Overhead line grounding fault positioning method and system based on double-frequency signals

Abstract

The invention discloses an overhead line grounding fault positioning method and system based on a double-frequency signal, relates to the technical field of line fault identification and positioning, and solves the problem of providing high enough accuracy and efficiency in fault detection and positioning tasks in a complex environment in the prior art. According to the method, an online learning algorithm is combined, the model can be dynamically adjusted according to new data, and the positioning accuracy and the response speed are further improved. Through carefully designed feature selection and construction processes, features most related to fault types and positions can be captured, so that rich information is provided for the model. Advanced machine learning algorithms such as random forests are used, a cross validation technology is combined, fault modes are effectively learned from data, and model parameters are optimized. The method based on data driving not only improves the accuracy of fault detection, but also enhances the generalization ability of the model to unknown fault modes, and provides powerful support for stable operation of a power system.

Description

Technical Field

[0001] The present invention relates to the technical field of line fault identification and positioning, and in particular to a method and system for positioning an overhead line grounding fault based on a dual-frequency signal. Background Art

[0002] Fault identification and location of power lines play a vital role in modern power systems. With the continuous increase in power demand and the complexity of power systems, the frequency and types of faults are becoming increasingly diverse. Timely and accurate identification and location of faults can not only reduce power outage time and economic losses, but also protect equipment safety and improve power supply quality. Fault identification involves real-time monitoring of the operating status of the power system, and timely detection of potential faults by analyzing parameters such as voltage and current. Fault location is the process of determining the fault location by analyzing fault signals and system topology. The accuracy of this process directly affects the efficiency of fault recovery and the stable operation of the power grid. Fault identification and location technology is particularly important in the operation and maintenance of power systems. At present, most 10kV distribution overhead lines use a small current grounding system, which is prone to single-phase grounding faults, causing line power outages and even damaging the safe operation of power system equipment. Therefore, it is necessary to find the fault line and locate the fault point in time.

[0003] However, existing line fault identification and location methods often face many challenges when dealing with fault detection in complex environments. Traditional methods rely on characteristic analysis of voltage and current signals, may not be able to accurately identify intermittent faults or high-impedance faults, and have limitations in the speed and accuracy of fault location. These shortcomings make troubleshooting more difficult, extend maintenance response time, and increase economic losses. Existing line fault identification and location methods have some shortcomings, especially when dealing with fault detection and location tasks in complex environments, which may not provide sufficiently high accuracy and efficiency. For example, traditional fault identification methods may have difficulty accurately identifying intermittent faults or high-impedance faults with unclear characteristics.

[0004] In view of this, a method and system for locating an overhead line grounding fault based on a dual-frequency signal is needed. Summary of the invention

[0005] In order to solve the problem of providing sufficiently high accuracy and efficiency in the fault detection and location tasks in complex environments in the prior art, the present invention provides a method and system for locating overhead line grounding faults based on dual-frequency signals, which can quickly and accurately locate faults by real-time monitoring of the occurrence of faults and by comparing the characteristics of the traveling wave signals at both ends of the line, and show stronger robustness and accuracy in complex environments. This method gradually reduces the fault area in an iterative manner until the fault point is accurately located, greatly improving the efficiency and accuracy of fault handling. The specific technical solution is as follows:

[0006] A method for locating an overhead line grounding fault based on a dual-frequency signal comprises the following steps:

[0007] Collect historical data, including line voltage, current, frequency, power, and fault type and location information;

[0008] Train machine learning models with the collected data to identify different types of faults and their possible locations;

[0009] Collect real-time data, including line voltage, current, frequency, and power, and identify fault type and location information through trained models;

[0010] Based on the location information, the fault scope is further narrowed down through the dichotomy method to locate the fault point.

[0011] Preferably, the process of training a machine learning model is as follows:

[0012] Collect historical data sets, including voltage, current, frequency, and power; then label the data sets with fault labels, which identify the fault type and location corresponding to each data point;

[0013] Perform data cleaning to handle missing values, outliers, and noise, and divide the data set into a training set and a test set;

[0014] Use statistical tests or model-based feature selection methods to evaluate the contribution of each feature to fault prediction, select the most influential features based on the evaluation results, and output the selected feature set;

[0015] Standardize or normalize the features in the feature set to obtain a converted feature set;

[0016] Select a machine learning model and initialize the model. Then input the training dataset and the corresponding fault labels for model training. Use cross-validation and grid search to optimize the model's hyperparameters.

[0017] Use the test dataset to evaluate the performance of the model and adjust the model structure or hyperparameters based on the evaluation results.

[0018] Preferably, the method further comprises the following steps:

[0019] Input the selected feature set, perform feature construction, create new features, obtain the created feature set, and input it into the model for training.

[0020] Preferably, the specific steps of further narrowing the fault scope by binary search are as follows:

[0021] Obtain real-time data and model prediction results, including location information;

[0022] Collect voltage and current traveling wave signals from both ends of the line corresponding to the position information, process the collected traveling wave signals, and extract key features, including wave head arrival time and wave speed;

[0023] Divide the line into two parts and determine which half of the line the fault is located by comparing the characteristics of the traveling wave signals at both ends;

[0024] The suspected fault area is further divided into two parts, and the characteristics of the traveling wave signals at both ends are repeatedly compared to determine which half of the line the fault is located in;

[0025] The scope of fault location is gradually narrowed down through multiple iterations until the fault area is narrowed down to the set threshold range.

[0026] Preferably, the process of determining which half of the line the fault is located in by comparing the characteristics of the traveling wave signals at both ends is as follows:

[0027] Preprocessing the collected traveling wave signal, including amplification, filtering and analog-to-digital conversion;

[0028] Perform digital signal processing to extract the characteristic parameters of the traveling wave, including wave head arrival time, waveform shape, and frequency components;

[0029] The time it takes for the fault wave to arrive at both ends of the line is recorded respectively. Based on the time difference when the wave arrives at both ends, a preliminary judgment is made as to which half of the line the fault is more likely to be located.

[0030] Preferably, the process of determining in which half of the line the fault is located by comparing the characteristics of the traveling wave signals at both ends further comprises the following steps:

[0031] If the time difference between the arrival of the traveling wave at both ends is within the set threshold, it is determined that the fault may be close to the midpoint;

[0032] If the time difference is not within the set threshold, the fault may be close to the end of the traveling wave that arrives later.

[0033] An overhead line ground fault location system based on dual-frequency signals, applied to the method as described above, comprises:

[0034] A host, including a display screen, a communication module and a signal transmission module, wherein the signal transmission module outputs DC high voltage and special frequency AC high voltage, and the display screen is used to display the information collected by the signal acquisition module, and the signal acquisition module is arranged in the operation unit;

[0035] The operating unit includes a signal acquisition module, and the signal acquisition module is connected to the host, and is used to collect the traveling wave signal of the line and transmit it to the host for preprocessing and display;

[0036] The terminal is connected to the host for communication, and includes a data processing terminal and a management terminal. The data processing terminal collects and stores historical line data and trained models, and identifies fault types and location information through the trained models. The management terminal is used to receive fault types and location information, as well as traveling wave signal information preprocessed by the host.

[0037] Preferably, it also includes an alarm unit, which is arranged in at least one place of the host and the terminal, and performs an alarm operation when the trained model identifies a fault, including an audible and visual alarm.

[0038] A computer-readable storage medium includes a stored program, wherein when the program is run, the device where the computer-readable storage medium is located is controlled to execute the overhead line grounding fault locating method based on dual-frequency signals as described above.

[0039] A processor is used to run a program, wherein when the program is run, the overhead line grounding fault location method based on dual-frequency signals as described above is executed.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention has efficient fault location capability. Compared with the prior art, the efficiency of fault location is significantly improved. In the power system. The dichotomy method quickly reduces the fault area to half by comparing the characteristics of the traveling wave signals at both ends of the power grid in real time, and then continues to divide the fault area into two until the fault point is accurately located. This method avoids the tedious process of checking one by one in the traditional method, greatly reducing the time for fault location. In addition, combined with the online learning algorithm, this method can dynamically adjust the model according to the new data, further improving the accuracy and response speed of positioning.

[0042] 2. Traditional fault location methods often rely on fixed rules and thresholds, which may no longer apply when system operating conditions change. Online learning algorithms can update models in real time and adjust prediction strategies based on the latest data, making fault detection and location more accurate. This adaptive capability enables the system to maintain high accuracy of fault location when facing new fault modes or system changes, reducing the possibility of false positives and false negatives.

[0043] 3. Through a carefully designed feature selection and construction process, the scheme is able to capture the features most relevant to the fault type and location, thus providing rich information for the model. Using advanced machine learning algorithms such as random forests, combined with techniques such as cross-validation, fault modes can be effectively learned from data and model parameters can be optimized. This data-driven approach not only improves the accuracy of fault detection, but also enhances the model's generalization ability for unknown fault modes. In this way, fast and accurate fault location can be achieved even in complex or atypical fault situations. The present invention improves the speed and accuracy of fault location, and also enhances the system's adaptability and generalization capabilities, providing strong support for the stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0045] Figure 1 is a flow chart of the method of the present invention;

[0046] Figure 2 It is a system principle diagram of the present invention.

[0047] Description of reference numerals:

[0048] 1. Ground rod; 2. Test line; 3. Signal collector; 4. Host; 5. Alarm module; 6. Display screen; 7. Ground port; 8. Overhead line. DETAILED DESCRIPTION

[0049] 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 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.

[0050] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0051] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0052] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0053] Combine the following Figure 1-Figure 2 , further illustrating the implementation mode of the present invention.

[0054] In one embodiment of the present invention, a method for locating an overhead line grounding fault based on a dual-frequency signal is provided, comprising the following steps:

[0055] S1: Collect historical data, including line voltage, current, frequency, power, and fault type and location information;

[0056] As can be understood, this step acquires a historical fault data set, including characteristics such as voltage, current, frequency, and power factor, and identifies the fault type and location corresponding to each data point. Subsequently, data cleaning and data segmentation are performed. Data cleaning specifically processes missing values, outliers, and noise. Data segmentation specifically divides the data set into a training set and a test set, usually in a ratio of 80:20 or 70:30. After step S1 is completed, the output results are the cleaned training set and test set.

[0057] S2: Train machine learning models using the collected data to identify different types of faults and their possible locations;

[0058] The process of training a machine learning model is as follows:

[0059] S201: Collect the cleaned training set and test set (output result of step S1);

[0060] S203: using statistical tests or model-based feature selection methods to evaluate the contribution of each feature to fault prediction, selecting the most influential features based on the evaluation results, and outputting the selected feature set;

[0061] S204: Standardize or normalize the features in the feature set to obtain a converted feature set; in addition, feature construction can be performed to create new features, obtain the created feature set, and input it into the model for training.

[0062] Feature engineering is the process of extracting, building, and selecting those features that are most useful for predicting the target variable. Good feature engineering can significantly improve the performance of the model. It includes the following operations:

[0063] Feature selection: Based on domain knowledge and exploratory data analysis (EDA), features that are strongly correlated with the fault type are selected.

[0064] Construct new features: Use Pandas’ diff() to calculate the rate of change, and use mean(), std(), etc. to calculate statistical features.

[0065] Feature encoding: One-Hot Encoding or Label Encoding for categorical data

[0066] Feature engineering outputs a set of carefully selected and constructed features that effectively represent the data and are used to train machine learning models.

[0067] S205: Select a machine learning model and initialize the model, then input a training data set and corresponding fault labels to perform model training, and use cross validation and grid search to optimize the model's hyperparameters;

[0068] S206: Use the test data set to evaluate the performance of the model, and adjust the model structure or hyperparameters based on the evaluation results.

[0069] Machine learning models can learn patterns from data and use these patterns to predict labels for unknown data. Choosing the right model and training method is crucial to improving prediction accuracy. It includes the following operations:

[0070] Model selection: Random forest was chosen due to its ability to handle non-linear relationships, built-in support for feature selection, and good generalization capabilities.

[0071] Training model: Use RandomForestClassifier for training and perform cross-validation through cross_val_score to evaluate the generalization ability of the model.

[0072] S3: Collect real-time data, including voltage, current, frequency, and power of the line, and identify the fault type and location information through the trained model; the location information described here is relatively broad location information, such as a problem with a certain line or a certain substation or the line between several power stations. Therefore, the location information obtained in this step is not accurate enough. This embodiment can obtain the precise fault location of the line, thereby generating step S4.

[0073] S4: Based on the location information, the fault scope is further narrowed down through the binary method to locate the fault point.

[0074] The specific steps to further narrow the fault scope through the binary method are as follows:

[0075] S401: Acquire real-time data and model prediction results, where the real-time data and model prediction results include location information;

[0076] S402: collecting voltage and current traveling wave signals from both ends of the line corresponding to the position information, processing the collected traveling wave signals, and extracting key features, including wave head arrival time and wave speed;

[0077] S403: Divide the line into two parts, and determine which half of the line the fault is located in by comparing the characteristics of the traveling wave signals at both ends;

[0078] S404: continue to divide the suspected fault area into two parts, and repeatedly compare the characteristics of the traveling wave signals at both ends to determine which half of the line the fault is located in;

[0079] S405: gradually narrowing the scope of fault location, through multiple iterations, until the fault area is narrowed down to a set threshold range.

[0080] The process of comparing the characteristics of the traveling wave signals at both ends to determine which half of the line the fault is located in is as follows:

[0081] S451: pre-processing the collected traveling wave signal, including amplification, filtering and analog-to-digital conversion;

[0082] S452: Perform digital signal processing to extract characteristic parameters of the traveling wave, including wave head arrival time, waveform shape, and frequency components;

[0083] S453: Record the time when the fault wave arrives at both ends of the line respectively, and preliminarily judge which half of the line the fault is more likely to be located in based on the time difference when the wave arrives at both ends. If the time difference when the wave arrives at both ends is within the set threshold range, it is determined that the fault may be close to the midpoint; if the time difference is not within the set threshold range, the fault may be close to the end where the wave arrives later.

[0084] Fault judgment principle:

[0085] When a ground fault occurs in the line, the signal generator sends a special frequency AC signal with a certain power to the fault line to reproduce the fault. According to Kirchhoff's law, the signal will flow to the earth through the grounding point, that is, a loop is formed between the signal source, the line, the grounding point and the earth. Then, a binary method is used on the path from the beginning of the line to the fault point. In the fault area, a signal with a value approximately equal to or less than the injected signal can be measured, while the current disappears in the non-fault area (including non-fault branches and after the fault point). The grounding interval is gradually reduced until the fault point is found.

[0086] In one embodiment of the present invention, a system for locating an overhead line ground fault based on a dual-frequency signal is proposed, which is applied to the method described above, such as Figure 2 As shown, including:

[0087] The host comprises a display screen, a communication module and a signal transmission module, wherein the signal transmission module outputs DC high voltage and special frequency AC high voltage, and the display screen is used to display the information collected by the signal acquisition module, and the signal acquisition module is arranged in the operation unit;

[0088] The operating unit includes a signal acquisition module, and the signal acquisition module is connected to the host, and is used to collect the traveling wave signal of the line and transmit it to the host for preprocessing and display; the operating unit also includes a grounding rod and a test line, and the grounding rod is connected to the host through the test line.

[0089] The terminal is connected to the host for communication, and includes a data processing terminal and a management terminal. The data processing terminal collects and stores historical line data and trained models, and identifies fault types and location information through the trained models. The management terminal is used to receive fault types and location information, as well as traveling wave signal information preprocessed by the host.

[0090] An alarm unit is provided at least in one place of the host and the terminal, and performs an alarm operation when a fault is identified by the trained model, including an audible and visual alarm.

[0091] In summary, the present invention has efficient fault location capability. Compared with the prior art, the efficiency of fault location is significantly improved. In the power system. The dichotomy method quickly reduces the fault area to half by comparing the characteristics of the traveling wave signals at both ends of the power grid in real time, and then continues to divide the fault into two until the fault point is accurately located. This method avoids the cumbersome process of checking one by one in the traditional method, greatly reducing the time of fault location. In addition, combined with the online learning algorithm, the method can dynamically adjust the model according to the new data, further improving the accuracy and response speed of positioning. Through the carefully designed feature selection and construction process, the present invention can capture the features most relevant to the fault type and location, thereby providing rich information for the model. Using advanced machine learning algorithms such as random forests, combined with cross-validation and other technologies, fault modes can be effectively learned from data and model parameters can be optimized. This data-driven method not only improves the accuracy of fault detection, but also enhances the generalization ability of the model for unknown fault modes. In this way, fast and accurate fault location can be achieved even in complex or atypical fault situations. The present invention improves the speed and accuracy of fault location, and also enhances the adaptive ability and generalization ability of the system, providing strong support for the stable operation of the power system.

[0092] 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.

[0093] 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, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0094] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0095] 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.

[0096] Finally, it should be noted 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned 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 method for locating an overhead line grounding fault based on a dual-frequency signal, characterized in that: The following steps are involved: Collect historical data, including line voltage, current, frequency, power, and fault type and location information; Train machine learning models with the collected data to identify different types of faults and their possible locations; Collect real-time data, including line voltage, current, frequency, and power, and identify fault type and location information through trained models; Based on the location information, the fault scope is further narrowed down through the dichotomy method to locate the fault point.

2. The method for locating an overhead line grounding fault based on a dual-frequency signal according to claim 1, characterized in that: The process of training a machine learning model is as follows: Collect historical data sets, including voltage, current, frequency, and power; then label the data sets with fault labels, which identify the fault type and location corresponding to each data point; Perform data cleaning to handle missing values, outliers, and noise, and divide the data set into a training set and a test set; Use statistical tests or model-based feature selection methods to evaluate the contribution of each feature to fault prediction, select the most influential features based on the evaluation results, and output the selected feature set; Standardize or normalize the features in the feature set to obtain a converted feature set; Select a machine learning model and initialize the model. Then input the training dataset and the corresponding fault labels for model training. Use cross-validation and grid search to optimize the model's hyperparameters. Use the test dataset to evaluate the performance of the model and adjust the model structure or hyperparameters based on the evaluation results.

3. The method for locating an overhead line grounding fault based on a dual-frequency signal according to claim 2, characterized in that: The following steps are also included: Input the selected feature set, perform feature construction, create new features, obtain the created feature set, and input it into the model for training.

4. The method for locating an overhead line grounding fault based on a dual-frequency signal according to claim 1, characterized in that: The specific steps to further narrow the scope of the fault through the binary method are as follows: Obtain real-time data and model prediction results, including location information; Collect voltage and current traveling wave signals from both ends of the line corresponding to the position information, process the collected traveling wave signals, and extract key features, including wave head arrival time and wave speed; Divide the line into two parts and determine which half of the line the fault is located by comparing the characteristics of the traveling wave signals at both ends; The suspected fault area is further divided into two parts, and the characteristics of the traveling wave signals at both ends are repeatedly compared to determine which half of the line the fault is located in; The scope of fault location is gradually narrowed down through multiple iterations until the fault area is narrowed down to the set threshold range.

5. The method for locating an overhead line grounding fault based on a dual-frequency signal according to claim 4, characterized in that: The process of comparing the characteristics of the traveling wave signals at both ends to determine which half of the line the fault is located in is as follows: Preprocessing the collected traveling wave signal, including amplification, filtering and analog-to-digital conversion; Perform digital signal processing to extract the characteristic parameters of the traveling wave, including wave head arrival time, waveform shape, and frequency components; The time it takes for the fault wave to arrive at both ends of the line is recorded respectively. Based on the time difference when the wave arrives at both ends, a preliminary judgment is made as to which half of the line the fault is more likely to be located.

6. The method for locating an overhead line grounding fault based on a dual-frequency signal according to claim 5, characterized in that: The process of comparing the characteristics of the traveling wave signals at both ends to determine which half of the line the fault is located in also includes the following steps: If the time difference between the arrival of the traveling wave at both ends is within the set threshold, it is determined that the fault may be close to the midpoint; If the time difference is not within the set threshold, the fault may be close to the end of the traveling wave that arrives later.

7. An overhead line grounding fault location system based on dual-frequency signals, characterized in that: The method applied to any one of claims 1 to 6 comprises: The host comprises a display screen, a communication module and a signal transmission module, wherein the signal transmission module outputs DC high voltage and special frequency AC high voltage, and the display screen is used to display the information collected by the signal acquisition module, and the signal acquisition module is arranged in the operation unit; The operating unit includes a signal acquisition module, and the signal acquisition module is connected to the host, and is used to collect the traveling wave signal of the line and transmit it to the host for preprocessing and display; The terminal is connected to the host for communication, and includes a data processing terminal and a management terminal. The data processing terminal collects and stores historical line data and trained models, and identifies fault types and location information through the trained models. The management terminal is used to receive fault types and location information, as well as traveling wave signal information preprocessed by the host.

8. The overhead line grounding fault location system based on dual-frequency signals according to claim 7, characterized in that: It also includes an alarm unit, which is set in at least one place of the host and the terminal. When the trained model identifies a fault, an alarm operation is performed, including an audible and visual alarm.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the overhead line grounding fault location method based on dual-frequency signals as described in any one of claims 1 to 6.

10. A processor, characterized in that: The processor is used to run a program, wherein when the program is run, the overhead line grounding fault location method based on dual-frequency signals as described in any one of claims 1 to 6 is executed.

Images