Power distribution network intelligent early warning and section positioning method based on transient recording data

Abstract

The invention discloses a power distribution network intelligent early warning and section positioning method based on transient recording data, and the method comprises the steps: carrying out the fault type recognition through the comprehensive analysis of the transient recording data, combining with an advanced algorithm, such as a relevance vector machine (RVM), carrying out the line abnormality recognition through a deep neural network, and accurately extracting the feature information of fault data, the limitation of an existing power distribution automation master station fault studying and judging method is effectively overcome, the problem of inaccurate fault studying and judging caused by the data problem of the intelligent terminal is avoided, the accuracy of fault studying and judging of the power distribution automation master station is integrally improved to a relatively high level, a reliable basis can be provided for fault processing, and the working efficiency is improved. Multi-source data such as a wave recording file, a topological structure and remote signaling and telemetering are fully fused, the operation state of the power distribution network is analyzed from multiple angles, compared with an analysis method of a single data source, the actual situation of the power distribution network can be reflected more comprehensively and accurately, and the reliability and accuracy of fault diagnosis and positioning are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent distribution network automation, and specifically provides a distribution network intelligent early warning and section positioning method based on transient recording data. Background Art

[0002] As a vital component of the power grid, distribution networks play a crucial role in safeguarding and improving people's livelihoods. Current domestic distribution lines are characterized by large scale, complex structure, numerous types and quantities of equipment, and complex operating environments. With economic and social development, people's demands for power supply reliability are increasing. However, 10kV distribution network failures account for over 80% of all grid failures. Efficient operation and maintenance will help promptly detect potential risks and reduce the number of distribution network failures.

[0003] At present, the fault analysis method of the distribution automation master station has limitations and fails to fully analyze the fault recording data uploaded by the smart terminal. At the same time, the distribution smart terminals are numerous and widespread, and there are differences in the success rate of on-site analysis and data upload, resulting in a low accuracy rate of fault analysis. Traditional fault location methods are difficult to quickly and accurately locate the fault section when faced with complex distribution network structures and various fault types, and cannot meet the growing demand for power supply reliability. Moreover, the existing technology has limited means for evaluating and identifying the causes of abnormal operation of medium and low voltage lines. It cannot detect potential problems in time before a fault occurs, and cannot achieve effective pre-emptive control, resulting in inefficient fault handling and long power outages, which brings inconvenience to users and causes significant economic losses to power companies.

[0004] Therefore, it is necessary to invent a distribution network intelligent early warning and section positioning method based on transient recording data to solve the above problems. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a distribution network intelligent early warning and section positioning method based on transient recording data to solve the problems raised in the above background technology of inaccurate fault analysis and judgment of the existing distribution automation master station, low fault positioning efficiency, and inability to achieve pre-emptive control.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a distribution network intelligent early warning and section positioning method based on transient recording data, comprising the following steps:

[0007] Step 1: Fault data acquisition and processing: Analyze the types of common faults on medium and low voltage lines, obtain waveform files, topology structures, telesignaling and telemetry data, and pre-process the acquired data to remove noise and outliers, unify the data format, and provide an accurate and effective data foundation for subsequent research.

[0008] Step 2: Evaluation of medium and low voltage line operation anomalies and identification of causes: Conduct in-depth analysis of pre-processed fault data, extract characteristic parameters of the fault data, and use support vector machine-related algorithms to perform feature selection and classification judgment on the characteristic parameters. By detecting and classifying the characteristic quantities of the input waveform data, it can automatically identify distribution network line anomalies caused by different reasons, evaluate the line operation status, and identify potential abnormal phenomena and the types of inducements.

[0009] Step 3: Line anomaly identification based on transient recording data. Utilizing the characteristic information in the transient recording data, a deep neural network-based method is used to automatically identify medium and low voltage line anomalies. This involves using a vector machine to classify and obtain the signal modal function. By describing the relationship between signal sequences, the optimal characteristic component values ​​are obtained. A deep neural network anomaly identification model is established. Convolution operations are used to learn associated features in different spaces. Long-short-term memory networks are used to process time series data, which is then transferred to the model to identify line anomalies.

[0010] Step 4: Based on the transient recording data, the line section is located. Combined with the line anomaly identification results, the fault location matrix is ​​used to achieve precise section location, quickly and accurately determining the specific section or equipment where the fault is located.

[0011] As a further description of the above technical solution, the method of obtaining the recording file described in step one includes collecting transient recording data when a fault occurs in real time through an intelligent terminal in the distribution network, and uploading the recording file to the distribution automation master station.

[0012] As a further description of the above technical solution, when preprocessing the data in step one, a filtering algorithm is used to remove noise, a data cleaning algorithm is used to remove outliers, and a data normalization algorithm is used to unify the data format.

[0013] As a further description of the above technical solution, the extraction of characteristic parameters of the fault data in step 2 includes extracting time domain characteristic parameters and frequency domain characteristic parameters of the fault waveform.

[0014] As a further description of the above technical solution, the characteristic parameters described in step 2 include the fault occurrence time, fault duration, fault current peak and fault voltage peak, and the frequency domain characteristic parameters include the fault frequency component and frequency amplitude distribution.

[0015] As a further description of the above technical solution, when using the support vector machine related algorithm to perform feature selection and classification judgment on feature parameters as described in step 2, a waveform data labeling and classification method based on the relevance vector machine (RVM) is adopted to establish models and theoretical analysis of different types of fault waveforms.

[0016] As a further description of the above technical solution, in step three, before establishing the deep neural network anomaly identification model, it also includes extracting fault time-frequency features of transient recording data based on wavelet transform to enhance the feature expression ability of the data.

[0017] As a further description of the above technical solution, when using the long short-term memory network to process time series data in step three, the long-term dependency of the data is learned through the memory and forgetting mechanism of the memory unit on historical data.

[0018] As a further description of the above technical solution, when the fault location matrix is ​​used to achieve precise section location in step 4, a multi-criteria fusion method based on fuzzy c-means clustering is combined to comprehensively consider various fault feature information to improve the accuracy of location.

[0019] As a further description of the above technical solution, it also includes the step of establishing a fault waveform library. During the analysis and processing of fault data, the waveform data of different types of faults and their corresponding characteristic parameters and fault type labels are stored in the fault waveform library to provide sample references for subsequent abnormality assessment, cause identification, abnormality recognition and section positioning.

[0020] As a further description of the above technical solution, when conducting abnormal operation assessment and cause identification of medium and low voltage lines, line abnormality identification and section positioning based on transient recording data, the real-time collected fault data is matched with samples in the fault waveform library, and the fault type and abnormal area are determined by calculating the sample test value.

[0021] As a further description of the above technical solution, when it is determined that there is an abnormality or fault in the line, an early warning message is generated and sent to the terminal device of the operation and maintenance personnel. At the same time, the abnormal or faulty section is marked on a map of the entire distribution automation master station.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] 1. The present invention comprehensively analyzes transient recording data, combines advanced algorithms such as relevance vector machines (RVM) to identify fault types, and uses deep neural networks to identify line anomalies. It can accurately extract characteristic information of fault data, effectively overcome the limitations of existing distribution automation master station fault analysis methods, avoid inaccurate fault analysis caused by smart terminal data problems, and improve the overall accuracy of distribution automation master station fault analysis to a higher level. It can provide a reliable basis for fault handling, fully integrate multi-source data such as recording files, topology structures, telesignaling and telemetry, and analyze the operating status of the distribution network from multiple angles. Compared with the analysis method of a single data source, it can more comprehensively and accurately reflect the actual situation of the distribution network, and improve the reliability and accuracy of fault diagnosis and positioning.

[0024] 2. The present invention is based on transient recording data and adopts a multi-criteria fusion method of fault location matrix combined with fuzzy c-means clustering, which can quickly and accurately locate the specific section or equipment where the fault is located. Compared with traditional positioning methods, it greatly shortens the fault location time and improves the positioning accuracy, so that operation and maintenance personnel can quickly reach the fault site for processing, reduce power outage time, reduce the inconvenience caused to users by power outages and reduce the economic losses of power companies. Through the abnormal operation assessment and inducement identification technology of medium and low voltage lines, it can timely discover potential abnormal phenomena and inducements of the line before the fault occurs, and realize advance warning of the fault. Operation and maintenance personnel can take measures in advance according to the early warning information to deal with equipment hidden dangers and defects, nip the fault in the bud, improve the operational reliability and stability of the distribution network, and reduce the number of faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0026] Figure 1 The present invention provides a flow chart of a distribution network intelligent early warning and section positioning method based on transient recording data. DETAILED DESCRIPTION

[0027] The following describes the implementation of the present invention using specific embodiments. Those skilled in the art will readily understand the other advantages and benefits of the present invention from the disclosure herein. Obviously, the embodiments described are only a portion of the present invention, not all of it. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are intended to fall within the scope of protection of the present invention.

[0028] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0029] Refer to the attached Figure 1 The distribution network intelligent early warning and section positioning method based on transient recording data of this embodiment includes the following steps:

[0030] Step 1: Fault data acquisition and processing: Analyze the types of common faults on medium and low voltage lines, obtain waveform files, topology structures, telesignaling and telemetry data, and pre-process the acquired data to remove noise and outliers, unify the data format, and provide an accurate and effective data foundation for subsequent research.

[0031] Step 2: Evaluation of medium and low voltage line operation anomalies and identification of causes: Conduct in-depth analysis of pre-processed fault data, extract characteristic parameters of the fault data, and use support vector machine-related algorithms to perform feature selection and classification judgment on the characteristic parameters. By detecting and classifying the characteristic quantities of the input waveform data, it can automatically identify distribution network line anomalies caused by different reasons, evaluate the line operation status, and identify potential abnormal phenomena and the types of inducements.

[0032] Step 3: Line anomaly identification based on transient recording data. Utilizing the characteristic information in the transient recording data, a deep neural network-based method is used to automatically identify medium and low voltage line anomalies. This involves using a vector machine to classify and obtain the signal modal function. By describing the relationship between signal sequences, the optimal characteristic component values ​​are obtained. A deep neural network anomaly identification model is established. Convolution operations are used to learn associated features in different spaces. Long-short-term memory networks are used to process time series data, which is then transferred to the model to identify line anomalies.

[0033] Step 4: Based on the transient recording data, the line section is located. Combined with the line anomaly identification results, the fault location matrix is ​​used to achieve precise section location, quickly and accurately determining the specific section or equipment where the fault is located.

[0034] It also includes the step of establishing a fault waveform library. During the analysis and processing of fault data, the waveform data of different types of faults and their corresponding characteristic parameters and fault type labels are stored in the fault waveform library to provide sample references for subsequent abnormality assessment, cause identification, abnormality recognition and section positioning.

[0035] The method of obtaining the recording file in step one includes collecting transient recording data when a fault occurs in real time through the intelligent terminal in the distribution network, and uploading the recording file to the distribution automation master station. When preprocessing the data in step one, a filtering algorithm is used to remove noise, a data cleaning algorithm is used to remove outliers, and a data normalization algorithm is used to unify the data format.

[0036] Extracting characteristic parameters of fault data in step 2 includes extracting time domain characteristic parameters and frequency domain characteristic parameters of the fault waveform. The characteristic parameters in step 2 include fault occurrence time, fault duration, fault current peak value and fault voltage peak value. The frequency domain characteristic parameters include fault frequency component and frequency amplitude distribution. When using support vector machine related algorithm to perform feature selection and classification judgment on characteristic parameters in step 2, a waveform data labeling and classification method based on relevance vector machine (RVM) is adopted, and models and theoretical analysis are established for different types of fault waveforms.

[0037] In step three, before establishing the deep neural network anomaly identification model, it also includes extracting the fault time-frequency characteristics of the transient recording data based on wavelet transform to enhance the feature expression ability of the data. In step three, when using the long short-term memory network to process time series data, the long-term dependency of the data is learned through the memory and forgetting mechanism of the memory unit on historical data.

[0038] In step 4, when using the fault location matrix to achieve precise section location, the multi-criteria fusion method based on fuzzy c-means clustering is combined to comprehensively consider various fault feature information to improve the accuracy of location.

[0039] When evaluating the abnormal operation of medium and low voltage lines and identifying their causes, as well as identifying line anomalies and locating sections based on transient waveform data, the fault data collected in real time is matched with samples in the fault waveform library, and the fault type and abnormal area are determined by calculating the sample test value. When it is determined that the line has an abnormality or fault, an early warning message is generated and sent to the terminal equipment of the operation and maintenance personnel. At the same time, the abnormal or faulty section is marked on a map of the entire distribution automation master station.

[0040] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A distribution network intelligent early warning and section positioning method based on transient recording data is characterized by: The following steps are involved: Step 1: Fault data acquisition and processing: Analyze the types of common faults on medium and low voltage lines, obtain waveform files, topology structures, telesignaling and telemetry data, and pre-process the acquired data to remove noise and outliers, unify the data format, and provide an accurate and effective data foundation for subsequent research. Step 2: Evaluation of medium and low voltage line operation anomalies and identification of causes: Conduct in-depth analysis of pre-processed fault data, extract characteristic parameters of the fault data, and use support vector machine-related algorithms to perform feature selection and classification judgment on the characteristic parameters. By detecting and classifying the characteristic quantities of the input waveform data, it can automatically identify distribution network line anomalies caused by different reasons, evaluate the line operation status, and identify potential abnormal phenomena and the types of inducements. Step 3: Line anomaly identification based on transient recording data. Utilizing the characteristic information in the transient recording data, a deep neural network-based method is used to automatically identify medium and low voltage line anomalies. This involves using a vector machine to classify and obtain the signal modal function. By describing the relationship between signal sequences, the optimal characteristic component values ​​are obtained. A deep neural network anomaly identification model is established. Convolution operations are used to learn associated features in different spaces. Long-short-term memory networks are used to process time series data, which is then transferred to the model to identify line anomalies. Step 4: Based on the transient recording data, the line section is located. Combined with the line anomaly identification results, the fault location matrix is ​​used to achieve precise section location, quickly and accurately determining the specific section or equipment where the fault is located.

2. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 1 is characterized in that: The method of obtaining the recording file described in step 1 includes collecting transient recording data when a fault occurs in real time through an intelligent terminal in the distribution network, and uploading the recording file to the distribution automation master station.

3. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 2 is characterized in that: When preprocessing the data as described in step 1, a filtering algorithm is used to remove noise, a data cleaning algorithm is used to remove outliers, and a data normalization algorithm is used to unify the data format.

4. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 3 is characterized in that: The extraction of characteristic parameters of fault data in step 2 includes extracting time domain characteristic parameters and frequency domain characteristic parameters of the fault waveform.

5. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 4 is characterized in that: The characteristic parameters described in step 2 include the fault occurrence time, fault duration, fault current peak value and fault voltage peak value, and the frequency domain characteristic parameters include the fault frequency component and frequency amplitude distribution.

6. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 5 is characterized in that: When using the support vector machine related algorithm to perform feature selection and classification judgment on feature parameters as described in step 2, a waveform data labeling and classification method based on relevance vector machine (RVM) is adopted, and models and theoretical analysis are established for different types of fault waveforms.

7. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 6 is characterized in that: In the step three, before establishing the deep neural network anomaly identification model, it also includes extracting the fault time-frequency feature of the transient recording data based on wavelet transform to enhance the feature expression ability of the data.

8. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 7 is characterized in that: In the step three, when the long short-term memory network is used to process time series data, the long-term dependency of the data is learned through the memory and forgetting mechanism of the memory unit on the historical data.

9. The method for intelligent early warning and section positioning of a distribution network based on transient recording data according to claim 8, characterized in that: In step 4, when the fault location matrix is ​​used to achieve accurate section location, a multi-criteria fusion method based on fuzzy c-means clustering is combined to comprehensively consider various fault feature information to improve the accuracy of location.

10. The method for intelligent early warning and section positioning of distribution network based on transient recording data according to claim 9, characterized in that: It also includes the step of establishing a fault waveform library. During the analysis and processing of fault data, the waveform data of different types of faults and their corresponding characteristic parameters and fault type labels are stored in the fault waveform library to provide sample references for subsequent abnormality assessment, cause identification, abnormality recognition and section positioning.

11. The method for intelligent early warning and section positioning of distribution network based on transient recording data according to claim 10, characterized in that: When evaluating the abnormal operation of medium and low voltage lines and identifying their causes, identifying line anomalies and locating sections based on transient recording data, the real-time collected fault data is matched with samples in the fault waveform library, and the fault type and abnormal area are determined by calculating the sample test value.

12. The method for intelligent early warning and section positioning of distribution network based on transient recording data according to claim 11, characterized in that: When it is determined that there is an abnormality or fault in the line, an early warning message is generated and sent to the terminal device of the operation and maintenance personnel. At the same time, the abnormal or faulty section is marked on a map of the entire station of the distribution automation master station.

Images