Fault diagnosis method for DC distribution network
By acquiring and analyzing the abnormal data of the slime mold model and the network information fusion data, combined with fuzzy logic processing and machine learning optimization, the problem of invalid positioning and invalid fault diagnosis in the fault diagnosis of DC distribution network line is solved, achieving higher accuracy and stability, improving maintenance efficiency.
Patent Information
- Application Number
- CN202411260970.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-10
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-09-10
Smart Images

Figure CN118797527B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of direct current distribution network, and more specifically, to a method for diagnosing line faults in a direct current distribution network. Background Art
[0002] DC distribution networks play an increasingly important role in power systems, especially with the rapid development of renewable energy and electric vehicles. Compared with AC distribution networks, DC distribution networks have many advantages, such as higher energy efficiency, lower energy loss, and better power quality control. However, one of the challenges associated with it is the complexity of diagnosing and locating line faults in DC networks.
[0003] The line characteristics of DC distribution network include unidirectional current transmission, voltage stability, low frequency operation, lower voltage loss, high current stability and strict requirements on circuit breakers and disconnectors. These characteristics make DC distribution network have unique advantages in energy transmission and distribution, such as higher energy efficiency and more stable power quality.
[0004] For example, the patent application with publication number: CN116449147A discloses a method for locating faults in a DC distribution network line. The method associates the line parameters according to the fault type of the DC distribution network line and establishes a fitness function, wherein the line parameters include the positioning parameters of the fault point, and the independent variable of the fitness function is the positioning parameter; the value of the positioning parameter is used as a population individual, and the slime mold optimization algorithm is used to determine the population individual with the best fitness, and the optimal solution of the positioning parameter is obtained, wherein the optimal solution of the positioning parameter is used to represent the location of the fault point in the DC distribution network line. In this way, the problem that it is difficult to realize automatic positioning of the current DC distribution network line fault can be solved.
[0005] For example, the patent application with publication number: CN116449147A discloses a DC distribution network fault diagnosis method based on Bayesian network information fusion, which improves the traditional Bayesian network power grid fault diagnosis model based on Bayesian network information fusion by targeting the commonalities and differences between DC distribution network and AC distribution network when faults occur. The steps are: Bayesian network information fusion, Bayesian network node conditional probability assignment, and DC distribution network fault diagnosis. The present invention improves the traditional Bayesian network power grid fault diagnosis model, realizes DC distribution network fault diagnosis, and verifies the reliability of the method through actual examples.
[0006] However, the inventors of this application found that the above technology has at least the following technical problems in the process of implementing the technical solution of the invention in the embodiment of this application:
[0007] (1) The first method uses the slime mold optimization algorithm to determine the individuals in the population with the best fitness in order to obtain the optimal positioning parameters. However, in practical applications, the accuracy of line parameters may be affected by many factors, such as sensor accuracy, environmental noise, etc., which may affect the effectiveness of the algorithm, resulting in invalid positioning and thus invalid fault diagnosis.
[0008] (2) In the second method, the power grid fault diagnosis model based on the Bayesian network is improved to realize fault diagnosis of the DC distribution network. However, in practical applications, the Bayesian network model may be affected by factors such as the diversity of fault types and parameter uncertainty, resulting in ineffective fault diagnosis.
[0009] (3) Fault diagnosis methods need to strike a balance between real-time performance and algorithm complexity. The above-mentioned complex algorithms may result in high computational costs, thus affecting the real-time performance of fault diagnosis; thus, they cannot meet the requirements for accurate diagnosis in complex fault situations.
[0010] In view of the above problems, the present invention proposes a solution. Summary of the invention
[0011] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for diagnosing line faults in a DC distribution network, which solves the problems raised in the above-mentioned background technology through a DC distribution network.
[0012] To achieve the above object, the present invention provides the following technical solutions:
[0013] A DC distribution network line fault diagnosis method comprises the following steps: obtaining slime mold model abnormal data before locating the line fault, and obtaining network information fusion data when diagnosing the line fault type; comparing the slime mold model abnormal data and the network information fusion data to obtain a first comparison result; combining the first comparison result with the constraint conditions of each diagnosis effect category to analyze and obtain the actual diagnosis effect category; optimizing the diagnosis scheme according to the actual diagnosis effect category; establishing an optimization effect evaluation model through machine learning according to the optimized slime mold model abnormal data and the network information fusion data to generate an optimization effect evaluation coefficient; and obtaining the power grid line fault diagnosis effect according to the actual diagnosis effect combined with the optimization effect evaluation model.
[0014] In a preferred embodiment, the slime mold model abnormal data includes an environmental noise interference coefficient and a line-related abnormality coefficient, and the network information fusion data includes a nonlinear load influence coefficient.
[0015] In a preferred embodiment, the environmental noise interference coefficient and the line-related abnormality coefficient are comprehensively processed using a fuzzy logic method to obtain a slime mold model effect evaluation value; the slime mold model effect evaluation value and the nonlinear load influence coefficient are compared and analyzed with a preset slime mold model effect standard value and a preset nonlinear load influence standard value, respectively, and the analysis result is calibrated as a first comparison result.
[0016] In a preferred embodiment, the specific method for optimizing the diagnostic scheme according to the actual diagnostic effect category is as follows: when the actual diagnostic effect category is a valid category, a normal DC distribution network line fault diagnostic method is performed; when the actual diagnostic effect category is an invalid category, a first-level alarm is issued to relevant technical personnel, and the ongoing DC distribution network line fault diagnostic method is terminated; when the actual diagnostic effect category is an optimizable category, a second-level alarm is issued to relevant technical personnel, and the ongoing DC distribution network line fault diagnostic method is optimized in process and external manner.
[0017] In a preferred embodiment, the specific method for obtaining the power grid line fault diagnosis effect based on the actual diagnostic effect combined with the optimization effect evaluation model is as follows: compare the optimization effect evaluation coefficient with the preset optimization effect evaluation threshold; when the optimization effect evaluation coefficient is greater than the preset optimization effect evaluation threshold, the actual diagnostic effect category of the optimized scheme is marked as a valid category; when the optimization effect evaluation coefficient is less than the preset optimization effect evaluation threshold, the actual diagnostic effect category of the optimized scheme is marked as an invalid category.
[0018] In a preferred embodiment, the method for obtaining the environmental noise interference coefficient is as follows:
[0019] The current signal and actual current value of the DC distribution network in the Z time interval within the G time are obtained, the current signal is used as the input signal, and the input signal set is established by labeling, and the actual current value is used as the observation value, and the observation value set is established by labeling; the expected value of the square of the difference between the observed value and the true value is calculated to obtain the noise mean square error; the real-time voltage signal is collected in the DC distribution network by using the measuring equipment, and the signal energy is calculated by the square of the amplitude of the real-time voltage signal; the signal power is calculated according to the ratio of the signal energy and the signal duration; the noise anomaly is calculated according to the noise mean square error and signal power analysis; the noise anomaly values at all time points are averaged to calculate the environmental noise interference coefficient.
[0020] In a preferred embodiment, the method for obtaining the line-related anomaly coefficient is as follows:
[0021] Obtain the line parameters to be used for slime mold optimization algorithm modeling, label and establish a line set, calculate the mean, difference, cumulative product sum and standard deviation of the line parameters for the relationship between two line parameters, and use the Pearson algorithm to calculate the actual line association number; combine the data in the line set with the preset line parameter expected relationship model to calculate the expected line relationship number; obtain the line association anomaly coefficient based on the difference between the actual line association number and the expected line relationship number between each line parameter.
[0022] In a preferred embodiment, the method for obtaining the nonlinear load influence coefficient is as follows:
[0023] The voltage signal waveform of the U time interval within the G time period when diagnosing the line fault type is obtained, and the harmonic components of the voltage signal waveform are extracted in combination with Fourier transform. The amplitude of each harmonic component is calculated by spectrum analysis; the amplitude of each harmonic component is squared and added to obtain the square sum of the harmonics; the maximum amplitude of all harmonic components is obtained; the maximum amplitude of all harmonic components is compared with the square sum of the harmonics, and the nonlinear load influence coefficient is calculated.
[0024] Technical effects and advantages of the DC distribution network line fault diagnosis method of the present invention:
[0025] 1. The present invention studies the environmental noise interference coefficient, line-related abnormality coefficient and nonlinear load influence coefficient in the fault location of the DC distribution network, and deeply explores their influence on the effect of the fault location method; first, the increase of the environmental noise interference coefficient will reduce the accuracy and stability of the fault location algorithm, because the high interference coefficient will make it difficult to distinguish the fault signal from the environmental noise, thereby affecting the accuracy and stability; secondly, the abnormality of the line-related abnormality coefficient will cause the fitness function to fail, thereby affecting the accuracy and reliability of the fault location, and reducing the credibility of the algorithm in practical applications; finally, the increase of the nonlinear load influence coefficient may cause harmonic disturbances, decreased voltage stability, current distortion and other problems, thereby affecting the accuracy of the Bayesian network model in estimating line parameters and fault diagnosis; therefore, by timely acquiring and analyzing these coefficients, it is possible to provide an important basis and reference for improving the fault location algorithm, improving the accuracy of fault diagnosis and maintenance efficiency.
[0026] 2. The present invention uses a fuzzy logic method to comprehensively process the environmental noise interference coefficient and the line-related abnormality coefficient to obtain a slime mold model effect evaluation value, and compares and analyzes it with a preset standard value to obtain a first comparison result; then, the diagnostic scheme is optimized according to the actual diagnostic effect category, including normal diagnosis, alarm issuance, and diagnostic method optimization steps; then, an optimization effect evaluation model is established through machine learning, the optimization effect evaluation coefficient is calculated, and it is compared with a preset optimization effect evaluation threshold to determine the actual diagnostic effect category of the optimized scheme; this method can effectively improve the accuracy and stability of DC distribution network line fault diagnosis, provide maintenance personnel with more reliable diagnostic results, and thus improve maintenance efficiency and reliability. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1 It is a structural schematic diagram of a DC distribution network line fault diagnosis method according to the present invention. DETAILED DESCRIPTION
[0028] 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.
[0029] Embodiment 1, Figure 1 A schematic diagram of the structure of a DC power distribution network line fault diagnosis method of the present invention is given, which includes the following steps:
[0030] Before locating the line fault, obtain the abnormal data of the slime mold model, and when diagnosing the line fault type, obtain the network information fusion data;
[0031] The abnormal data of the slime mold model is compared with the network information fusion data to obtain a first comparison result;
[0032] The first comparison result is combined with the constraint conditions of each diagnostic effect category to analyze and obtain the actual diagnostic effect category;
[0033] Optimize the diagnostic scheme according to the actual diagnostic effect category;
[0034] An optimization effect evaluation model is established through machine learning based on the optimized slime mold model abnormal data and network information fusion data;
[0035] The power grid line fault diagnosis effect is obtained based on the actual diagnosis effect combined with the optimization effect evaluation model.
[0036] The abnormal data of the slime mold model includes an environmental noise interference coefficient and a line-related abnormality coefficient.
[0037] The operating environment of the DC distribution network may be affected by many factors, including environmental noise such as electromagnetic interference, mechanical noise, electronic noise, and natural noise. There may also be interference from factors such as equipment failure and external interference. These noises and interferences may cause the signals collected by sensors or monitoring equipment to contain distortion or distortion, thereby affecting the accuracy and reliability of the fault location algorithm. In order to cope with these problems, the fault location algorithm needs to have anti-interference capabilities and take corresponding signal processing and data correction measures to reduce the impact of environmental noise and interference on fault location.
[0038] The environmental noise interference coefficient has the following influence and effect on the effect of analyzing the method for solving line faults:
[0039] Accuracy impact: The increase in environmental noise interference coefficient may cause more interference to the sensor signal, thus affecting the accuracy of the fault location algorithm. Too high an environmental noise interference coefficient will increase the noise component, making it difficult to distinguish between the fault signal and the environmental interference, thereby reducing the accuracy of fault location.
[0040] Stability impact: Changes in the environmental noise interference coefficient may affect the stability of the fault location algorithm. When the environmental noise interference coefficient changes, the fault location algorithm may need to adjust parameters in a timely manner or take appropriate measures to adapt to environmental changes in order to maintain the stability of fault location.
[0041] Signal-to-noise ratio impact: The increase in environmental noise interference coefficient will reduce the signal-to-noise ratio between the signal and the noise, making it difficult to distinguish between the fault signal and the environmental noise. A low signal-to-noise ratio may cause the fault location algorithm to fail to correctly identify the fault signal, thus affecting the accuracy and reliability of fault location.
[0042] Impact of algorithm anti-interference ability: The increase of environmental noise interference coefficient may increase the anti-interference requirement of the fault location algorithm. For algorithms with higher anti-interference ability, even in the case of high environmental noise interference coefficient, they can effectively identify and locate fault signals, thereby ensuring the accuracy of fault location.
[0043] Therefore, timely acquisition and analysis of the environmental noise interference coefficient can accurately and effectively analyze and determine the effectiveness of the method used to solve line faults.
[0044] The method for obtaining the environmental noise interference coefficient is as follows:
[0045] The current signal and actual current value of the DC distribution network within the time interval Z within the time G are obtained, the current signal is used as the input signal, the input signal set is established by labeling, and the actual current value is used as the observation value, the observation value set is established by labeling; the expected value of the square of the difference between the observed value and the true value is calculated to obtain the noise mean square error;
[0046] Use measurement equipment to collect real-time voltage signals in the DC distribution network, calculate the signal energy by the square of the amplitude of the real-time voltage signal; calculate the signal power based on the ratio of the signal energy and the signal duration;
[0047] The noise outlier value is calculated based on the noise mean square error and signal power analysis; the noise outlier values at all time points are averaged to calculate the environmental noise interference coefficient.
[0048] The calculation formula of the environmental noise interference coefficient is as follows:
[0049]
[0050] In the formula, is the environmental noise interference coefficient, To obtain the number of signals, is the current signal, is the actual current value, is the real-time voltage signal, Sampling time interval, is the duration of the signal.
[0051] It can be seen from the calculation formula of the environmental noise interference coefficient that when the performance value of the environmental noise interference coefficient is larger, it indicates that the effect of the method used to solve the line fault is worse, and vice versa, it indicates that the effect of the method used to solve the line fault is better.
[0052] In the DC distribution network line fault location method, inaccurate association of line parameters may cause the fitness function to fail, thus affecting the accuracy of fault location. If the line parameters are not correctly associated with the fault type when establishing the fitness function, the solution generated by the algorithm may not match the actual fault nature, and the fault point location cannot be accurately determined. It is crucial to ensure that the line parameters and fault types are correctly and completely associated.
[0053] The line-related anomaly coefficient has the following effects and functions on analyzing the effect of the method for solving line faults:
[0054] Decreased positioning accuracy: Inaccurate association between line parameters and fault types will cause the fitness function to fail, thus affecting the accurate positioning of the fault point. Because the fitness function cannot accurately reflect the nature and location of the fault, the accuracy of fault positioning will be seriously affected.
[0055] Increased difficulty in solving the problem: If the line parameters are not fully associated with the fault type, some key features may be ignored, making it difficult for the algorithm to correctly identify the fault location, especially in complex fault situations.
[0056] Misleading algorithm results: Line-related anomaly coefficients may mislead the algorithm to produce erroneous solutions, which may not match the actual faults, making the fault location results lose their reference value and unable to help maintenance personnel accurately find the fault point.
[0057] Reduced reliability of the method: Abnormal line correlation coefficients will reduce the reliability and stability of the fault location method, significantly reducing the credibility of the algorithm in practical applications, affecting the practicality and effectiveness of the method.
[0058] Therefore, timely acquisition and analysis of line-related abnormality coefficients can accurately and effectively analyze and determine the effectiveness of methods used to resolve line faults.
[0059] The method for obtaining the line-related anomaly coefficient is as follows:
[0060] Obtain the line parameters to be used for slime mold optimization algorithm modeling, label and establish a line set, calculate the mean, difference, cumulative product sum and standard deviation of the line parameters for the relationship between two line parameters, and calculate the actual line association number through the Pearson algorithm;
[0061] The data in the line set is combined with a preset line parameter expected relationship model to calculate the expected line relationship number;
[0062] The line correlation anomaly coefficient is obtained by calculating the difference between the actual line correlation number and the expected line correlation number between various line parameters.
[0063] The calculation formula of the actual line association number is as follows:
[0064]
[0065] The calculation formula of the line association anomaly coefficient is as follows:
[0066]
[0067] In the formula, is the actual line association number, is the data volume of a single line parameter, and is the label of the line type in the line set, is the number of line types, is the expected number of line relationships.
[0068] It should be noted that the preset line parameter expected relationship model is determined according to the physical relationship or theoretical model between the line parameters, so as to determine the expected correlation coefficient between the line parameters.
[0069] It can be seen from the calculation formula of the line-related abnormality coefficient that when the performance value of the line-related abnormality coefficient is larger, it indicates that the effect of the method used to solve the line fault is worse, and vice versa, it indicates that the effect of the method used to solve the line fault is better.
[0070] The network information fusion data includes a nonlinear load influence coefficient.
[0071] The load on the power grid may be nonlinear, such as the nonlinear load introduced by arc furnaces, inverters, etc. These nonlinear loads may cause changes in the voltage and current waveforms of the power grid, thereby affecting the accuracy of the Bayesian network model in estimating line parameters and fault diagnosis.
[0072] The nonlinear load influence coefficient has the following influence and effect on the effect of analyzing the method for solving line faults:
[0073] Harmonic disturbance: Nonlinear loads will introduce harmonic currents and harmonic voltages, which will change the current and voltage waveforms in the power grid. Harmonic disturbances may cause voltage distortion, current distortion and other problems in the power grid, which in turn affects the accurate estimation of line parameters by the Bayesian network model.
[0074] Voltage stability: Nonlinear loads can change the voltage waveform of the power grid line, causing the voltage stability of the power grid to decrease. For example, equipment such as arc furnaces may cause voltage fluctuations, while equipment such as inverters may cause voltage instability. Such voltage fluctuations and instabilities may affect the parameter estimation of the Bayesian network model, thereby affecting the accuracy of fault diagnosis.
[0075] Current distortion: Harmonic currents introduced by nonlinear loads may cause current distortion in the power grid, making the current waveform no longer a sine wave. This current distortion may affect the estimation of line parameters by the Bayesian network model, especially the estimation accuracy of current-related parameters (such as resistance, inductance, etc.).
[0076] Overload and overcurrent problems: Nonlinear loads may cause overload and overcurrent problems in power grid lines, which may cause damage or failure of line equipment. Overload and overcurrent problems may cause the parameter estimation of the Bayesian network model to fail, thus affecting the accuracy of fault diagnosis.
[0077] Therefore, timely acquisition and analysis of the nonlinear load influence coefficient can accurately and effectively analyze and determine the effectiveness of the method used to solve the line fault.
[0078] The method for obtaining the nonlinear load influence coefficient is as follows:
[0079] The voltage signal waveform of the U time interval within the G time period when diagnosing the line fault type is obtained, and the harmonic components of the voltage signal waveform are extracted in combination with Fourier transform. The amplitude of each harmonic component is calculated by spectrum analysis; the amplitude of each harmonic component is squared and added to obtain the square sum of the harmonics; the maximum amplitude of all harmonic components is obtained; the maximum amplitude of all harmonic components is compared with the square sum of the harmonics, and the nonlinear load influence coefficient is calculated.
[0080] The method for obtaining the harmonic components is as follows:
[0081]
[0082] The method for obtaining the nonlinear load influence coefficient is as follows:
[0083]
[0084] In the formula, is the nonlinear load influence coefficient, is the amplitude of the harmonic component, is the number of voltage signal waveforms, is the frequency, For time, is the maximum amplitude of all harmonic components.
[0085] It can be seen from the calculation formula of the nonlinear load influence coefficient that when the performance value of the nonlinear load influence coefficient is larger, it indicates that the effect of the method used to solve the line fault is worse, and vice versa, it indicates that the effect of the method used to solve the line fault is better.
[0086] This embodiment studies the environmental noise interference coefficient, line-related abnormal coefficient and nonlinear load influence coefficient in the fault location of the DC distribution network, and deeply explores their influence on the effect of the fault location method. First, the increase of the environmental noise interference coefficient will reduce the accuracy and stability of the fault location algorithm, because the high interference coefficient will make it difficult to distinguish the fault signal from the environmental noise, thereby affecting the accuracy and stability. Secondly, the abnormality of the line-related abnormal coefficient will cause the fitness function to fail, thereby affecting the accuracy and reliability of the fault location, and reducing the credibility of the algorithm in practical applications. Finally, the increase of the nonlinear load influence coefficient may cause harmonic disturbances, decreased voltage stability, current distortion and other problems, thereby affecting the accuracy of the Bayesian network model in estimating line parameters and fault diagnosis. Therefore, by timely acquiring and analyzing these coefficients, it is possible to provide important basis and reference for improving the fault location algorithm, improving the accuracy of fault diagnosis and maintenance efficiency.
[0087] In Example 2, the abnormal data of the slime mold model is compared with the network information fusion data to obtain a first comparison result;
[0088] The specific method for obtaining the first comparison result is as follows:
[0089] The fuzzy logic method is used to comprehensively process the environmental noise interference coefficient and the line-related anomaly coefficient to obtain the slime mold model effect evaluation value.
[0090] Compare and analyze the slime mold model effect evaluation value and the nonlinear load influence coefficient with the preset slime mold model effect standard value and the preset nonlinear load influence standard value, respectively, and mark the analysis result as the first comparison result;
[0091] It should be noted that the fuzzy logic method is as follows:
[0092] The environmental noise interference coefficient and the line-related abnormality coefficient are fuzzified and converted into fuzzy sets;
[0093] A set of fuzzy rules is formulated to describe the relationship between the environmental noise interference coefficient and the line-related anomaly coefficient;
[0094] According to the set fuzzy rules, fuzzy reasoning is performed to obtain a fuzzy output, which represents the effect evaluation value of the slime mold model;
[0095] Convert the fuzzy output into a specific evaluation value.
[0096] The first comparison result is combined with the constraint conditions of each diagnostic effect category to analyze and obtain the actual diagnostic effect category;
[0097] The diagnostic effect categories include effective categories, ineffective categories, and optimizable categories;
[0098] The constraints of each diagnostic effect category are as follows:
[0099] Constraints on valid categories: .
[0100] Invalid category constraints: .
[0101] Constraints for the optimizable categories:
[0102] In the formula, is the slime mold model effect evaluation value, is the standard value of the slime mold model effect, Standard value for nonlinear load influence;
[0103] The first comparison result is used to obtain an actual diagnosis effect category based on the satisfied constraint conditions.
[0104] Optimize the diagnostic scheme according to the actual diagnostic effect category;
[0105] Specifically, when the actual diagnosis effect category is a valid category, a normal DC distribution network line fault diagnosis method is performed;
[0106] When the actual diagnosis effect category is an invalid category, a first-level alarm is issued to relevant technical personnel, and the ongoing DC distribution network line fault diagnosis method is terminated;
[0107] When the actual diagnosis effect category is an optimizable category, a secondary alarm is issued to the relevant technical personnel to perform process and external optimization on the ongoing DC distribution network line fault diagnosis method.
[0108] According to the abnormal data of the optimized slime mold model and the fusion data of network information, an optimization effect evaluation model is established through machine learning to generate an optimization effect evaluation coefficient;
[0109] Specifically, the environmental noise interference coefficient, line-related abnormality coefficient and nonlinear load influence coefficient after the optimization method are obtained to establish an optimization effect evaluation model.
[0110] The calculation formula of the optimization effect evaluation coefficient is as follows:
[0111]
[0112] In the formula, To optimize the effect evaluation coefficient, is the preset proportional coefficient of the environmental noise interference coefficient, is the preset proportional coefficient of the line-related anomaly coefficient, is the preset proportionality factor of the nonlinear load influence coefficient, is the environmental noise interference coefficient, is the line-related anomaly coefficient, is the nonlinear load influence coefficient, and , , Both are greater than 0.
[0113] The power grid line fault diagnosis effect is obtained based on the actual diagnosis effect combined with the optimization effect evaluation model.
[0114] Specifically, the optimization effect evaluation coefficient is compared with a preset optimization effect evaluation threshold. When the optimization effect evaluation coefficient is greater than the preset optimization effect evaluation threshold, the actual diagnostic effect category of the optimized scheme is marked as a valid category. When the optimization effect evaluation coefficient is less than the preset optimization effect evaluation threshold, the actual diagnostic effect category of the optimized scheme is marked as an invalid category.
[0115] In this embodiment, the environmental noise interference coefficient and the line-related abnormality coefficient are comprehensively processed by a fuzzy logic method to obtain a slime mold model effect evaluation value, and a comparison analysis is performed with a preset standard value to obtain a first comparison result; then, the diagnostic scheme is optimized according to the actual diagnostic effect category, including steps such as normal diagnosis, alarm issuance, and diagnostic method optimization; then, an optimization effect evaluation model is established through machine learning, the optimization effect evaluation coefficient is calculated, and it is compared with a preset optimization effect evaluation threshold to determine the actual diagnostic effect category of the optimized scheme; this method can effectively improve the accuracy and stability of DC distribution network line fault diagnosis, provide maintenance personnel with more reliable diagnostic results, and thus improve maintenance efficiency and reliability.
[0116] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.
[0117] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.
[0118] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. 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 this application.
[0119] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.
[0120] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
[0121] 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 protection scope of the present invention.
Claims
1. DC distribution network line fault diagnosis method, It is characterized in that The steps include: Before locating the line fault, obtain the abnormal data of the slime mold model, and when diagnosing the line fault type, obtain the network information fusion data; The slime mold model abnormal data includes an environmental noise interference coefficient and a line-related abnormality coefficient, and the network information fusion data includes a nonlinear load influence coefficient; The method for obtaining the environmental noise interference coefficient is as follows: Obtain the current signal and actual current value of the DC distribution network within the time interval Z within the time G, use the current signal as the input signal, and establish an input signal set with a label, and use the actual current value as the observation value, and establish an observation value set with a label; Calculate the expected value of the square of the difference between the observed value and the true value to get the noise mean square error; Use measurement equipment to collect real-time voltage signals in the DC distribution network, calculate the signal energy by the square of the amplitude of the real-time voltage signal; calculate the signal power based on the ratio of the signal energy and the signal duration; The noise anomaly value is calculated based on the noise mean square error and signal power analysis; the noise anomaly values at all time points are averaged to calculate the environmental noise interference coefficient; The method for obtaining the line-related anomaly coefficient is as follows: Obtain the line parameters to be used for slime mold optimization algorithm modeling, label and establish a line set, calculate the mean, difference, cumulative product sum and standard deviation of the line parameters for the relationship between two line parameters, and calculate the actual line association number through the Pearson algorithm; The data in the line set is combined with a preset line parameter expected relationship model to calculate the expected line relationship number; According to the difference between the actual line correlation number and the expected line correlation number between the various line parameters, the line correlation anomaly coefficient is obtained; The abnormal data of the slime mold model is compared with the network information fusion data to obtain a first comparison result; The first comparison result is combined with the constraint conditions of each diagnostic effect category to analyze and obtain the actual diagnostic effect category; Optimize the diagnostic scheme according to the actual diagnostic effect category; According to the abnormal data of the optimized slime mold model and the fusion data of network information, an optimization effect evaluation model is established through machine learning to generate an optimization effect evaluation coefficient; The calculation formula of the optimization effect evaluation coefficient is as follows: In the formula, To optimize the effect evaluation coefficient, is the preset proportional coefficient of the environmental noise interference coefficient, is the preset proportional coefficient of the line-related anomaly coefficient, is the preset proportionality factor of the nonlinear load influence factor, is the environmental noise interference coefficient, is the line-related anomaly coefficient, is the nonlinear load influence coefficient, and , , All are greater than 0; The power grid line fault diagnosis effect is obtained based on the actual diagnosis effect combined with the optimization effect evaluation model.
2. The DC power distribution network line fault diagnosis method according to claim 1, It is characterized in that The specific method for obtaining the first comparison result is as follows: The fuzzy logic method is used to comprehensively process the environmental noise interference coefficient and the line-related anomaly coefficient to obtain the slime mold model effect evaluation value. The slime mold model effect evaluation value and the nonlinear load influence coefficient are compared and analyzed with the preset slime mold model effect standard value and the preset nonlinear load influence standard value, respectively, and the analysis result is calibrated as the first comparison result.
3. The DC power distribution network line fault diagnosis method according to claim 2, It is characterized in that The actual diagnostic effect categories include effective categories, ineffective categories, and optimizable categories; The constraints of each diagnostic effect category are as follows: Constraints for valid categories: ; Invalid category constraints: ; Constraints for the optimizable categories: In the formula, is the slime mold model effect evaluation value, is the standard value of the slime mold model effect, The standard value for the influence of nonlinear loads. is the nonlinear load influence coefficient.
4. The DC power distribution network line fault diagnosis method according to claim 3, It is characterized in that The specific method for optimizing the diagnostic scheme according to the actual diagnostic effect category is as follows: When the actual diagnosis effect category is a valid category, a normal DC distribution network line fault diagnosis method is performed; When the actual diagnosis effect category is an invalid category, a first-level alarm is issued to relevant technical personnel, and the ongoing DC distribution network line fault diagnosis method is terminated; When the actual diagnosis effect category is an optimizable category, a secondary alarm is issued to the relevant technical personnel to perform process and external optimization on the ongoing DC distribution network line fault diagnosis method.
5. The DC power distribution network line fault diagnosis method according to claim 4, It is characterized in that The specific method for obtaining the power grid line fault diagnosis effect based on the actual diagnosis effect combined with the optimization effect evaluation model is as follows: The optimization effect evaluation coefficient is compared with the preset optimization effect evaluation threshold. When the optimization effect evaluation coefficient is greater than the preset optimization effect evaluation threshold, the actual diagnostic effect category of the optimized scheme is marked as a valid category. When the optimization effect evaluation coefficient is less than the preset optimization effect evaluation threshold, the actual diagnostic effect category of the optimized scheme is marked as an invalid category.
6. The DC power distribution network line fault diagnosis method according to claim 5, It is characterized in that The method for obtaining the nonlinear load influence coefficient is as follows: The voltage signal waveform of the U time interval within the G time period when diagnosing the line fault type is obtained, and the harmonic components of the voltage signal waveform are extracted in combination with Fourier transform. The amplitude of each harmonic component is calculated by spectrum analysis; the amplitude of each harmonic component is squared and added to obtain the square sum of the harmonics; the maximum amplitude of all harmonic components is obtained; the maximum amplitude of all harmonic components is compared with the square sum of the harmonics, and the nonlinear load influence coefficient is calculated.
Citation Information
Patent Citations
DC power distribution network line fault positioning method, apparatus and device, and storage medium
CN116449147A
Single-phase earth fault positioning method based on improved variational mode decomposition
CN114966324A
