A smart grid fault location system

Through the data acquisition and analysis module of the smart grid fault positioning system, combined with dimensionality reduction processing and fault type coefficient calculation, accurate fault type prediction of abnormal data in the power fault area is achieved, and the inaccurate fault type prediction in the existing technology is solved. The problem has improved maintenance efficiency and economicality.

CN118735230BActive Publication Date: 2024-11-01HUNAN ELECTRICAL COLLEGE OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411226636.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-03
Publication Date
2024-11-01
Estimated Expiration
2044-09-03

AI Technical Summary

Technical Problem

The existing grid fault positioning algorithm cannot accurately predict the fault type, especially when abnormal data exists in the power fault area, resulting in uneconomical positioning errors and maintenance.

Method used

A smart grid fault positioning system was designed, including data acquisition module, fault positioning module and personnel configuration module. Fault characteristic data is collected through IoT devices, area judgment, location judgment and fault type analysis are carried out, and dimensionality reduction processing and identification are used to identify and judge. Criteria functions and fault type coefficient calculations to achieve accurate fault type prediction of abnormal data and provide staffing suggestions.

Benefits of technology

Improve the accuracy of fault type prediction, reduce maintenance time and losses, and improve the maintenance efficiency and economicality of the power system by matching maintenance personnel and fault type.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118735230B_ABST
    Figure CN118735230B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of power grid management, and specifically discloses an intelligent power grid fault location system, including a data acquisition module, a fault location module, and a fault analysis module. The data acquisition module acquires power grid fault feature data through an Internet of Things device and uploads the data to the fault location module; the fault location module sequentially performs regional judgment and position judgment on the fault according to the fault feature data acquired by the acquisition module; the fault analysis module analyzes the fault that has completed regional judgment and position judgment, and determines its fault type. Based on the data of regional judgment and position judgment, the present invention makes a more accurate prediction of the fault type of the fault location where abnormal data appears in the power fault area, and provides more information for power maintenance. The matching of maintenance personnel and fault type can effectively reduce the maintenance time, and minimize and recover the loss caused by the power failure.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power grid management, and in particular to a smart grid fault location system. Background Art

[0002] Grid fault location has a strong guiding role in grid maintenance and can reduce the losses caused by grid faults. Existing fault location algorithms can only adapt to small data samples and are highly dependent on data. If there are abnormal data in the data, it will cause large positioning errors. The common method is to process the abnormal data to improve the accuracy of positioning prediction. Obviously, the current positioning algorithm does not have the function of predicting the type of fault after positioning, and the existing fault type prediction algorithm usually predicts the main body of the fault area, while the abnormal faults in the main body of the corresponding fault area are difficult to predict. Predicting each fault location separately is a huge waste of computing power and is obviously not economical.

[0003] The benefits of fault type prediction are obvious. It can help power units select appropriate personnel for maintenance to reduce maintenance time and further reduce power failure losses. In view of this, the present invention proposes an intelligent grid fault location system that provides more accurate fault type prediction based on power fault location. Summary of the invention

[0004] The purpose of the present invention is to provide a smart grid fault location system to solve the following technical problems:

[0005] The problem is how to provide a more accurate prediction of fault type based on power fault location.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A smart grid fault location system, comprising:

[0008] A data acquisition module, a fault location module, a fault analysis module and a personnel configuration module. The data acquisition module collects power grid fault feature data through an Internet of Things device and uploads it to the fault location module;

[0009] The fault location module sequentially determines the area and position of the fault based on the fault feature data collected by the acquisition module;

[0010] The fault analysis module analyzes the faults that have completed area judgment and location judgment to determine the fault type;

[0011] The personnel configuration module provides personnel configuration suggestions for the maintenance process according to the fault type and personnel status.

[0012] Through the above technical scheme: a scheme is proposed to make regional judgment and position judgment on the fault, and further judge its fault type based on the judgment process of the two. The present invention is based on the data basis of regional judgment and position judgment, and makes a more accurate prediction of the fault type of the fault location where abnormal data appears in the power fault area, providing more information for power maintenance. The matching of maintenance personnel and fault type can effectively reduce the maintenance time, minimize and recover the losses caused by power failures.

[0013] As a further technical solution of the present invention: the process of sequentially determining the area and the position of the fault includes:

[0014] Perform dimensionality reduction processing on the collected fault feature data to obtain low-dimensional data;

[0015] Constructing a fault region identification and discrimination criterion function based on low-dimensional data and performing region judgment to obtain region judgment results;

[0016] Based on low-dimensional data, a fault location identification criterion function is constructed, and then the residual difference is obtained and constraints are set based on the residual difference. The objective function is obtained in combination with the constraints, and the location is judged by the objective function to obtain the location judgment result.

[0017] As a further technical solution of the present invention: the process of determining the fault type includes:

[0018] Among the fault characteristic data uploaded by all fault locations, the corresponding data uploaded by the fault location determined to be abnormal data is removed to obtain the second characteristic data;

[0019] Selecting characteristic data with high correlation with the fault type from the second characteristic data as target data;

[0020] Standardize each classification in the target data to obtain the corresponding standard value and set the weight value, and then calculate the fault type coefficient based on the standard value and the weight value;

[0021] The basic fault type corresponding to the fault location of the normal data is determined based on the fault type coefficient, and then the fault type corresponding to the fault location of the abnormal data is determined based on the basic fault type.

[0022] As a further technical solution of the present invention: the process of calculating and obtaining the fault type coefficient includes:

[0023] By formula:

[0024] in, is the fault type coefficient, is the number of normal data in the actual monitoring results of the fault area, is the value of the hth data in the target data, is the standard value preset based on the data type of the hth data, It is a weight value preset based on the data type of the hth data.

[0025] Through the above technical scheme: a process of performing area judgment and position judgment, and a process of judging the fault type based on area judgment and position judgment are provided. Specifically, the present invention first constructs a fault area identification judgment criterion function and a fault location identification judgment criterion function, and then constructs constraints and combines the constraints with the two to obtain a target function. Based on the target function, a more accurate judgment can be made on the fault location of the area where the abnormal data is located, and there is a high probability that the problem of abnormal data appearing at the location is because the corresponding fault type is different from the main fault type of the area where it is located, resulting in a large amount of deviation in the data, thereby distinguishing different fault types in a fault area.

[0026] As a further technical solution of the present invention: the process of determining the basic fault type based on the fault type coefficient includes:

[0027] Substitute the fault type coefficient into the preset selection function, and the selection function outputs the basic fault type;

[0028] The number of fault types output by the selection function is multiple and is accompanied by distribution probabilities based on sample acquisition.

[0029] As a further technical solution of the present invention: the process of determining the fault type corresponding to the fault location of the abnormal data includes:

[0030] Based on the fault feature data corresponding to the fault position of the abnormal data, a classification vector is constructed using the value of the actual fault position, the value of the residual difference corresponding to the fault position, and the fault type coefficient;

[0031] The classification vectors composed of abnormal data of the samples are divided into several identification groups based on the fault type, and a classification vector is selected from each identification group as the central vector;

[0032] The vector distance between the actually obtained classification vector and all the center vectors is calculated, and the group of the center vector corresponding to the minimum vector distance is selected as the target group, and the predicted type is obtained based on the fault type of the target group.

[0033] As a further technical solution of the present invention: the process of obtaining the predicted type based on the fault type of the target group includes:

[0034] Setting the fault type of the target group to the first prediction type;

[0035] Compare the first prediction type with the multiple fault types output by the selection function, and if the multiple fault types output by the selection function include the first prediction type, output the first prediction type as the target fault type corresponding to the fault position of the abnormal data;

[0036] If the multiple fault types output by the selection function do not include the first predicted type, based on the basic fault type, the fault type with the highest concurrency probability is selected from the concurrency statistical set as the target fault type corresponding to the fault position of the abnormal data.

[0037] Through the above technical scheme: a process for predicting the fault type of different fault locations in the fault area is provided. The present invention is based on the abnormal data in the fault location judgment process, and predicts the fault type by constructing a classification vector. In particular, the fault types in the power system have concurrent characteristics. Therefore, the difficult-to-predict fault types are distinguished by the comparison method, and the concurrent probability is used as the target fault type to further improve the accuracy of the prediction.

[0038] As a further technical solution of the present invention: the process of providing personnel configuration suggestions for the maintenance process according to the fault type and the personnel status includes:

[0039] Based on the evaluation of each personnel's maintenance performance on different fault types, obtain the fault type score corresponding to each personnel;

[0040] For on-the-job personnel, the recommendation coefficient is obtained based on the fault type, number of faults and fault type score in the area;

[0041] The person with the largest recommendation coefficient is selected as the recommendation target for the fault area and pushed.

[0042] As a further technical solution of the present invention: the process of obtaining the recommendation coefficient includes:

[0043] For a certain fault area, the formula is:

[0044] Get the recommendation factor ,in is the number of basic fault types, is the fault type score of the basic fault type of the person, is the target fault type corresponding to the fault location corresponding to the k abnormal data. Fault type scores for each target fault type.

[0045] Through the above technical solution: a solution for recommending employees to carry out maintenance is provided. Specifically, the employees' ability to handle different types of faults is scored. Based on different fault types and quantities, the employee's recommendation coefficient for a certain fault area is obtained. The larger the recommendation coefficient, the less time the employee will spend in this area while ensuring quality.

[0046] Beneficial effects of the present invention:

[0047] (1) Based on the data of area judgment and location judgment, the present invention makes a more accurate prediction of the fault type of the fault location where abnormal data appears in the power fault area, providing more information for power maintenance. The matching of maintenance personnel and fault type can effectively reduce the maintenance time and minimize and recover the losses caused by power failures.

[0048] (2) The present invention first constructs a fault area identification and discrimination criterion function and a fault location identification and discrimination criterion function, and then constructs constraints and combines the constraints with the two to obtain an objective function. Based on the objective function, a more accurate judgment can be made on the fault location in the area where the abnormal data is located. The problem of abnormal data appearing at the location is likely to be because the corresponding fault type is different from the main fault type in the area, resulting in a large amount of data deviation. Therefore, different fault types in a fault area can be distinguished.

[0049] (3) The present invention is based on the abnormal data in the fault location judgment process, and predicts the fault type by constructing a classification vector. In particular, the fault types in the power system have concurrent characteristics. Therefore, the difficult-to-predict fault types are distinguished by the comparison method, and the concurrent probability is used as the target fault type to further improve the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present invention will be further described below in conjunction with the accompanying drawings.

[0051] Figure 1 It is a schematic diagram of the module composition of the present invention;

[0052] Figure 2 It is a schematic diagram of the functional composition of the fault location module of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] See also Figure 1 and Figure 2 As shown, in one embodiment, a smart grid fault location system is provided, including: a data acquisition module, a fault location module, a fault analysis module and a personnel configuration module, the data acquisition module collects grid fault feature data through an Internet of Things device and uploads it to the fault location module; the fault location module performs regional judgment and position judgment on the fault in turn according to the fault feature data collected by the acquisition module; the fault analysis module analyzes the fault that has completed regional judgment and position judgment, and determines its fault type; the personnel configuration module provides personnel configuration suggestions for the maintenance process according to the fault type and personnel status.

[0055] In this embodiment, a solution is proposed to perform regional judgment and location judgment on the fault, and further judge the fault type based on the judgment process of the two. Based on the data of regional judgment and location judgment, the present invention makes a more accurate prediction of the fault type of the fault location where abnormal data appears in the power fault area, and provides more information for power maintenance. The matching of maintenance personnel and fault type can effectively reduce the maintenance time and minimize and recover the losses caused by power failures.

[0056] The process of sequentially determining the area and the location of the fault includes:

[0057] The collected fault feature data is processed by dimensionality reduction to obtain low-dimensional data. Assuming that Faults, each fault occurs dimensional data, constituting the original Data Collection ;

[0058] Pair Collection Perform dimensionality reduction to obtain a new low-dimensional data set Y, wherein the dimensionality reduction method is preferably a data screening method based on manifold learning;

[0059] Based on low-dimensional data, a fault region identification and discrimination criterion function is constructed and region judgment is performed to obtain a region judgment result. In this embodiment, when faults occur in the same region, they can be regarded as the same type of faults. Based on this, the fault region identification and discrimination criterion function is:

[0060] in, is the fault area mapping function based on each sample point, is the intra-class distribution matrix obtained after the fault types in the fault area are regarded as the same type of faults, is the inter-class difference distribution matrix, yes The transposed result of is the fault area identification criterion function, the result of which is the minimum value selected from multiple functions based on the matrix operation results;

[0061] A fault location identification criterion function is constructed based on low-dimensional data, and then the residual difference is obtained and constraints are set based on the residual difference. The objective function is obtained in combination with the constraints, and then the location is judged by the objective function to obtain the location judgment result. The residual difference is the difference between the predicted value and the normal fault distance. Preferably, in this embodiment, the constraints are defined as:

[0062] in, is the residual error, The first occurrence in the sample The actual fault location of the secondary fault, is the fault location calculated after multi-scale mapping. It is the preset standard value of residual error. The specific calculation process belongs to the existing technology and will not be described in detail. The purpose of the constraint condition is to make the residual error grow only linearly, suppress the loss increase, and reduce the impact of abnormal data on the accuracy of the model. It is to sum the contents after the summation symbol from 1 to k;

[0063] As an example, the objective function obtained in this embodiment in combination with the constraint conditions is:

[0064] in, , are the number of normal data and abnormal data in the actual monitoring results of the fault area, respectively. , The data is normal data. The data is abnormal data. Similarly, is the objective function to be sought, It is to sum the contents after the summation symbol from 1 to k. It is to sum the contents after the summation symbol from 1 to d.

[0065] The process of determining the fault type includes:

[0066] Among the fault characteristic data uploaded by all fault locations, the corresponding data uploaded by the fault location determined to be abnormal data is removed to obtain the second characteristic data, that is, the corresponding data uploaded by the fault location determined to be normal data;

[0067] Selecting characteristic data with high correlation with the fault type from the second characteristic data as target data;

[0068] Standardize each classification in the target data to obtain the corresponding standard value and set the weight value, and then calculate the fault type coefficient based on the standard value and the weight value;

[0069] The basic fault type corresponding to the fault location of the normal data is determined based on the fault type coefficient, and then the fault type corresponding to the fault location of the abnormal data is determined based on the basic fault type.

[0070] The process of calculating the fault type coefficient includes:

[0071] By formula:

[0072] in, is the fault type coefficient, is the number of normal data in the actual monitoring results of the fault area, is the value of the hth data in the target data, is the standard value preset based on the data type of the hth data, is the preset weight value based on the data type of the hth data. It is to sum the contents after the summation symbol from 1 to d.

[0073] In this embodiment, a process for performing area judgment and position judgment, as well as a process for judging the fault type based on area judgment and position judgment are provided. Specifically, the present invention first constructs a fault area identification judgment criterion function and a fault location identification judgment criterion function, and then constructs constraints and combines the constraints with the two to obtain an objective function. Based on the objective function, a more accurate judgment can be made on the fault location of the area where the abnormal data is located. In addition, there is a high probability that the problem of abnormal data appearing at the location is because the corresponding fault type is different from the main fault type of the area where it is located, resulting in a large amount of deviation in the data. Therefore, different fault types in a fault area can be distinguished.

[0074] The process of determining the basic fault type based on the fault type coefficient includes:

[0075] Substitute the fault type coefficient into the preset selection function, and the selection function outputs the basic fault type;

[0076] The number of fault types output by the selection function is multiple and is accompanied by distribution probabilities based on sample acquisition.

[0077] The process of determining the fault type corresponding to the fault location of abnormal data includes:

[0078] Based on the fault characteristic data corresponding to the fault location of the abnormal data, the first The classification vector is constructed by using the value of the actual fault position of the secondary fault, the value of the residual difference corresponding to the fault position, and the fault type coefficient;

[0079] The classification vectors composed of abnormal data of the samples are divided into several identification groups based on the fault type, and a classification vector is selected from each identification group as the central vector;

[0080] The vector distance between the actually obtained classification vector and all the center vectors is calculated, and the group of the center vector corresponding to the minimum vector distance is selected as the target group, and the predicted type is obtained based on the fault type of the target group.

[0081] The process of obtaining the predicted type based on the fault type of the target group includes:

[0082] Setting the fault type of the target group to the first prediction type;

[0083] Compare the first prediction type with the multiple fault types output by the selection function, and if the multiple fault types output by the selection function include the first prediction type, output the first prediction type as the target fault type corresponding to the fault position of the abnormal data;

[0084] If the multiple fault types output by the selection function do not include the first predicted type, based on the basic fault type, the fault type with the highest concurrency probability is selected from the concurrency statistical set as the target fault type corresponding to the fault position of the abnormal data.

[0085] In this embodiment, a process for predicting fault types at different fault locations in a fault area is provided. The present invention is based on abnormal data in the fault location judgment process and predicts the fault type by constructing a classification vector. In particular, the fault types in the power system have concurrent characteristics. Therefore, the difficult-to-predict fault types are distinguished by a comparison method, and the concurrent probability is used as the target fault type to further improve the accuracy of the prediction.

[0086] The process of providing personnel allocation recommendations for the maintenance process based on the fault type and personnel status includes:

[0087] Based on the evaluation of each personnel's maintenance performance on different fault types, the fault type score corresponding to each personnel is obtained. There is no restriction on the specific evaluation and scoring process. The percentage system is preferably adopted, that is, the more skilled the operation, the higher the score;

[0088] For on-the-job personnel, the recommendation coefficient is obtained based on the fault type, number of faults and fault type score in the area. On-the-job personnel refer to the on-the-job workers who complete the process of evaluating the maintenance performance of different fault types;

[0089] The person with the largest recommendation coefficient is selected as the recommendation target for the fault area and pushed.

[0090] The process of obtaining the recommendation coefficient includes:

[0091] For a certain fault area, the formula is:

[0092] Get the recommendation factor ,in is the number of basic fault types, is the fault type score of the basic fault type of the person, is the target fault type corresponding to the fault location corresponding to the k abnormal data. Fault type scores for target fault types, It is to sum the contents after the summation symbol from 1 to k.

[0093] In this embodiment, a scheme for recommending employees to perform maintenance is provided. Specifically, the employees' ability to handle different types of faults is scored. Based on different fault types and quantities, a recommendation coefficient for the employee for a certain fault area is obtained. The larger the recommendation coefficient, the less time the employee will spend in this area while ensuring quality.

[0094] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A smart grid fault location system, comprising a data acquisition module, a fault location module, a fault analysis module and a personnel configuration module, Features: The data acquisition module collects power grid fault characteristic data through the Internet of Things device and uploads it to the fault location module; The fault location module sequentially determines the area and position of the fault based on the fault feature data collected by the acquisition module; The process of determining the area and location of the fault in sequence includes: Perform dimensionality reduction processing on the collected fault feature data to obtain low-dimensional data; Constructing a fault region identification and discrimination criterion function based on low-dimensional data and performing region judgment to obtain region judgment results; Based on low-dimensional data, a fault location identification criterion function is constructed, and then the residual difference is obtained and constraints are set based on the residual difference. The objective function is obtained by combining the constraints, and the location is judged by the objective function to obtain the location judgment result. The fault analysis module analyzes the faults that have completed area judgment and location judgment to determine the fault type; The process of determining the fault type includes: Among the fault characteristic data uploaded by all fault locations, the corresponding data uploaded by the fault location determined to be abnormal data is removed to obtain the second characteristic data; Selecting characteristic data with high correlation with the fault type from the second characteristic data as target data; Standardize each classification in the target data to obtain the corresponding standard value and set the weight value, and then calculate the fault type coefficient based on the standard value and the weight value; Determine the basic fault type corresponding to the fault position of the normal data based on the fault type coefficient, and then determine the fault type corresponding to the fault position of the abnormal data based on the basic fault type; The personnel configuration module provides personnel configuration suggestions for the maintenance process according to the fault type and personnel status.

2. A smart grid fault location system according to claim 1, It is characterized in that The process of calculating the fault type coefficient includes: By formula: in, is the fault type coefficient, is the number of normal data in the actual monitoring results of the fault area, is the value of the hth data in the target data, is the standard value preset based on the data type of the hth data, It is a weight value preset based on the data type of the hth data.

3. A smart grid fault location system according to claim 1, It is characterized in that The process of determining the basic fault type based on the fault type coefficient includes: Substitute the fault type coefficient into the preset selection function, and the selection function outputs the basic fault type; The number of fault types output by the selection function is multiple and is accompanied by distribution probabilities based on sample acquisition.

4. A smart grid fault location system according to claim 1, It is characterized in that The process of determining the fault type corresponding to the fault location of abnormal data includes: Based on the fault feature data corresponding to the fault position of the abnormal data, a classification vector is constructed using the value of the actual fault position, the value of the residual difference corresponding to the fault position, and the fault type coefficient; The classification vectors composed of abnormal data of the samples are divided into several identification groups based on the fault type, and a classification vector is selected from each identification group as the central vector; The vector distance between the actually obtained classification vector and all the center vectors is calculated, and the group of the center vector corresponding to the minimum vector distance is selected as the target group, and the predicted type is obtained based on the fault type of the target group.

5. A smart grid fault location system according to claim 4, It is characterized in that The process of obtaining the predicted type based on the fault type of the target group includes: Setting the fault type of the target group to the first prediction type; Compare the first prediction type with the multiple fault types output by the selection function, and if the multiple fault types output by the selection function include the first prediction type, output the first prediction type as the target fault type corresponding to the fault position of the abnormal data; If the multiple fault types output by the selection function do not include the first predicted type, based on the basic fault type, the fault type with the highest concurrency probability is selected from the concurrency statistical set as the target fault type corresponding to the fault position of the abnormal data.

6. A smart grid fault location system according to claim 1, It is characterized in that The process of providing personnel allocation recommendations for the maintenance process based on the fault type and personnel status includes: Based on the evaluation of each personnel's maintenance performance on different fault types, obtain the fault type score corresponding to each personnel; For on-the-job personnel, the recommendation coefficient is obtained based on the fault type, number of faults and fault type score in the area; The person with the largest recommendation coefficient is selected as the recommendation target for the fault area and pushed.

7. A smart grid fault location system according to claim 6, It is characterized in that The process of obtaining the recommendation coefficient includes: For a certain fault area, the formula is: Get the recommendation factor ,in is the number of basic fault types, is the fault type score of the basic fault type of the person, is the target fault type corresponding to the fault location corresponding to the k abnormal data. Fault type scores for each target fault type.