Power grid early warning method and system based on different factor types

Abstract

The invention is suitable for the technical field of power grid early warning, and provides a power grid early warning method and system based on different factor types, and the method comprises the steps: collecting original data of each link of a power grid, and carrying out the preprocessing of the original data; performing text information extraction on the preprocessed data by using a Transform model to obtain target key information; matching the target key information with a preset power grid early warning factor library, and determining a factor type corresponding to the target key information, the power grid early warning factor library comprising a plurality of influence factor types of power grid early warning; and calculating the weight of the factor type corresponding to the target key information based on the optimized index weight method, and determining an early warning result according to the calculated weight. By analyzing the global data of the power grid, the early warning is more comprehensive, the early warning accuracy of the power grid is effectively improved, and the loss and the influence are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid early warning, and in particular to a power grid early warning method and system based on different types of factors. Background Art

[0002] The power grid is responsible for delivering electricity from power plants to end users. It consists of several key components, including the power generation system, transmission system, substation system and distribution system. The power generation system generates electricity through power generation facilities such as thermal power plants, hydropower plants and nuclear power plants. The transmission system uses high-voltage transmission lines and substations to transmit electricity over long distances. The substation system is responsible for voltage rise and fall conversion to meet different transmission and power demand, and the distribution system distributes electricity to specific power load points. In addition, the power grid also includes a monitoring system for real-time monitoring and early warning of the power grid operation status to ensure the reliability and safety of power supply.

[0003] In the existing technology, the power grid monitoring and early warning system has some problems based on these aspects, such as limited data collection, the collected data is limited to the power grid's own factors, such as voltage, frequency and circuit breaker opening, etc., lacking a global perspective, and it is difficult to form a comprehensive understanding of the overall operation status of the power grid; insufficient integration capabilities, the collected data lacks limited integration and comprehensive analysis, resulting in potential fault points that are difficult to accurately predict and identify; insufficient early warning capabilities, slow response speed of the early warning system, limited early warning capabilities, and inability to effectively prevent accidents in a timely manner. The emergence of these problems will easily lead to more serious losses due to early warning prediction or inadequate early warning. Summary of the invention

[0004] The embodiments of the present application provide a power grid early warning method and system based on different types of factors, which are used to solve the problem of more serious losses caused by early warning prediction or inadequate early warning.

[0005] A first aspect of an embodiment of the present application provides a power grid early warning method based on different types of factors, including:

[0006] Collecting raw data from all links of the power grid and preprocessing the raw data;

[0007] Use the Transformer model to extract text information from the preprocessed data to obtain the target key information;

[0008] Matching the target key information with a preset power grid early warning factor library, and determining the factor type corresponding to the target key information, wherein the power grid early warning factor library includes several types of influencing factors of power grid early warning;

[0009] The weights of the factor types corresponding to the target key information are calculated based on the optimized indicator weight method, and the warning results are determined according to the calculated weights.

[0010] Furthermore, the Transformer model is used to extract text information from the preprocessed data to obtain target key information, including:

[0011] Use the encoding network in the Transformer model to encode the preprocessed data sequence to obtain the representation of entity mentions;

[0012] Calculate the correlation between the representation mentioned by the entity and the set influencing factor variable;

[0013] When the calculated correlation meets the preset conditions, the decoding network in the Transformer model is used for decoding to output the target key information.

[0014] Furthermore, the encoding network in the Transformer model is used to encode the preprocessed data sequence, and the expression of the representation of the entity mention is obtained as follows:

[0015] L={l1,l2,…,l N}=BERT(D);

[0016] D={t1,t2,…,t n ,};

[0017] Where: N is the number of words in the preprocessed data sequence, n is the number of sequences after segmentation by the BERT segmenter, L is the entity mention representation, and D is the preprocessed data sequence.

[0018] Furthermore, the calculating of the correlation between the representation mentioned by the entity and the set influencing factor variable includes:

[0019] The cosine similarity between the entity mention and the set influencing factor variable is calculated, and the expression is as follows:

[0020]

[0021] Where: A and B are entity mention and influencing factor variables respectively, a i and b i are the values ​​of the i-th dimension of A and B respectively.

[0022] Furthermore, the types of influencing factors of the several power grid early warnings include meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors.

[0023] Furthermore, the weight of the factor type corresponding to the target key information is calculated based on the optimized indicator weight method, including:

[0024] The first-order weighting calculation is performed on the meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors using the principal weighting method;

[0025] The objective weighting method is used to perform second-order weighting calculation on the meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors;

[0026] The weight of the factor type corresponding to the target key information is generated according to the first-order weighted calculation result and the second-order weighted calculation result.

[0027] Furthermore, the calculation formula for generating the weight of the factor type corresponding to the target key information according to the first-order weighted calculation result and the second-order weighted calculation result is as follows:

[0028]

[0029] Where: k is the combined weight of subjective and objective, ω i is the weight obtained by subjective weighting calculation, ω j The weight is calculated by objective weighting.

[0030] A second aspect of the embodiment of the present application provides a power grid early warning system based on different types of factors, including:

[0031] The raw data acquisition unit is used to collect the raw data of each link of the power grid and pre-process the raw data;

[0032] A target key information determination unit is used to extract text information from the preprocessed data using a Transformer model to obtain target key information;

[0033] A factor type determination unit corresponding to the target key information, used to match the target key information with a preset power grid early warning factor library and determine the factor type corresponding to the target key information, wherein the power grid early warning factor library includes several types of influencing factors of power grid early warning;

[0034] The early warning result determination unit is used to calculate the weight of the factor type corresponding to the target key information based on the optimized indicator weight method, and determine the early warning result according to the calculated weight.

[0035] A third aspect of an embodiment of the present application provides a computer device, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein when the processor executes the computer-readable instructions, the steps of the power grid early warning method based on different factor types as described in any one of the above are implemented.

[0036] A fourth aspect of an embodiment of the present application provides a readable storage medium, which stores computer-readable instructions. When the computer-readable instructions are executed by a processor, the steps of the power grid early warning method based on different factor types as described in any one of the above are implemented.

[0037] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0038] This application first pre-processes the data based on the original data collected from each link of the power grid, and then uses the Transformer model to extract text information from the pre-processed data to obtain the target key information, and then matches the target key information with the preset power grid early warning factor library, and determines the factor type corresponding to the target key information, and calculates the weight of the factor type corresponding to the target key information based on the optimized indicator weight method, and finally determines the early warning result according to the calculated weight. This application analyzes the global data of the power grid to make the early warning more comprehensive, and improves the accurate data foundation for the identification and calculation of power grid fault early warning results based on the type of influencing factors, effectively improving the accuracy of power grid early warning and reducing losses and impacts. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of an embodiment flow of a power grid early warning based on different types of factors in the present invention;

[0040] Figure 2 It is a structural block diagram of an embodiment of a power grid early warning system based on different types of factors in the present invention;

[0041] Figure 3 The figure is a structural block diagram of a computer device in the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0043] In this embodiment, the power grid early warning method based on different types of factors is used to improve the accuracy of power grid early warning. In this embodiment, the implementation method can be implemented in the system, can be implemented in the server, and can also be implemented in the terminal, and the specific details are not clearly limited.

[0044] Embodiment 1

[0045] See also Figure 1 , an embodiment of a power grid early warning method based on different factor types in the present invention comprises the following steps:

[0046] S11. Collect the original data of each link of the power grid and pre-process the original data;

[0047] In this embodiment, the original data of each link of the power grid is collected in real time through the Internet of Things sensor equipment, where the original data includes the data generated by the operation of all links involved in the power grid. The system preprocesses the collected data, which includes data cleaning, denoising and standardization. Specifically, the sliding average filtering algorithm is used to implement data preprocessing. First, the collected original data is processed, and a sliding window size is set, such as N sampling points. The average value of the data in the window is calculated as the filter value of the current point. The sliding window moves forward point by point, and the entire signal sequence is filtered; valid data is screened by comparing the signal waveforms before and after filtering.

[0048] S12. Use the Transformer model to extract text information from the preprocessed data to obtain target key information;

[0049] In this embodiment, the Transformer model is used to extract text information from the preprocessed data to obtain target key information, including:

[0050] 1. Use the encoding network in the Transformer model to encode the preprocessed data sequence to obtain the representation of entity mentions;

[0051] First, the preprocessed data is converted into a word vector sequence, and then encoded using the encoding network in the Transformer model to obtain the entity mention representation, which is expressed as follows:

[0052] L={l1,l2,…,l N}=BERT(D);

[0053] D={t1,t2,…,t n ,};

[0054] Where: N is the number of words in the preprocessed data sequence, n is the number of sequences after segmentation by the BERT segmenter, L is the entity mention representation, and D is the preprocessed data sequence.

[0055] 2. Calculate the correlation between the representation mentioned by the entity and the set influencing factor variables;

[0056] Specifically, the representation of entity mentions is manifested as entity representation and entity relationship representation. For example, if the word vector sequence contains entity-related entities such as weather, equipment, and environment, the correlation between all influencing factor variables of weather, equipment, and environment entities is calculated respectively. It is necessary to calculate the pairwise correlation between each entity and each variable in the influencing factor. The calculation formula is as follows:

[0057]

[0058] Where: A and B are entity mention and influencing factor variables respectively, a i and b i are the values ​​of the i-th dimension of A and B respectively.

[0059] 3. When the calculated correlation meets the preset conditions, the decoding network in the Transformer model is used for decoding to output the target key information.

[0060] Specifically, the relevant preset conditions are set to two types here, one is related, and the other is irrelevant. A threshold is set. When the calculated correlation coefficient is greater than or equal to the threshold, it is related; when the calculated correlation coefficient is less than the threshold, it is irrelevant. When the calculated correlation meets the conditions, the decoding network in the Transformer model is used for decoding and the target key information is output. At this time, the target key information may contain one key information or multiple key information. If it contains multiple key information, a key information data set is generated, and then the subsequent matching process is performed based on the data set.

[0061] S13. Match the target key information with the preset power grid warning factor library, and determine the factor type corresponding to the target key information. The power grid warning factor library includes several types of influencing factors of power grid warning;

[0062] In this embodiment, several types of influencing factors of power grid early warning include meteorological factors, inherent attribute factors of equipment, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors. The types of influencing factors are set according to the influencing factors. For example, the power grid early warning factor library is set with meteorological factor types, inherent attribute factor types of equipment, operation factor types, operation and maintenance factor types, geographical environment factor types, social environment factor types and other factor types. For example, when the target key obtained in the previous step includes thunderstorms, typhoons and topography, the key information is matched with the type, and it can be confirmed that the key information involves meteorological factor types and geographical environment factor types. Thunderstorms, typhoons and topography are used as matching identifiers. When the type contains the identifier information, it belongs to this type.

[0063] S14. Calculate the weight of the factor type corresponding to the target key information based on the optimized indicator weight method, and determine the warning result according to the calculated weight.

[0064] In this embodiment, the weight of the factor type corresponding to the target key information is calculated based on the optimized indicator weight method, including the following steps:

[0065] 1. Use the principal weighting method to perform first-order weighting calculations on meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors;

[0066] Suppose the meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors are x1, x2, x3, x4, x5, x6, x7, and select the most important factor from these indicators, denoted as x i , if the index x i The importance of x j When it is high, it is recorded as x i ≥x j , calculate the importance relationship of these indicators, for example, the relationship of indicator importance is: x1≥x2≥x3≥x4≥x5≥x6≥x7, then establish a sequence relationship, and after determining the sequence relationship, perform a quantitative analysis on the importance of each indicator, the expression is as follows:

[0067]

[0068] Where: k is the subjective weight of the kth indicator, γ k It is the importance ratio between two adjacent indicators.

[0069] Based on the γ given by experts k The subjective weight of each indicator is calculated using the rational assignment method.

[0070] 2. Use the objective weighting method to perform second-order weighting calculations on meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors;

[0071] The objective weighting method here is to calculate the information entropy H of each indicator j and entropy weight ω j , where the calculation formulas for information entropy and entropy weight are as follows:

[0072]

[0073] where p j is the characteristic weight of index j.

[0074] 3. Generate the weights of the factor types corresponding to the target key information based on the first-order weighted calculation results and the second-order weighted calculation results.

[0075] The subjective weight obtained by the first-order weighting calculation above is combined with the objective weight calculated by the second-order weighting calculation to obtain the combined weight. The calculation formula is as follows:

[0076]

[0077]

[0078] Where: k is the combined weight of subjective and objective, ω i is the weight obtained by subjective weighting calculation, ω j The weight is calculated by objective weighting.

[0079] In summary, the early warning results are generated according to the calculated combined weights. The early warning results can be set to be visualized at different levels, and the weights of each factor type are different, so the early warning results are displayed in different ways. The early warning results are set according to the weights of each factor type. Finally, according to different early warning results, the system sets up decision support, synchronizes the corresponding decision plans according to different early warning results, and pushes them to managers at the same time as the early warning results, so that managers can perform targeted maintenance.

[0080] Embodiment 2

[0081] See also Figure 2 , an embodiment of a power grid early warning system based on different factor types in the present invention comprises the following steps:

[0082] The raw data acquisition unit 101 is used to collect raw data from various links of the power grid and pre-process the raw data;

[0083] The target key information determination unit 102 is used to extract text information from the preprocessed data using a Transformer model to obtain the target key information;

[0084] The factor type determination unit 103 corresponding to the target key information is used to match the target key information with a preset power grid early warning factor library and determine the factor type corresponding to the target key information. The power grid early warning factor library includes several types of influencing factors of power grid early warning;

[0085] The early warning result determination unit 104 is used to calculate the weight of the factor type corresponding to the target key information based on the optimized indicator weight method, and determine the early warning result according to the calculated weight.

[0086] For the specific definition of the system, please refer to the definition of the method above, which will not be repeated here. Each module in the above system can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0087] Embodiment 3

[0088] See also Figure 3 The present invention provides a computer device, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the above method are implemented.

[0089] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0090] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored. In addition, each functional unit in each embodiment of the present invention can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

[0091] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0092] It can be understood that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A power grid early warning method based on different types of factors, characterized in that: include: Collecting raw data from all links of the power grid and preprocessing the raw data; Use the Transformer model to extract text information from the preprocessed data to obtain the target key information; Matching the target key information with a preset power grid early warning factor library, and determining the factor type corresponding to the target key information, wherein the power grid early warning factor library includes several types of influencing factors of power grid early warning; The weights of the factor types corresponding to the target key information are calculated based on the optimized indicator weight method, and the warning results are determined according to the calculated weights.

2. The power grid early warning method based on different factor types according to claim 1 is characterized in that: The Transformer model is used to extract text information from the preprocessed data to obtain target key information, including: Use the encoding network in the Transformer model to encode the preprocessed data sequence to obtain the representation of entity mentions; Calculate the correlation between the representation mentioned by the entity and the set influencing factor variable; When the calculated correlation meets the preset conditions, the decoding network in the Transformer model is used for decoding to output the target key information.

3. The power grid early warning method based on different factor types according to claim 2 is characterized in that: The encoding network in the Transformer model is used to encode the preprocessed data sequence, and the expression of the representation of the entity mention is obtained as follows: L={l1,l2,...,l N }=BERT(D); D={t1,t2,…,t n ,}; Where: N is the number of words in the preprocessed data sequence, n is the number of sequences after segmentation by the BERT segmenter, L is the entity mention representation, and D is the preprocessed data sequence.

4. The power grid early warning method based on different factor types according to claim 2 is characterized in that: The calculating the correlation between the representation mentioned by the entity and the set influencing factor variable includes: The cosine similarity between the entity mention and the set influencing factor variable is calculated, and the expression is as follows: Where: A and B are entity mention and influencing factor variables respectively, a i and b i are the values ​​of the i-th dimension of A and B respectively.

5. The power grid early warning method based on different factor types according to claim 1 is characterized in that: The types of influencing factors of the several power grid early warnings include meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors.

6. The power grid early warning method based on different factor types according to claim 5 is characterized in that: The calculating the weight of the factor type corresponding to the target key information based on the optimized indicator weight method includes: The first-order weighting calculation is performed on the meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors using the principal weighting method; The objective weighting method is used to perform second-order weighting calculation on the meteorological factors, equipment inherent attribute factors, operation factors, operation and maintenance factors, geographical environment factors, social environment factors and other factors; The weight of the factor type corresponding to the target key information is generated according to the first-order weighted calculation result and the second-order weighted calculation result.

7. The power grid early warning method based on different factor types according to claim 6 is characterized in that: The calculation formula for generating the weight of the factor type corresponding to the target key information according to the first-order weighted calculation result and the second-order weighted calculation result is as follows: Where: k is the combined weight of subjective and objective, ω i is the weight obtained by subjective weighting calculation, ω j The weight is calculated by objective weighting.

8. A power grid early warning system based on different types of factors, characterized in that: include: The raw data acquisition unit is used to collect the raw data of each link of the power grid and pre-process the raw data; A target key information determination unit is used to extract text information from the preprocessed data using a Transformer model to obtain target key information; A factor type determination unit corresponding to the target key information, used to match the target key information with a preset power grid early warning factor library and determine the factor type corresponding to the target key information, wherein the power grid early warning factor library includes several types of influencing factors of power grid early warning; The early warning result determination unit is used to calculate the weight of the factor type corresponding to the target key information based on the optimized indicator weight method, and determine the early warning result according to the calculated weight.

9. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, characterized in that: When the processor executes the computer-readable instructions, the steps of the power grid early warning method based on different factor types as described in any one of claims 1 to 7 are implemented.

10. A readable storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by a processor, the steps of the power grid early warning method based on different factor types as described in any one of claims 1 to 7 are implemented.

Images