory, Long-Short memory iterative Neural network), and the like). Then, the state of the power system can be analyzed by using the corresponding relation, so that the state analysis of the power system is realized, each node of the power system does not need to be disconnected in sequence in a violent enumeration mode, the time of safety analysis can be saved, and the computing resources of the safety analysis can be saved.

In order to realize the state analysis of the power system, a training phase and an analysis phase can be involved, and in the training phase, the corresponding relation between the power grid characteristics and the characteristic values can be trained. In the analysis stage, the state of the power system can be analyzed based on the corresponding relationship obtained in the training stage. The following describes the processing procedure of the training phase and the processing procedure of the analysis phase with reference to specific application scenarios.

Referring to fig. 1A, the flow shown by the dotted line is the processing in the training phase, and the flow shown by the solid line is the processing in the analysis phase. For the sake of convenience of distinction, the power plan of the training phase is referred to as a first power plan, and the grid node layout of the training phase is referred to as a first grid node layout. The power plan of the analysis phase is referred to as a second power plan, and the grid node layout of the analysis phase is referred to as a third grid node layout. In practical applications, the grid node topology may include, but is not limited to, a grid thermodynamic diagram.

Referring to fig. 1A, in a training phase, a first power plan may be obtained, a first power grid node layout is obtained according to the first power plan, the first power grid node layout is output to a machine learning network, and the machine learning network trains a corresponding relationship between a power grid characteristic and a characteristic value according to the first power grid node layout.

In the analysis stage, a second power plan may be obtained, a third power grid node layout may be obtained according to the second power plan, and the third power grid node layout may be output to the machine learning network, and the machine learning network obtains a characteristic value according to the third power grid node layout and the correspondence, and outputs the obtained characteristic value.

When the machine learning network trains the corresponding relation between the power grid characteristics and the characteristic values according to the power grid node layout, the accuracy of the training result is determined by the number of the power grid node layouts, and therefore a large number of power grid node layouts can be provided. However, in practical applications, the number of abnormal first power plans may be relatively small, and the number of abnormal first grid node layouts is relatively small, so the processing flow shown in fig. 1B may also be adopted.

As shown in fig. 1B, the flow shown by the dotted line is the processing in the training phase, and the flow shown by the solid line is the processing in the analysis phase. For convenience of distinguishing, the power plan in the training stage is called a first power plan, the power grid node layout obtained based on the first power plan is called a first power grid node layout, and the power grid node layout obtained based on the first power grid node layout is called a second power grid node layout. The power plan of the analysis phase is referred to as a second power plan and the grid node topology of the analysis phase is referred to as a third grid node topology.

Referring to fig. 1B, in the training phase, a first power plan may be obtained, a first power grid node layout may be obtained according to the first power plan, and the first power grid node layout may be output to the machine learning network. In addition, the first grid node layout may be output to the generative countermeasure network, and a second grid node layout (e.g., a plurality of second grid node layouts) may be generated by the generative countermeasure network from the first grid node layout and output to the machine learning network. After the processing, the machine learning network may receive the first power grid node layout and the second power grid node layout, and train a corresponding relationship between the power grid characteristic and the characteristic value according to the received first power grid node layout and the second power grid node layout.

In one example, when the machine learning network trains the correspondence between the grid characteristics and the characteristic values, the machine learning network may use a supervised machine learning algorithm (such as a deep learning algorithm) or an unsupervised machine learning algorithm to train the correspondence between the grid characteristics and the characteristic values. Wherein, the supervised machine learning algorithm is as follows: in the training process, corresponding label data (the label data is used for indicating whether the first power grid node layout and the second power grid node layout are normal power grid node layouts or abnormal power grid node layouts, and in the subsequent process, the label data is called as a label value of the power grid node layouts) are provided for the input first power grid node layout and second power grid node layout. The unsupervised machine learning algorithm refers to: in the training process, corresponding marking data do not exist for the input first power grid node layout and the input second power grid node layout. For convenience of description, the following will describe a training process of the machine learning network by taking a supervised machine learning algorithm as an example.

Referring to fig. 2A, which is a flowchart of an analysis method of a power system, fig. 2A is a flowchart of a training phase, and fig. 2A is a flowchart based on the application scenario shown in fig. 1A, the method may include:

in step 211, a first power plan is obtained.

Step 212, a first grid node topology is obtained according to the first power plan.

Step 213, training a corresponding relation between the grid characteristics and the characteristic values according to the first grid node layout and the label values of the first grid node layout; wherein, the corresponding relation is used for analyzing the state of the power system.

In an example, the execution sequence is only an example given for convenience of description, and in practical applications, the execution sequence between steps may also be changed, and the execution sequence is not limited. Moreover, in other embodiments, the steps of the respective methods do not have to be performed in the order shown and described herein, and the methods may include more or less steps than those described herein. Moreover, a single step described in this specification may be broken down into multiple steps for description in other embodiments; multiple steps described in this specification may be combined into a single step in other embodiments.

With respect to step 211, in one example, the power management apparatus may maintain a historical database for storing power plans, and the power plans stored in the historical database are referred to as a first power plan. Based on this, the power management apparatus may acquire the first power plan from the history database.

In one example, each time a new power plan is obtained by the power management device, the new power plan may be stored in the historical database, such that the power plan in the historical database is continually updated to accommodate changes in the external environment. On this basis, the power management apparatus may execute the above steps 211 to 213 when the power schedule in the history database is changed; alternatively, the steps 211 to 213 may be performed periodically according to a preset period, for example, every 24 hours, and the process is not limited.

In one example, the historical database may also record a tag value of the first power plan, which may be the first identifier (e.g., 1) or the second identifier (e.g., 0). The first flag indicates that the first power plan is a normal power plan, and the second flag indicates that the first power plan is an abnormal power plan.

For example, the historical database includes a first power plan 1, a first power plan 2, and a first power plan 3, the tag values of the first power plan 1 and the first power plan 2 are first identifiers, and the tag value of the first power plan 3 is a second identifier. In practical application, the number of the first power plans in the historical database is far greater than 3, and the number of the first power plans is not limited.

With respect to step 212, in one example, the first grid node layout may include, but is not limited to: power information for each node in the power system. The manner of obtaining the first grid node layout according to the first power plan is not limited herein, as long as the first grid node layout can include power information of each node in the power system. The following describes a process of obtaining a first grid node layout with reference to the specific embodiment.

Specifically, the process of "obtaining the first grid node layout according to the first power plan" may include, but is not limited to, the following ways: acquiring a power grid topological structure and power information of each node in the power grid topological structure from the first power plan; the power grid topology structure may include a topology structure of each node in the power system; and generating a first power grid node layout according to the power grid topological structure and the power information of each node.

Further, the process of "generating a first grid node layout according to the grid topology and the power information of each node" may include: generating a first intermediate node layout according to the topological structure of each node in the power system (namely the power grid topological structure); for convenience of distinguishing, the intermediate node layout generated based on the power grid topological structure in the training stage is called as a first intermediate node layout. Then, based on the power information of each node, marking color information corresponding to the power information of the node at a position corresponding to each node in the first intermediate node layout, thereby converting the first intermediate node layout into a first power grid node layout.

Nodes may include, but are not limited to: lines, generator sets, transformers, power units, etc., although other types of nodes are possible, without limitation. The power information may include, but is not limited to: the electric quantity value, such as the electric quantity value generated by the generator set, the electric quantity value transmitted by the line, the electric quantity value consumed by the electric unit and the like. When the power information of the nodes is counted, the nodes may be a single node or a plurality of nodes. For example, the power plant a has a plurality of generator sets corresponding to a node, and the sum of the electric quantity values generated by the plurality of generator sets is the electric quantity value of the node, i.e., the electric power information of the node. For another example, the manufacturer a has a plurality of power consumption units corresponding to a node, and the sum of the power consumption values used by the plurality of power consumption units is the power consumption value of the node, that is, the power information of the node.

The above process is described below with reference to the first power plan 1, and the other power plans such as the first power plan 2 and the first power plan 3 are similar to the first power plan 1 and will not be described again here.

The first power plan 1 may include, but is not limited to, a power grid topology, and power information of each node in the power grid topology. Referring to fig. 2B, as an example of the grid topology in the first power plan 1, the nodes in fig. 2B may be lines, gensets, transformers, power units, etc. Further, the first power plan 1 may also include power information of each node (i.e., an electric quantity value of each node).

The power grid topological structure can comprise 11 nodes, wherein the generator set A is a node 1, the generator set B is a node 2, the lines 1-6 are respectively a node 3-a node 8, and the power utilization units C-E are respectively a node 9-a node 11. If the electric quantity value generated by the generator set A is 1000, the electric quantity value generated by the generator set B is 500, the electric power information of the node 1 is 1000, and the electric power information of the node 2 is 500. If the electric power values transmitted by the lines 1 to 6 are 400, 100, 300 and 500, respectively, the electric power information of the nodes 3 to 8 is 400, 100, 300 and 500, respectively. If the power consumption values of the power consumption units C to E are 500, 200 and 800, respectively, the power information of the nodes 9 to 11 is 500, 200 and 800, respectively.

In the application scenario, the power grid topology shown in fig. 2B may be analyzed from first power plan 1, and the power information of nodes 1 to 11 in the power grid topology may be analyzed from first power plan 1. The power grid topology shown in fig. 2B is converted into the first intermediate node layout shown in fig. 2C, and the conversion process of the first intermediate node layout is not described again, as long as the first intermediate node layout includes each node in the power grid topology, which is described by taking fig. 2C as an example. In fig. 2C, the area where the generator set a is located is the position corresponding to the node 1, that is, the position in the frame where the node 1 is located; the area where the generator set B is located is the position corresponding to the node 2, namely the position in the frame where the node 2 is located; the area where the line 1 is located is the position corresponding to the node 3, namely the position in the frame where the node 3 is located; by analogy, other nodes are not described in detail.

Based on the power information 1000 of the node 1, color information corresponding to the power information 1000 is marked at a position corresponding to the node 1 (i.e., a frame in which the node 1 is located) in the first intermediate node layout. Based on the power information 500 of the node 2, color information corresponding to the power information 500 is marked at a position corresponding to the node 2 in the first intermediate node layout. By analogy, after the color information corresponding to the electric power information is marked on the 11 nodes in the first intermediate node layout, the first power grid node layout can be obtained, and the structure of the first power grid node layout is not limited.

In order to label the color information corresponding to the power information, a corresponding relationship between the power information and the color information may be configured, so that after the power information of the node is obtained, the color information corresponding to the power information may be obtained by querying the corresponding relationship, and the color information corresponding to the power information is labeled.

For example, the correspondence of power information 1-10 to color 1, the correspondence of power information 11-20 to color 2, and so on are configured, the color is darker as the power information is larger. In practical applications, the different colors may be different RGB (Red Green Blue ) colors, which is not limited herein.

In one example, when the correspondence relationship between the power information and the color information is arranged, cold colors and warm colors may be distinguished, the warm colors may be red, orange, and the like, and the cold colors may be green, blue, and the like. Specifically, if the node is a node that generates an electric quantity value or transmits an electric quantity value, the color information corresponding to the electric power information of the node is a warm color, and the warm color is darker as the electric power information is larger. If the node is a node consuming an electric quantity value, the color information corresponding to the electric power information of the node is a cold color, and the cold color is darker as the electric power information is larger.

In summary, the correspondence between the power information and the color information is not limited, and it is only required to find the corresponding color information based on the power information and label the color information at the position corresponding to the node.

For step 213, in an example, after obtaining the first grid node layout, a tag value of the first grid node layout may also be determined, where the tag value is a first identifier or a second identifier, the first identifier indicates that the first grid node layout is a normal grid node layout, and the second identifier indicates that the first grid node layout is an abnormal grid node layout. For example, a tag value of a first power plan, that is, a tag value of a first grid node layout corresponding to the first power plan, may be obtained from a history database, and the tag value may be a first identifier (e.g., 1) or a second identifier (e.g., 0).

Wherein the first identifier indicates that the first grid node layout is a normal grid node layout, that is, the first grid node layout is a normal node layout sample, that is, a normal sample; furthermore, the second identifier indicates that the first grid node layout is an abnormal grid node layout, that is, the first grid node layout is an abnormal node layout sample, that is, an abnormal sample.

After the first power grid node layout and the label values of the first power grid node layout are obtained, the corresponding relation between the power grid characteristics and the characteristic values can be trained according to the first power grid node layout and the label values of the first power grid node layout, and in an analysis stage, the state of the power system can be analyzed according to the corresponding relation.

In one example, the process for "training the correspondence between the grid feature and the feature value according to the first grid node layout and the label value of the first grid node layout" may include, but is not limited to, the following ways: the first grid node layout and the label value of the first grid node layout may be output to a machine learning network (e.g., a neural network such as CNN, LSTM, etc.), and the machine learning network trains a corresponding relationship between the grid characteristic and the characteristic value according to the first grid node layout and the label value of the first grid node layout.

Since a large number of first power plans may be stored in the historical database, a large number of first grid node layouts, and tag values for these first grid node layouts, may be obtained. A number of first grid node layouts, and tag values for those first grid node layouts, may then be output to the machine learning network. Based on a large amount of input data, the machine learning network can adopt a machine learning algorithm (such as a deep learning algorithm) to train the corresponding relation between the power grid characteristics and the characteristic values, and the training process is not repeated. Finally, the corresponding relation between the grid characteristics and the characteristic values can be output to a classifier of the machine learning network.

The grid characteristic may be an image characteristic of the first grid node layout, such as a color characteristic, a texture characteristic, a shape characteristic, a spatial relationship characteristic, and the like, which is not limited to this grid characteristic.

The characteristic value may be a probability value of the normal power plan or a probability value of the abnormal power plan. For example, if the machine learning network is used to train the correspondence between the grid characteristics and the probability value of the normal power plan, the characteristic value may be a probability value of the normal power plan, such as 95%, where 95% indicates that the probability value of the power plan being the normal power plan is 95%. If the machine learning network is used to train the correspondence between the grid characteristics and the probability value of the abnormal power plan, the characteristic value may be a probability value of the abnormal power plan, such as 95%, where 95% indicates that the probability value of the abnormal power plan is 95%.

According to the power grid node layout, when the corresponding relation between the power grid characteristics and the characteristic values is trained, the accuracy of the training result is determined by the number of the power grid node layouts, and therefore a large number of power grid node layouts, such as a large number of normal samples (normal power grid node layouts) and a large number of abnormal samples (abnormal power grid node layouts), can be provided.

The process flow shown in fig. 2D may also be used in consideration that the first power plan number may be relatively small (e.g., the abnormal first power plan number may be relatively small), resulting in a relatively small number of first grid node layouts. Referring to fig. 2D, which is a flowchart of an analysis method of a power system, fig. 2B is a flowchart of a training phase, and fig. 2D is a flowchart based on the application scenario shown in fig. 1B, the method may include:

step 221, a first power plan is obtained.

Step 222, a first grid node layout is obtained according to the first power plan.

Step 223, processing the first power grid node layout by using a generated countermeasure network algorithm to obtain a plurality of second power grid node layouts, wherein the label values of the second power grid node layouts are the same as those of the first power grid node layouts.

And 224, training the corresponding relation between the power grid characteristics and the characteristic values according to the label values of the first power grid node layout and the label values of the second power grid node layout and the second power grid node layout.

Wherein, in the analysis stage, the state of the power system can be analyzed using the correspondence.

Step 221 and step 222 are similar to step 211 and step 212, and are not described herein again.

For step 223, if the tag value of the first power grid node layout is the first identifier (indicating a normal power grid node layout), the tag value of the second power grid node layout obtained when the first power grid node layout is processed by using the generated countermeasure network algorithm is the first identifier. If the label value of the first power grid node layout is the second identifier (representing abnormal power grid node layout), the label value of the second power grid node layout obtained when the first power grid node layout is processed by adopting a generation countermeasure network algorithm is the second identifier.

Specifically, after a large number of first power grid node layouts are obtained, the first power grid node layouts with the label values as the first identifications are selected from the first power grid node layouts, and the selected first power grid node layouts are output to the generation countermeasure network. And the generation countermeasure network is based on the first power grid node layouts, a large number of second power grid node layouts are obtained by adopting a generation countermeasure network algorithm, and label values of the second power grid node layouts are first marks. In addition, after a large number of first power grid node layouts are obtained, the first power grid node layouts with the label values as the second identifications are selected from the first power grid node layouts, and the selected first power grid node layouts are output to the generation countermeasure network. And the generation countermeasure network is based on the first power grid node layouts, a large number of second power grid node layouts are obtained by adopting a generation countermeasure network algorithm, and label values of the second power grid node layouts are second identifications.

Among them, generating a countermeasure network (GANs) is a highly innovative method in the field of deep learning, and its main idea is: through the learning of the existing samples, more new samples are generated, the new samples can reflect the inherent attributes of the original samples, the original samples are effectively supplemented, and the new samples have the diversity of information, so that the reliability of the samples is higher, and the identification precision is higher.

The generation countermeasure network can be composed of two parts, the first part is a generator for generating new samples, and the second part is a discriminator for discriminating and generating new samples. By means of the mode of generating and judging countermeasures, existing unlabeled data can be fully mined in the scene of lacking of enough and supervised labeled data, and possible normal or abnormal changes can be found and distinguished, so that more effective power characteristics can be obtained.

In summary, the first power grid node layout may be output to the generation countermeasure network, so that the generation countermeasure network generates more second power grid node layouts (i.e., new samples) based on the first power grid node layout, and the generation process is not limited, where the second power grid node layouts are images with higher precision and better image quality.

In the above embodiment, the number of the second grid node layouts generated by the generation countermeasure network may be configured empirically, or the generation countermeasure network may be generated according to actual needs, which is not limited to this.

For step 224, in an example, after obtaining the tag values of the first grid node layout and the first grid node layout, and the tag values of the second grid node layout and the second grid node layout, the corresponding relationship between the grid characteristic and the characteristic value may be trained according to the tag values of the first grid node layout and the first grid node layout, and the tag values of the second grid node layout and the second grid node layout. Further, in the analysis stage, the state of the power system can be analyzed according to the corresponding relation between the power grid characteristics and the characteristic values.

In one example, the process of training a correspondence between a grid characteristic and a characteristic value according to the tag values of the first grid node layout and the first grid node layout, and the tag values of the second grid node layout and the second grid node layout may include: the label values of the first grid node layout and the first grid node layout, and the label values of the second grid node layout and the second grid node layout may be output to a machine learning network (e.g., a neural network such as CNN, LSTM), and the machine learning network trains the corresponding relationship between the grid characteristics and the characteristic values according to the label values of the first grid node layout and the first grid node layout, and the label values of the second grid node layout and the second grid node layout.

After a large number of power grid node layouts (such as a first power grid node layout and a second power grid node layout) and label values of the power grid node layouts are obtained, the large number of power grid node layouts and the label values of the power grid node layouts can be output to the machine learning network. Based on a large amount of input data, the machine learning network can adopt a machine learning algorithm (such as a deep learning algorithm) to train the corresponding relation between the power grid characteristics and the characteristic values, and the training process is not repeated. Finally, the corresponding relation between the grid characteristics and the characteristic values can be output to a classifier of the machine learning network. The power grid features may be image features of a power grid node layout, such as color features, texture features, shape features, spatial relationship features, and the like, which are not limited to this power grid feature. Further, the characteristic value may be a probability value of a normal power plan, or a probability value of an abnormal power plan.

For the training stage (fig. 2A and 2D), the corresponding relationship between the grid characteristics and the characteristic values may be periodically trained, so that the corresponding relationship between the grid characteristics and the characteristic values is maintained, and the reliability is high.

Referring to fig. 2E, which is a flowchart of an analysis method of a power system, fig. 2E is a flowchart of an analysis stage, and an applicable scenario thereof may be as shown in fig. 1A and fig. 1B, where the method may include:

in step 231, a second power plan requiring analysis of the state is obtained.

Step 232, a third grid node layout is obtained according to the second power plan.

And 233, inquiring the corresponding relation according to the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics. And the corresponding relation is the corresponding relation between the power grid characteristic trained according to the first power plan and the characteristic value.

The status of the second power plan is analyzed based on the obtained characteristic values, step 234.

With respect to step 231, in one example, the process for "obtaining a second power plan that requires analyzing the status" may include, but is not limited to: and acquiring a power generation plan, a power demand forecast and a power grid topological structure, and acquiring a second power plan according to the power generation plan, the power demand forecast and the power grid topological structure.

The power generation plan may be an electric quantity value generated by the generator set, and the power management device may acquire the power generation plan from the power plant, where an acquisition manner of the power generation plan is not limited.

The power demand prediction may be a predicted power consumption value of the power consumption unit, and the power management device may predict the power consumption value of the power consumption unit according to the historical data, without limiting an acquisition manner of the power demand prediction. For example, if it is known that the average daily consumed electricity value of the electricity consumption unit is 500 based on the history data, it may be predicted that the consumed electricity value of the electricity consumption unit is 500; alternatively, when the electricity consumption value consumed by the electricity consumption unit in the previous day is known to be 520, it can be predicted that the electricity consumption value consumed by the electricity consumption unit is 520.

The power grid topology may be a topology of each node in the power system, and the node may include but is not limited to: lines, generator sets, transformers, power units, etc. The power management device may generate a power grid topology structure according to the position of the generator set, the position of the transformer, the position of the power consumption unit, all lines among the generator set, the transformer, and the power consumption unit, and the like, and the obtaining manner of the power grid topology structure is not limited. For example, the power management device may generate the grid topology shown in fig. 2B.

After the power generation plan, the power demand forecast, and the grid topology are obtained, a second power plan can be obtained according to the power generation plan, the power demand forecast, and the grid topology. The second power plan may include the power grid topology and power information of each node in the power grid topology.

Taking the power grid topology shown in fig. 2B as an example, based on the power generation plan, if the electric quantity value generated by the generator set a is 1000, the electric quantity value generated by the generator set B is 500, the electric power information of the node 1 corresponding to the generator set a is 1000, and the electric power information of the node 2 corresponding to the generator set B is 500. Based on the power demand prediction, if the power consumption values of the power consumption units C and E are predicted to be 500, 200, and 800, respectively, the power information of the nodes 9 to 11 corresponding to the power consumption units C and E is 500, 200, and 800, respectively.

Based on the power generation schedule and the power demand forecast, it is known that the amount of power transmitted by lines 1-6 is 500, that of lines 1 and 2 is 200, that of lines 3 and 4 is 800, that of lines 5 and 6 is 1000, and that of lines 6 is 500. To achieve the above object, examples of the electric quantity values transmitted by the lines 1 to 6 may be 400, 100, 300, 500, i.e. the electric power information of the nodes 3 to 8 corresponding to the lines 1 to 6 is 400, 100, 300, 500, respectively.

Of course, the above process is only an example of "obtaining the second power plan according to the power generation plan, the power demand prediction, and the grid topology", and the obtaining manner is not limited.

With respect to step 232, in one example, the third grid node layout may include, but is not limited to: power information for each node in the power system. The manner of obtaining the third grid node layout according to the second power plan is not limited herein, as long as the third grid node layout can include power information of each node in the power system. The process of obtaining the third grid node layout is described below with reference to the specific embodiment.

Specifically, the process of "obtaining the third grid node layout according to the second power plan" may include, but is not limited to, the following ways: acquiring a power grid topological structure and power information of each node in the power grid topological structure from the second power plan; the power grid topology structure may include a topology structure of each node in the power system; and generating a third power grid node layout according to the power grid topological structure and the power information of each node.

Further, the process of "generating a third grid node layout according to the grid topology and the power information of each node" may include: generating a second intermediate node layout according to the topology structure of each node in the power system (namely the power grid topology structure); for convenience of distinguishing, the intermediate node layout generated based on the power grid topological structure in the analysis stage is called as a second intermediate node layout. And then, marking color information corresponding to the power information of the node at the position corresponding to each node in the second intermediate node layout based on the power information of each node, so as to convert the second intermediate node layout into a third power grid node layout.

The process of "obtaining the third grid node layout according to the second power plan" is similar to the process of "obtaining the first grid node layout according to the first power plan" in step 212, and is not described herein again.

In step 233, after the third power grid node layout is obtained, the power grid characteristics of the third power grid node layout may be determined, and the correspondence is queried according to the power grid characteristics (i.e., in the training phase, the trained correspondence between the power grid characteristics and the characteristic values), so as to obtain the characteristic values corresponding to the power grid characteristics.

The process of obtaining the characteristic value corresponding to the power grid characteristic by querying the corresponding relationship for the power grid characteristic laid out by the third power grid node may include: and outputting the third power grid node layout to a machine learning network, determining the power grid characteristics corresponding to the third power grid node layout by the machine learning network, and inquiring the corresponding relation generated in the training stage through the power grid characteristics to obtain the characteristic value corresponding to the power grid characteristics.

The grid feature may be an image feature of the third grid node layout, such as a color feature, a texture feature, a shape feature, a spatial relationship feature, and the like, which is not limited to this grid feature. Further, the above-described feature value may be a probability value of a normal power plan, or a probability value of an abnormal power plan.

For step 234, in one example, the process for "analyzing the status of the second power plan according to the obtained characteristic values" may include, but is not limited to, the following: if the characteristic value is the probability value of the normal power plan, when the characteristic value is larger than a preset first threshold value, determining that the state of the second power plan is normal; otherwise, determining that the state of the second power plan is abnormal. Or if the characteristic value is the probability value of the abnormal power plan, when the characteristic value is greater than a preset second threshold value, determining that the state of the second power plan is abnormal; otherwise, determining that the state of the second power plan is normal.

The preset first threshold may be configured empirically, the preset second threshold may be configured empirically, and the preset first threshold and the preset second threshold may be the same or different. For example, the preset first threshold may be 90%, 85%, etc., and the preset second threshold may be 90%, 95%, etc.

In one example, after determining that the status of the second power plan is normal, the second power plan may also be stored in a historical database, and a tag value of the second power plan is recorded in the historical database, where the tag value is a first identifier (e.g., 1) indicating a normal power plan. After determining that the status of the second power plan is abnormal, the second power plan may be stored in the historical database, and a tag value of the second power plan, which is a second identifier (e.g., 0) indicating an abnormal power plan, may be recorded in the historical database.

In an example, after determining that the status of the second power plan is abnormal, a new second power plan may be obtained according to the power generation plan, the power demand prediction, and the power grid topology, and based on the new second power plan, the above steps 232 to 234 may be performed, and if the status of the second power plan is still abnormal, the new second power plan may be obtained according to the power generation plan, the power demand prediction, and the power grid topology, and so on.

In addition, after the second power plan is determined to be abnormal, the abnormal information can be notified to the power management personnel, and the power management personnel can adjust the power generation plan, the power demand prediction and the power grid topological structure, so that the purposes of guiding the power grid plan, operating and scheduling and overhauling the power failure plan are achieved.

Based on the same application concept as the method, the embodiment of the application also provides an analysis method of the power system, and the method can be applied to a power management device which is used for analyzing the state of the power system, such as performing static safety analysis. The power management device may be a PC, a notebook computer, a mobile terminal, a server, or the like, and the type of the power management device is not limited in detail.

Referring to fig. 3, a flow chart of an analysis method of a power system is shown, which may include:

step 301, a third power plan requiring analysis of the state is obtained.

Step 302, inquiring whether a normal power plan matching with the third power plan exists in the normal sample library. If so, step 303 may be performed; if not, step 304 may be performed.

In step 303, it is analyzed that the status of the third power plan is normal.

At step 304, it is analyzed that the status of the third power plan is abnormal.

With respect to step 301, in one example, the process for "obtaining a third power plan that requires analyzing the status" may include, but is not limited to: and acquiring a power generation plan, a power demand prediction and a power grid topological structure, and acquiring a third power plan according to the power generation plan, the power demand prediction and the power grid topological structure.

The processing procedure of step 301 is similar to the processing procedure of step 231, and for convenience of distinction, the power plan whose state needs to be analyzed is referred to as a third power plan, which is not repeated herein.

With respect to step 302, in one example, the power management device may maintain a normal sample library that is used to record a large number of normal power plans. Based on this, after the third power plan is obtained, it may be queried whether there is a normal power plan matching the third power plan in the normal sample library. If yes, namely the third power plan exists in the normal sample library, the third power plan is the normal power plan. If not, namely the third power plan does not exist in the normal sample library, the third power plan is an abnormal power plan.

In order to record a large number of normal power plans in the normal sample library, the method may further include: acquiring a normal power plan (which can be a plurality of normal power plans) from a historical database, and processing the normal power plan by adopting a generated countermeasure network algorithm to obtain a plurality of new normal power plans; and then, storing the normal power plans acquired from the historical database and a plurality of normal power plans obtained by adopting a generated countermeasure network algorithm into a normal sample library so as to record a large number of normal power plans in the normal sample library.

Among other things, the power management device may maintain a historical database for storing power plans, which may include normal power plans and abnormal power plans. For example, a tag value of each power plan may be recorded in the historical database, and the tag value may be a first flag (e.g., 1) indicating that the power plan is a normal power plan or a second flag (e.g., 0) indicating that the power plan is an abnormal power plan. Based on this, the power management apparatus may acquire the normal power plan from the history database, and store the acquired normal power plan in the normal sample library.

After the power management equipment acquires the normal power plan from the historical database, the power management equipment can also process the normal power plan by adopting a generation countermeasure network algorithm to obtain a plurality of new normal power plans, and the new normal power plans are stored in a normal sample library. For example, the normal power plans may be output to a generative countermeasure network, which is based on these normal power plans, a generative countermeasure network algorithm may be employed to derive a number of new normal power plans, and these new normal power plans may be stored to a normal sample repository.

The content of generating the countermeasure network is described above, and is not described herein again. In summary, after outputting the normal power plan to the generation countermeasure network, the generation countermeasure network may generate more normal power plans (i.e., new samples) based on the normal power plan, without limitation to the generation process.

In one example, after analyzing that the state of the third power plan is normal, the third power plan may be stored in the normal sample library and the third power plan may be stored in the historical database, and the tag value of the third power plan, i.e., the first identifier (e.g., 1), is recorded in the historical database and represents the normal power plan.

In an example, after analyzing that the state of the third power plan is abnormal, a new third power plan may be obtained according to the power generation plan, the power demand prediction, and the power grid topology, and based on the new third power plan, the step 302 may be executed again, and if the state of the third power plan is still abnormal, the new third power plan may be obtained according to the power generation plan, the power demand prediction, and the power grid topology, and so on.

In addition, after the third power plan is determined to be abnormal, the abnormal information can be notified to the power management personnel, and the power management personnel can adjust the power generation plan, the power demand prediction and the power grid topological structure, so that the purposes of guiding the power grid plan, operating and scheduling and overhauling the power failure plan are achieved.

Based on the technical scheme, in the embodiment of the application, the state of the power system can be analyzed based on the normal power plan in the normal sample library, the power grid planning, the operation scheduling and the maintenance power failure plan are guided based on the result of the safety analysis, the static safety analysis of the power system can be rapidly carried out, and the power planning and the scheduling are supported. The above mode does not need to adopt a violent enumeration mode to sequentially disconnect each node of the power system, so that the time of safety analysis can be saved, and the computing resources of the safety analysis can be saved. Under emergency, safety analysis can be carried out in a short time, so that electric power operators can make scheduling decisions quickly, and large-scale electric power faults are avoided.

Based on the same application concept as the method, the embodiment of the present application further provides an analysis apparatus of an electric power system, which can be applied to an electric power management device, as shown in fig. 4, and is a structural diagram of the apparatus.

An obtaining module 401, configured to obtain a first power plan; acquiring a first power grid node layout according to the first power plan; a training module 402, configured to train a correspondence between a power grid characteristic and a characteristic value according to the first power grid node layout and a label value of the first power grid node layout; the corresponding relation is used for analyzing the state of the power system.

The obtaining module 401 is specifically configured to, in a process of obtaining a first power grid node layout according to the first power plan, obtain a power grid topology structure and power information of each node in the power grid topology structure from the first power plan; the power grid topological structure comprises a topological structure of each node in a power system; generating a first power grid node layout according to the power grid topological structure and the power information of each node;

the training module 402 is specifically configured to, in the process of training a correspondence between a power grid feature and a feature value according to the first power grid node layout and a tag value of the first power grid node layout, process the first power grid node layout by using a generation countermeasure network algorithm to obtain a plurality of second power grid node layouts, where the tag value of the second power grid node layout is the same as the tag value of the first power grid node layout; and training the corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout, the label values of the first power grid node layout, the plurality of second power grid node layouts and the label values of the second power grid node layouts.

The obtaining module 401 is further configured to obtain a second power plan of a state to be analyzed; acquiring a third power grid node layout according to the second power plan; inquiring the corresponding relation according to the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics of the third power grid node layout;

the device further comprises (not shown in the figures):

Based on the same application concept as the method, the embodiment of the application also provides a power management device, which may include a processor; wherein: the processor is used for obtaining a first power plan; acquiring a first power grid node layout according to the first power plan; training a corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout and the label values of the first power grid node layout; wherein the corresponding relation is used for analyzing the state of the power system.

Based on the same application concept as the method, the embodiment of the present application further provides a machine-readable storage medium, where the machine-readable storage medium may be applied to a power management device, and the machine-readable storage medium has stored thereon several computer instructions, where the computer instructions, when executed, perform the following processes: acquiring a first power plan; acquiring a first power grid node layout according to the first power plan; training a corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout and the label values of the first power grid node layout; wherein the corresponding relation is used for analyzing the state of the power system.

Based on the same application concept as the method, the embodiment of the present application further provides an analysis apparatus of an electric power system, which can be applied to an electric power management device, as shown in fig. 5, and is a structural diagram of the apparatus.

An obtaining module 501, configured to obtain a second power plan of a state to be analyzed; obtaining a third power grid node layout according to the second power plan; inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values;

an analyzing module 502, configured to analyze a state of the second power plan according to the obtained feature value.

The obtaining module 501 is specifically configured to, in the process of obtaining the layout of the third power grid node according to the second power plan, obtain a power grid topology structure and power information of each node in the power grid topology structure from the second power plan; the power grid topological structure comprises a topological structure of each node in a power system; generating a third power grid node layout according to the power grid topological structure and the power information of each node;

the analysis module 502 is specifically configured to determine that the state of the second power plan is normal if the feature value is a probability value of a normal power plan, and when the feature value is greater than a preset first threshold; or if the characteristic value is a probability value of an abnormal power plan, and when the characteristic value is greater than a preset second threshold, determining that the state of the second power plan is abnormal.

Based on the same application concept as the method, an embodiment of the present application further provides a power management apparatus, including: the processor is used for acquiring a second power plan of a state needing to be analyzed; obtaining a third power grid node layout according to the second power plan; inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values; and analyzing the state of the second power plan according to the obtained characteristic value.

Based on the same application concept as the method, the embodiment of the present application further provides a machine-readable storage medium, which can be applied to a power management device, where the machine-readable storage medium stores thereon several computer instructions, and when the computer instructions are executed, the computer instructions perform the following processes: acquiring a second power plan of a state needing to be analyzed; obtaining a third power grid node layout according to the second power plan; inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values; and analyzing the state of the second power plan according to the obtained characteristic value.

Based on the same application concept as the method, the embodiment of the present application further provides an analysis apparatus of an electric power system, which can be applied to an electric power management device, as shown in fig. 6, and is a structural diagram of the apparatus.

An obtaining module 601, configured to obtain a third power plan of a state to be analyzed;

a query module 602, configured to query whether a normal power plan matching the third power plan exists in a normal sample library;

an analysis module 603, configured to, if the query result is yes, analyze that the state of the third power plan is normal; and when the query result is negative, analyzing that the state of the third power plan is abnormal.

Based on the same application concept as the method, the embodiment of the application also provides a power management device, which may include a processor; wherein: the processor is used for acquiring a third power plan of a state needing to be analyzed; querying whether a normal power plan matched with the third power plan exists in a normal sample library; if yes, analyzing that the state of the third power plan is normal; if not, analyzing that the state of the third power plan is abnormal.

Based on the same application concept as the method, the embodiment of the present application further provides a machine-readable storage medium, where the machine-readable storage medium may be applied to a power management device, and the machine-readable storage medium has stored thereon several computer instructions, where the computer instructions, when executed, perform the following processes: acquiring a third power plan of a state to be analyzed; querying whether a normal power plan matched with the third power plan exists in a normal sample library; if yes, analyzing that the state of the third power plan is normal; if not, analyzing that the state of the third power plan is abnormal.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. A typical implementation device is a computer, which may take the form of a personal computer, laptop computer, cellular telephone, camera phone, smart phone, personal digital assistant, media player, navigation device, email messaging device, game console, tablet computer, wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the units may be implemented in one or more software and/or hardware when implementing the present application.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application may 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, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may 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, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

Furthermore, these computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above description is only an example of the present application and is not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method of analyzing an electrical power system, the method comprising:

acquiring a first power plan;

acquiring a first power grid node layout according to the first power plan;

2. The method of claim 1,

the first grid node layout comprises power information of each node in the power system; the tag value is a first identifier or a second identifier; the first identifier represents that the first power grid node layout is a normal power grid node layout, and the second identifier represents that the first power grid node layout is an abnormal power grid node layout.

3. The method of claim 1,

the process of obtaining a first power grid node layout according to the first power plan specifically includes:

acquiring a power grid topological structure and power information of each node in the power grid topological structure from a first power plan; the power grid topological structure comprises a topological structure of each node in a power system;

and generating a first power grid node layout according to the power grid topological structure and the power information of each node.

4. The method according to claim 3, wherein the process of generating a first grid node topology from the grid topology and the power information of each node specifically comprises:

generating a first intermediate node layout according to the topological structure of each node in the power system;

and marking color information corresponding to the power information of the node at the position corresponding to each node in the first intermediate node layout based on the power information of each node to obtain the first power grid node layout.

5. The method of claim 1, wherein training the correspondence between the grid characteristics and the characteristic values according to the first grid node layout and the label values of the first grid node layout comprises:

processing the first power grid node layout by adopting a generation countermeasure network algorithm to obtain a plurality of second power grid node layouts, wherein the label values of the second power grid node layouts are the same as those of the first power grid node layouts;

and training the corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout, the label values of the first power grid node layout, the plurality of second power grid node layouts and the label values of the second power grid node layouts.

6. The method according to claim 1 or 5, wherein training the correspondence between the grid characteristics and the characteristic values according to the first grid node layout and the label values of the first grid node layout comprises:

and outputting the first power grid node layout and the label value to a machine learning network, and training the corresponding relation between the power grid characteristics and the characteristic value by the machine learning network according to the first power grid node layout and the label value.

7. The method of claim 1,

after training the corresponding relation between the grid characteristics and the characteristic values according to the first grid node layout and the label values of the first grid node layout, the method further comprises the following steps:

acquiring a second power plan of a state needing to be analyzed;

obtaining a third power grid node layout according to the second power plan;

inquiring the corresponding relation according to the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics of the third power grid node layout;

8. The method of claim 7,

the process of obtaining the second power plan of the state to be analyzed specifically includes:

the method comprises the steps of obtaining a power generation plan, a power demand forecast and a power grid topological structure, and obtaining a second power plan according to the power generation plan, the power demand forecast and the power grid topological structure.

9. The method of claim 7,

the process of obtaining a third power grid node layout according to the second power plan specifically includes:

acquiring a power grid topological structure and power information of each node in the power grid topological structure from a second power plan; the power grid topological structure comprises a topological structure of each node in a power system;

and generating a third power grid node layout according to the power grid topological structure and the power information of each node.

10. The method according to claim 9, wherein the process of generating a third grid node topology from the grid topology and the power information of each node specifically includes:

generating a second intermediate node layout according to the topological structure of each node in the power system;

and marking color information corresponding to the power information of the node at the position corresponding to each node in the second intermediate node layout based on the power information of each node to obtain the third power grid node layout.

11. The method of claim 7,

the process of querying the corresponding relationship through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics of the third power grid node layout specifically includes:

and outputting the third power grid node layout to a machine learning network, determining the power grid characteristics corresponding to the third power grid node layout by the machine learning network, and inquiring the corresponding relation according to the power grid characteristics to obtain the characteristic value corresponding to the power grid characteristics.

12. The method of claim 7,

the characteristic value is a probability value of a normal power plan or a probability value of an abnormal power plan;

the process of analyzing the state of the second power plan according to the obtained characteristic value specifically includes:

if the characteristic value is the probability value of the normal power plan, when the characteristic value is larger than a preset first threshold value, determining that the state of the second power plan is normal; or,

and if the characteristic value is the probability value of the abnormal power plan, when the characteristic value is larger than a preset second threshold value, determining that the state of the second power plan is abnormal.

13. A method of analyzing an electrical power system, the method comprising:

acquiring a second power plan of a state needing to be analyzed;

obtaining a third power grid node layout according to the second power plan;

14. The method of claim 13,

15. The method of claim 13,

16. The method according to claim 15, wherein the process of generating a third grid node topology from the grid topology and the power information of each node specifically comprises:

17. The method of claim 13,

the process of obtaining the characteristic value corresponding to the power grid characteristic by querying the corresponding relation of the power grid characteristic of the third power grid node layout specifically includes:

18. The method of claim 13,

19. A method of analyzing an electrical power system, the method comprising:

acquiring a third power plan of a state to be analyzed;

if yes, analyzing that the state of the third power plan is normal;

if not, analyzing that the state of the third power plan is abnormal.

20. The method of claim 19,

the process of obtaining the third power plan of the state to be analyzed specifically includes:

the method comprises the steps of obtaining a power generation plan, a power demand forecast and a power grid topological structure, and obtaining a third power plan according to the power generation plan, the power demand forecast and the power grid topological structure.

21. The method of claim 19, wherein prior to said querying whether a normal power plan matching the third power plan exists in the normal sample library, the method further comprises:

acquiring a normal power plan, and processing the normal power plan by adopting a generated countermeasure network algorithm to obtain a plurality of new normal power plans;

and storing the obtained normal power plans and a plurality of normal power plans obtained by adopting a generated countermeasure network algorithm into the normal sample library.

22. An analysis device for an electric power system, the device comprising:

23. The apparatus of claim 22,

the obtaining module is specifically configured to obtain a power grid topology structure and power information of each node in the power grid topology structure from the first power plan in a process of obtaining a first power grid node layout according to the first power plan; the power grid topological structure comprises a topological structure of each node in a power system; generating a first power grid node layout according to the power grid topological structure and the power information of each node;

the training module is specifically configured to, in the process of training a corresponding relationship between a power grid characteristic and a characteristic value according to the first power grid node layout and a label value of the first power grid node layout, process the first power grid node layout by using a generation countermeasure network algorithm to obtain a plurality of second power grid node layouts, where a label value of each of the second power grid node layouts is the same as a label value of each of the first power grid node layouts; and training the corresponding relation between the power grid characteristics and the characteristic values according to the first power grid node layout, the label values of the first power grid node layout, the plurality of second power grid node layouts and the label values of the second power grid node layouts.

24. The apparatus of claim 22,

the acquisition module is further used for acquiring a second power plan of a state to be analyzed; acquiring a third power grid node layout according to the second power plan; inquiring the corresponding relation according to the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics of the third power grid node layout;

the device further comprises:

25. An analysis device for an electric power system, the device comprising:

26. The apparatus of claim 25,

the obtaining module is specifically configured to obtain a power grid topology structure and power information of each node in the power grid topology structure from the second power plan in a process of obtaining a third power grid node layout according to the second power plan; the power grid topological structure comprises a topological structure of each node in a power system; generating a third power grid node layout according to the power grid topological structure and the power information of each node;

the analysis module is specifically configured to determine that the state of the second power plan is normal when the feature value is greater than a preset first threshold value if the feature value is a probability value of a normal power plan; or if the characteristic value is a probability value of an abnormal power plan, and when the characteristic value is greater than a preset second threshold, determining that the state of the second power plan is abnormal.

27. An analysis device for an electric power system, the device comprising:

28. A power management device, characterized in that the power management device comprises:

29. A power management device, characterized in that the power management device comprises: a processor for obtaining a second power plan requiring analysis of a state; obtaining a third power grid node layout according to the second power plan; inquiring a corresponding relation through the power grid characteristics of the third power grid node layout to obtain a characteristic value corresponding to the power grid characteristics; the corresponding relation is the corresponding relation between the power grid characteristics trained according to the first power plan and the characteristic values; and analyzing the state of the second power plan according to the obtained characteristic value.

30. A power management device, characterized in that the power management device comprises: