Relay protection outlet signal intelligent analysis system

Through intelligent analysis system, the error and refusal events of the relay protection device are identified and adjusted, and the reliability and safety problems caused by the error and refusal events in traditional methods are solved, and real-time monitoring and optimization of complex power systems are realized. .

CN118920414AInactive Publication Date: 2024-11-08HUAIAN OF JIANGSU ELECTRIC POWER CO POWER SUPPLY

Patent Information

Application Number
CN202410843180.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-27
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional relay protection devices have false and refusal events in complex power systems, which affects the reliability and safety of the system. The existing analysis methods cannot fully cover multiple influencing factors and lack adaptability.

Method used

A relay protection exit signal intelligent analysis system is designed, including data acquisition, false event analysis, refusal event analysis, feature extraction, model training and real-time monitoring modules, and the false event and refusal are identified through filtering processing and machine learning algorithms. Causes of events and generate improvement suggestions, adjust set thresholds and optimize device layout to improve anti-interference capabilities.

Benefits of technology

The accuracy of malfunction and refusal event identification of relay protection devices is improved, the frequency of event occurrence is reduced, the reliability and stability of the power system is enhanced, and adaptive protection strategy adjustment is achieved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118920414A_ABST
    Figure CN118920414A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of signal analysis, in particular to a relay protection outlet signal intelligent analysis system, which comprises the following modules: a data acquisition module, which is used for acquiring operation data of a power system and action records of a relay protection device; the maloperation event analysis module is used for identifying the cause of the maloperation event; the action refusing event analysis module is used for identifying the reason of the action refusing event; the feature extraction module is used for extracting key feature parameters from the filtered data; the model training module is used for training a maloperation and rejection event recognition model by adopting a machine learning algorithm based on historical maloperation and rejection event data; the real-time monitoring module is used for predicting future maloperation and refusal events; and the improvement suggestion module is used for generating targeted improvement suggestions. According to the method, the identification accuracy of the maloperation and rejection events is improved, the maloperation and rejection events can be effectively reduced through dynamic adjustment of model parameters and real-time feature extraction, and the reliability and stability of a power system are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of signal analysis, and in particular to an intelligent analysis system for relay protection outlet signals. Background Art

[0002] The relay protection system is a vital component of the power system. It is responsible for quickly removing the fault area when a fault occurs, preventing the fault from spreading, and ensuring the safety and stability of the power system. However, traditional relay protection devices rely on fixed thresholds and preset logic for fault judgment, which has certain limitations. In complex power systems, traditional methods may lead to false operation and refusal to operate events, affecting the reliability and safety of the system.

[0003] A false action event refers to the erroneous action of the relay protection device when there is no fault, which may be caused by environmental factors, electromagnetic interference, equipment aging, or unreasonable threshold settings. A refusal to act event refers to the failure of the relay protection device to act in time when a fault occurs, which may be caused by fault type identification errors, protection device failures, or communication delays. In order to solve these problems and improve the operational reliability and fault handling capabilities of relay protection devices, an intelligent analysis system is urgently needed.

[0004] In the existing technology, the analysis of false operation and refusal to operate events mostly adopts a single data processing and analysis method, which cannot fully cover multiple influencing factors, resulting in insufficient accuracy and pertinence of the analysis results. In addition, the existing system lacks adaptive capabilities when dealing with the dynamically changing power system environment and cannot adjust and optimize the protection strategy in real time. Summary of the invention

[0005] Based on the above purpose, the present invention provides an intelligent analysis system for relay protection outlet signals.

[0006] Relay protection export signal intelligent analysis system, including the following modules:

[0007] Data acquisition module: used to collect power system operation data and action records of relay protection devices, including current, voltage, frequency, action time, and action type information, and filter the collected data to improve data quality;

[0008] Misoperation event analysis module: Based on the filtered data, analyze the misoperation events of the relay protection device and identify the causes of the misoperation events, including environmental factors, electromagnetic interference, equipment aging, and unreasonable threshold settings;

[0009] Refusal to operate event analysis module: Based on the filtered data, the refusal to operate event of the relay protection device is analyzed to identify the cause of the refusal to operate event, including fault type identification error, protection device failure, and communication delay;

[0010] Feature extraction module: extracts key feature parameters from the filtered data, including current waveform, harmonic components, and action time distribution;

[0011] Model training module: Based on historical misoperation and refusal event data, a machine learning algorithm is used to train the misoperation and refusal event recognition model;

[0012] Real-time monitoring module: inputs real-time data into the trained false operation and refusal to operate event recognition model, monitors the operating status of the relay protection device in real time, and predicts future false operation and refusal to operate events;

[0013] Improvement suggestion module: Generate targeted improvement suggestions based on the analysis results, including adjusting the set threshold, optimizing the equipment layout, and improving the anti-interference ability.

[0014] Furthermore, the data acquisition module specifically includes:

[0015] Various types of sensors, including current sensors, voltage sensors, and frequency sensors, are used to collect current, voltage, and frequency data of the power system in real time;

[0016] An action record collection unit is used to collect action records of relay protection devices, including action time and action type information;

[0017] A data aggregation unit aggregates the data from various sensors and action recording and collecting units, and transmits the data to the filtering processing unit through a data bus;

[0018] The filtering processing unit performs filtering on the aggregated data to eliminate noise and interference and improve data quality.

[0019] Furthermore, the malfunction event analysis module specifically includes:

[0020] Environmental monitoring unit: monitors the temperature data around the relay protection device through the temperature sensor, and analyzes the impact of temperature changes on the operation of the relay protection device;

[0021] Electromagnetic interference detection unit: Use electromagnetic interference detectors to monitor electromagnetic interference signals in the power system in real time, and determine whether electromagnetic interference is the cause of the malfunction event by analyzing the correlation between the filtered data and the electromagnetic interference signal;

[0022] Equipment status evaluation unit: Based on the historical operation data and current performance parameters of the equipment, the aging degree of the relay protection device is evaluated, and combined with the filtered data, the impact of equipment aging on false operation events is analyzed;

[0023] Threshold rationality analysis unit: Perform rationality analysis on the set threshold of the relay protection device, compare the actual operation data after filtering with the set threshold, identify whether the set threshold is reasonable, and whether there is a possibility of false operation.

[0024] Furthermore, the effect of the temperature change on the action of the relay protection device is described by the current change, and the relationship between the temperature and the current is linear, which is expressed as:

[0025] I T =I0+k T (T-T0), where I T : Current value at current temperature

[0026] I0: current value at reference temperature, k T : Temperature coefficient (current change caused by each degree of temperature change), T: current temperature, T0: reference temperature,

[0027] Determine whether the temperature change causes a false trip by calculating the current change caused by the temperature change: ΔI T =k T (T-T0), by analyzing ΔI T The phase coefficient of the false operation event is used to determine the impact of temperature changes on the relay protection device.

[0028] Furthermore, the electromagnetic interference analysis includes electromagnetic interference correlation analysis. Assuming that the electromagnetic interference signal is E(t), the current signal after filtering is I(t), and the cross-correlation function R EI (τ) to analyze the correlation between the two:

[0029] Among them, R EI (τ): cross-correlation function between electromagnetic interference signal and current signal, E(t): electromagnetic interference signal, I(t): current signal, τ: time delay parameter, when τ = 0, if R EI (0) The value is large, indicating that there is a significant correlation between the electromagnetic interference signal and the current signal, which is the cause of the false operation event.

[0030] Furthermore, the analysis of the impact of equipment aging on malfunction events includes an aging degree assessment. The change of equipment performance over time is described by an exponential decay model. Assuming that the equipment performance parameter is P(t), its change relationship is: P(t) = P0·e -λt , where P(t): performance parameter of the device at time t, P0: initial performance parameter, λ: aging rate constant, t: time;

[0031] By comparing the current performance parameter P(t) with historical data, the impact of equipment aging on false trip events can be evaluated, including the calculation of the attenuation of the performance parameter:

[0032] ΔP=P0-P(t)=P0·(1-e -λt ).

[0033] Furthermore, the threshold rationality analysis unit includes setting a threshold comparison, comparing the actual operation data {I i} and set threshold I set Compare and calculate the probability of malfunction Pmalfunction:

[0034] Among them, P 误动 : Probability of false operation, N 误动 : The number of times the actual current value exceeds the set threshold, N 总 : Total number of sampling times;

[0035] If the probability of malfunction P 误动 If it is too high, it means that the threshold setting is unreasonable and needs to be adjusted and the reasonable threshold range needs to be recalculated:

[0036] I new =μ I +k·σ I , where I new : New threshold value, μ I : Average value of current data, σ I : standard deviation of current data, k: constant.

[0037] Furthermore, the refusal event analysis module includes:

[0038] Fault type identification unit: Based on the filtered data, the fault type of the power system is analyzed through pattern recognition and feature matching algorithms to identify the fault type identification error. The pattern recognition and feature matching include using a support vector machine model to identify the fault type of the filtered data. Assuming the current data is I(t), the identified fault type is F, and the training model is obtained: F = Model (I(t)), where Model represents the trained support vector machine model. The identification result is compared with the actual fault type, and the fault type identification error rate is calculated: in, is the indicator function, when F i ≠F actual,i When , it takes 1, otherwise it takes 0, and N is the total number of samples;

[0039] Protection device status monitoring unit: real-time monitoring of the operating status and performance parameters of the relay protection device, combined with the filtered data, analyzes the probability of protection device failure, monitors the response time, contact status, internal voltage and current of the protection device, and calculates the changes in performance parameters through the filtered data. The changes in performance parameters include the response time T of the protection device. r It is expressed as:

[0040] T r =T measured -T expected , where T measured is the actual measured response time, T expected is the expected response time, if T r If it deviates from 0, there is a protection device failure;

[0041] Communication delay detection unit: monitors the communication delay between the relay protection device and the control center, identifies the refusal to operate event caused by communication delay by analyzing the filtered data and communication log. Let the communication delay be D(t), calculate the distribution and statistical characteristics of the delay through the filtered data and communication log, and calculate the average and standard deviation of the communication delay:

[0042]

[0043] Among them, μ D is the average value of communication delay, σ D is the standard deviation of communication delay, N is the total number of communication samples, and if the communication delay significantly exceeds the preset threshold, a refusal event occurs.

[0044] Further, the feature extraction module extracts the current waveform, harmonic components and action time distribution from the filtered data, specifically;

[0045] Current waveform extraction: The current signal I(t) after filtering is decomposed by discrete wavelet transform to extract the characteristics of each frequency band:

[0046] Among them, c j,k is the wavelet coefficient, ψ j,k (t) is the wavelet basis function;

[0047] Harmonic component extraction: The filtered current signal I(t) is converted to the frequency domain through fast Fourier transform to extract the harmonic components:

[0048] Where I(f) is the current signal in the frequency domain, and f is the frequency;

[0049] Action time distribution extraction: Count the action time of the relay protection device, generate an action time distribution histogram, and extract the statistical characteristics of the action time, namely the mean and variance:

[0050]

[0051] Among them, T i is the i-th action time, μ T is the mean action time, is the action time variance, and N is the number of actions.

[0052] Furthermore, the model training module uses a machine learning algorithm to train a misoperation and refusal event recognition model based on historical misoperation and refusal event data, specifically including:

[0053] The historical malfunction and refusal to operate event data are normalized to ensure the uniformity and validity of data input, and the extracted key characteristic parameters (wavelet coefficients of current waveform, amplitude and phase of harmonic components, and statistical characteristics of action time) are combined into a characteristic vector X:

[0054] X=[c j,k ,I(f),μ T ,σ T ];

[0055] Based on the support vector machine SVM, the model training is carried out, and the model training process is as follows:

[0056] Among them, w and b are model parameters, C is the regularization parameter, and y i is the label of the i-th sample, including false action or refusal to act, X i is the feature vector of the i-th sample;

[0057] Model Validation: Use cross-validation methods to evaluate the performance of the model and adjust the model parameters to achieve the best performance.

[0058] Beneficial effects of the present invention:

[0059] The present invention extracts key characteristic parameters, such as current waveform, harmonic components and action time distribution, from the filtered data. These characteristic parameters are combined with the algorithm to train the false operation and refusal event recognition model, realizing real-time monitoring and event prediction of the operating status of the relay protection device, improving the recognition accuracy of false operation and refusal events, and can effectively reduce the occurrence of false operation and refusal events by dynamically adjusting model parameters and real-time feature extraction, thereby improving the reliability and stability of the power system.

[0060] The improvement suggestion module of the present invention can generate specific improvement measures based on the analysis results of false operation and refusal to operate events, including adjusting the set threshold of the relay protection device, optimizing the equipment layout, and improving the anti-interference ability. Through detailed statistical analysis and model inference, the system can accurately identify the main causes of false operation and refusal to operate events and propose corresponding optimization strategies. For example, by recalculating the current threshold and optimizing the equipment layout according to the distribution of interference sources, the impact of environmental interference and electromagnetic interference on the equipment can be effectively reduced.

[0061] The present invention realizes intelligent monitoring and event prediction of relay protection devices through real-time data input, feature extraction and model inference. The system can automatically adjust the recognition model and threshold setting according to the dynamic environment and operation data during real-time monitoring, maintaining efficient adaptive capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0063] Figure 1 A schematic diagram of system function modules according to an embodiment of the present invention;

[0064] Figure 2 A schematic diagram of a malfunction event analysis module according to an embodiment of the present invention;

[0065] Figure 3 Schematic diagram of a refusal event analysis module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0066] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0067] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0068] like Figure 1-Figure 3 As shown in the figure, the relay protection export signal intelligent analysis system includes the following modules:

[0069] Data acquisition module: used to collect power system operation data and action records of relay protection devices, including current, voltage, frequency, action time, and action type information, and filter the collected data to improve data quality;

[0070] Misoperation event analysis module: Based on the filtered data, analyze the misoperation events of the relay protection device and identify the causes of the misoperation events, including environmental factors, electromagnetic interference, equipment aging, and unreasonable threshold settings;

[0071] Refusal to operate event analysis module: Based on the filtered data, the refusal to operate event of the relay protection device is analyzed to identify the cause of the refusal to operate event, including fault type identification error, protection device failure, and communication delay;

[0072] Feature extraction module: extracts key feature parameters from the filtered data, including current waveform, harmonic components, and action time distribution;

[0073] Model training module: Based on historical misoperation and refusal event data, a machine learning algorithm is used to train the misoperation and refusal event recognition model;

[0074] Real-time monitoring module: inputs real-time data into the trained false operation and refusal to operate event recognition model, monitors the operating status of the relay protection device in real time, and predicts future false operation and refusal to operate events;

[0075] Improvement suggestion module: Generate targeted improvement suggestions based on the analysis results, including adjusting the set threshold, optimizing the equipment layout, and improving the anti-interference ability.

[0076] The data acquisition module specifically includes:

[0077] Various types of sensors, including current sensors, voltage sensors, and frequency sensors, are used to collect current, voltage, and frequency data of the power system in real time;

[0078] An action record collection unit is used to collect action records of relay protection devices, including action time and action type information;

[0079] A data aggregation unit aggregates the data from various sensors and action recording and collecting units, and transmits the data to the filtering processing unit through a data bus;

[0080] The filtering processing unit performs filtering on the aggregated data to eliminate noise and interference and improve data quality.

[0081] The filtering processing unit includes the following components:

[0082] Low-pass filter: used to eliminate high-frequency noise interference and ensure that the collected data is smooth and stable;

[0083] High-pass filter: used to remove low-frequency interference and drift to ensure the baseline level of the signal;

[0084] Bandpass filter: used to extract signals within a specific frequency band, retain information on the fault characteristic frequency band, and exclude noise from irrelevant frequency bands;

[0085] Adaptive filter: Dynamically adjust filter parameters according to the real-time noise environment to improve the filtering effect.

[0086] The action types mainly include the following:

[0087] Overcurrent protection action: responds to the situation where the current in the power system exceeds the set value, and is used to protect lines and equipment from damage caused by short circuits and overloads.

[0088] Differential protection action: used to detect faults inside equipment such as transformers, generators, and busbars, and to perform fault detection based on the difference in current inside and outside the protection zone.

[0089] Distance protection action: Fault detection is performed based on the distance between the fault point and the protection device. It is used for the protection of transmission lines and is divided into impedance relays, directional impedance relays, etc.

[0090] Overvoltage protection action: Detects when the voltage in the power system exceeds the set value and protects the equipment from damage caused by overvoltage.

[0091] Undervoltage protection action: It responds to the situation where the voltage in the power system is lower than the set value, and is used to protect the safety of equipment such as motors under low voltage conditions.

[0092] Frequency protection action: Detects situations where the frequency in the power system deviates from the normal range, including overfrequency protection and underfrequency protection, which are often used to prevent the impact of abnormal grid frequency on system stability.

[0093] Demagnetization protection action: In case the generator loses excitation current, it is used to protect the generator and power system from the adverse effects of demagnetization.

[0094] Reclosing action: After the relay protection action clears the fault, the circuit breaker is automatically or manually reclosed to restore line operation. It can be divided into single reclosing and multiple reclosing.

[0095] The misoperation event analysis module specifically includes:

[0096] Environmental monitoring unit: monitors the temperature data around the relay protection device through the temperature sensor, and analyzes the impact of temperature changes on the operation of the relay protection device;

[0097] Electromagnetic interference detection unit: Use electromagnetic interference detectors to monitor electromagnetic interference signals in the power system in real time, and determine whether electromagnetic interference is the cause of the malfunction event by analyzing the correlation between the filtered data and the electromagnetic interference signal;

[0098] Equipment status evaluation unit: Based on the historical operation data and current performance parameters of the equipment, the aging degree of the relay protection device is evaluated, and combined with the filtered data, the impact of equipment aging on false operation events is analyzed;

[0099] Threshold rationality analysis unit: Perform rationality analysis on the set threshold of the relay protection device, compare the actual operation data after filtering with the set threshold, identify whether the set threshold is reasonable, and whether there is a possibility of false operation.

[0100] The effect of temperature change on the action of the relay protection device is described by the change of current. The relationship between temperature and current is linear, which can be expressed as:

[0101] I T =I0+k T (T-T0), where I T : Current value at current temperature

[0102] I0: current value at reference temperature, k T : Temperature coefficient (current change caused by each degree of temperature change), T: current temperature, T0: reference temperature,

[0103] Determine whether the temperature change causes a false trip by calculating the current change caused by the temperature change: ΔI T =k T (T-T0), by analyzing ΔI TThe phase coefficient of the false operation event is used to determine the impact of temperature changes on the relay protection device. T When calculating the correlation between the current change caused by temperature change and the false trip event, the correlation coefficient is used to quantify ΔI T Linear correlation with the number of false trip events:

[0104] Among them, r: Pearson correlation coefficient, X i : Current change value ΔI caused by the i-th temperature change T , Y i : The number of false events in the i-th time period, The average value of the current change caused by temperature change, The average number of false operation events. The correlation coefficient r ranges from [-1,1]. A value close to 1 or 1 indicates a strong correlation, and a value close to 0 indicates a weak correlation.

[0105] The analysis of electromagnetic interference includes electromagnetic interference correlation analysis. Let the electromagnetic interference signal be E(t), the current signal after filtering be I(t), and the cross-correlation function R EI (τ) to analyze the correlation between the two:

[0106] Among them, R EI (τ): cross-correlation function between electromagnetic interference signal and current signal, E(t): electromagnetic interference signal, I(t): current signal, τ: time delay parameter, when τ = 0, if R EI (0) The value is large, indicating that there is a significant correlation between the electromagnetic interference signal and the current signal, which is the cause of the false operation event.

[0107] The analysis of the impact of equipment aging on misoperation events includes the evaluation of the degree of aging. The change of equipment performance over time is described by the exponential decay model. Assuming that the equipment performance parameter is P(t), its change relationship is: P(t) = P0·e -λt , where P(t): performance parameter of the device at time t, P0: initial performance parameter, λ: aging rate constant, t: time;

[0108] By comparing the current performance parameter P(t) with historical data, the impact of equipment aging on false trip events can be evaluated, including the calculation of the attenuation of the performance parameter:

[0109] ΔP=P0-P(t)=P0·(1-e -λt ).

[0110] After calculating the attenuation ΔP of the performance parameter, the impact of equipment aging on false trip events can be determined by the following method:

[0111] 1. Threshold setting: Set a threshold for a device performance parameter. If the current performance parameter is lower than the threshold, it is considered that the device aging has a significant impact on the malfunction event, for example:

[0112] P threshold =P0·(1-λ threshold ), if P(t) <P threshold, , the equipment aging degree reaches the warning level.

[0113] 2. Comparison of the probability of malfunction events: Compare the probability of malfunction events under different aging degrees. Assume that the probability of malfunction events under good equipment performance and after aging is P good and P aged Statistical analysis was performed to confirm whether the two were significantly different.

[0114] 3. Regression analysis: Use regression model to analyze the relationship between the equipment aging degree λt and the number of malfunction events:

[0115] E=α+β(λt)+∈. If the regression coefficient β is significant and positive, it means that equipment aging has a significant impact on false operation events.

[0116] The threshold rationality analysis unit includes setting a threshold comparison, comparing the actual operation data after filtering {I i} and set threshold I set For comparison, calculate the probability of malfunction Pmalfunction:

[0117] Among them, P 误动 : Probability of false operation, N 误动 : The number of times the actual current value exceeds the set threshold, N 总 : Total number of sampling times;

[0118] If the probability of malfunction P 误动 If it is too high, it means that the threshold setting is unreasonable and needs to be adjusted and the reasonable threshold range needs to be recalculated:

[0119] I new =μ I +k·σ I , where I new : New threshold value, μ I : average value of current data, σ I : standard deviation of current data, k: constant, determined according to specific needs (such as 1.96 corresponds to 95% confidence interval).

[0120] The denial-of-action event analysis module includes:

[0121] Fault type identification unit: Based on the filtered data, the fault type of the power system is analyzed through pattern recognition and feature matching algorithms to identify the fault type identification error. Pattern recognition and feature matching include using a support vector machine model to identify the fault type of the filtered data. Suppose the current data is I(t), and the identified fault type is F. By training the model, we get: F = Model(I(t)), where Model represents the trained support vector machine model. The identification result is compared with the actual fault type, and the fault type identification error rate is calculated: Among them, Ⅱ is the indicator function, when F i ≠F actual,i When , it takes 1, otherwise it takes 0, and N is the total number of samples;

[0122] Protection device status monitoring unit: real-time monitoring of the operating status and performance parameters of the relay protection device, combined with the filtered data, analyzes the probability of protection device failure, monitors the response time, contact status, internal voltage and current of the protection device, and calculates the changes in performance parameters through the filtered data. The changes in performance parameters include the response time T of the protection device. r It is expressed as:

[0123] T r =T measured -T expected , where T measured is the actual measured response time, T expected is the expected response time, if T r If it deviates from 0, there is a protection device failure;

[0124] Communication delay detection unit: monitors the communication delay between the relay protection device and the control center, identifies the refusal to operate event caused by communication delay by analyzing the filtered data and communication log. Let the communication delay be D(t), calculate the distribution and statistical characteristics of the delay through the filtered data and communication log, and calculate the average and standard deviation of the communication delay:

[0125]

[0126]

[0127] Among them, μ D is the average value of communication delay, σ D is the standard deviation of communication delay, N is the total number of communication samples, and if the communication delay significantly exceeds the preset threshold, a refusal event occurs.

[0128] The feature extraction module extracts the current waveform, harmonic components and action time distribution from the filtered data, specifically;

[0129] Current waveform extraction: The current signal I(t) after filtering is decomposed by discrete wavelet transform to extract the characteristics of each frequency band:

[0130] Among them, c j,k is the wavelet coefficient, ψ j,k (t) is the wavelet basis function;

[0131] Harmonic component extraction: The filtered current signal I(t) is converted to the frequency domain through fast Fourier transform to extract the harmonic components:

[0132] Where I(f) is the current signal in the frequency domain, and f is the frequency;

[0133] Action time distribution extraction: Count the action time of the relay protection device, generate an action time distribution histogram, and extract the statistical characteristics of the action time, namely the mean and variance:

[0134]

[0135] Among them, T i is the i-th action time, μ T is the mean action time, is the action time variance, and N is the number of actions.

[0136] The model training module uses machine learning algorithms to train the misoperation and refusal event recognition model based on historical misoperation and refusal event data, including:

[0137] The historical malfunction and refusal to operate event data are normalized to ensure the uniformity and validity of data input, and the extracted key characteristic parameters (wavelet coefficients of current waveform, amplitude and phase of harmonic components, and statistical characteristics of action time) are combined into a characteristic vector X:

[0138] X=[c j,k ,I(f),μ T ,σ T ];

[0139] Based on the support vector machine SVM, the model training is carried out, and the model training process is as follows:

[0140] Among them, w and b are model parameters, C is the regularization parameter, and y i is the label of the i-th sample, including false action or refusal to act, X i is the feature vector of the i-th sample;

[0141] Model Validation: Use cross-validation methods to evaluate the performance of the model and adjust the model parameters to achieve the best performance.

[0142] The real-time feature vector X real-time Input the trained misoperation and refusal to operate event recognition model for real-time judgment. Using the support vector machine (SVM) model, the model inference process is:

[0143] y real-time =sign(w T X real-time +b), where y real-time is the real-time prediction result, and sign(·) is the sign function, indicating false action (positive) or refusal to act (negative).

[0144] The improvement suggestion module includes:

[0145] Analysis result receiving unit: receiving the analysis result of the misoperation and refusal to operate event recognition model;

[0146] Threshold setting adjustment unit: according to the analysis results, adjust the threshold setting of the relay protection device to reduce false operation and refusal to operate events;

[0147] Equipment layout optimization unit: Based on the analysis results of fault types and environmental factors, it generates equipment layout optimization suggestions to reduce the impact of environmental interference on relay protection devices;

[0148] Anti-interference capability improvement unit: Based on the electromagnetic interference analysis results, it generates suggestions for improving the anti-interference capability of the relay protection device, including measures such as increasing shielding and improving grounding.

[0149] Workflow and details:

[0150] Receiving analysis results:

[0151] The analysis result receiving unit receives the real-time analysis result from the model inference unit, including the predicted malfunction and refusal event information.

[0152] To set the threshold adjustment:

[0153] Threshold adjustment strategy: Recalculate the set threshold of the relay protection device based on the analysis results and actual operating data. The adjustment process includes transmitting the new threshold parameters to the relay protection device and updating the protection settings.

[0154] Equipment layout optimization:

[0155] Interference source identification: Identify the main interference sources around the relay protection device based on environmental factors and electromagnetic interference analysis results.

[0156] Optimization suggestion generation: Generate optimization suggestions for equipment layout based on the location of interference sources, including adjusting the equipment installation location, adding isolation barriers, and other measures. For example, for electromagnetic interference in a certain frequency band, it can be recommended to add shielding materials;

[0157] Improved anti-interference ability:

[0158] Shielding and grounding improvements: Based on the results of the electromagnetic interference analysis, it is recommended to add shielding measures around key equipment and improve equipment grounding to improve anti-interference capabilities. For example, it is recommended to use a more efficient grounding system;

[0159] Add anti-interference equipment: It is recommended to install anti-interference equipment, such as filters and anti-interference capacitors, to reduce the impact of electromagnetic interference.

[0160] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.

[0161] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the claims. Therefore, any omissions, 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. Relay protection export signal intelligent analysis system, characterized by: Includes the following modules: Data acquisition module: used to collect power system operation data and action records of relay protection devices, including current, voltage, frequency, action time, and action type information, and filter the collected data to improve data quality; Misoperation event analysis module: Based on the filtered data, analyze the misoperation events of the relay protection device and identify the causes of the misoperation events, including environmental factors, electromagnetic interference, equipment aging, and unreasonable threshold settings; Refusal to operate event analysis module: Based on the filtered data, the refusal to operate event of the relay protection device is analyzed to identify the cause of the refusal to operate event, including fault type identification error, protection device failure, and communication delay; Feature extraction module: extracts key feature parameters from the filtered data, including current waveform, harmonic components, and action time distribution; Model training module: Based on historical misoperation and refusal event data, a machine learning algorithm is used to train the misoperation and refusal event recognition model; Real-time monitoring module: inputs real-time data into the trained false operation and refusal to operate event recognition model, monitors the operating status of the relay protection device in real time, and predicts future false operation and refusal to operate events; Improvement suggestion module: Generate targeted improvement suggestions based on the analysis results, including adjusting the set threshold, optimizing the equipment layout, and improving the anti-interference ability.

2. The intelligent analysis system for relay protection output signals according to claim 1 is characterized in that: The data acquisition module specifically includes: Various types of sensors, including current sensors, voltage sensors, and frequency sensors, are used to collect current, voltage, and frequency data of the power system in real time; An action record collection unit is used to collect action records of relay protection devices, including action time and action type information; A data aggregation unit aggregates the data from various sensors and action recording and collecting units, and transmits the data to the filtering processing unit through a data bus; The filtering processing unit performs filtering on the aggregated data to eliminate noise and interference and improve data quality.

3. The intelligent analysis system for relay protection output signals according to claim 1 is characterized in that: The malfunction event analysis module specifically includes: Environmental monitoring unit: monitors the temperature data around the relay protection device through the temperature sensor, and analyzes the impact of temperature changes on the operation of the relay protection device; Electromagnetic interference detection unit: Use electromagnetic interference detectors to monitor electromagnetic interference signals in the power system in real time, and determine whether electromagnetic interference is the cause of the malfunction event by analyzing the correlation between the filtered data and the electromagnetic interference signal; Equipment status evaluation unit: Based on the historical operation data and current performance parameters of the equipment, the aging degree of the relay protection device is evaluated, and combined with the filtered data, the impact of equipment aging on false operation events is analyzed; Threshold rationality analysis unit: Perform rationality analysis on the set threshold of the relay protection device, compare the actual operation data after filtering with the set threshold, identify whether the set threshold is reasonable, and whether there is a possibility of false operation.

4. The intelligent analysis system for relay protection output signals according to claim 3 is characterized in that: The effect of temperature change on the action of the relay protection device is described by current change. The relationship between temperature and current is linear, which can be expressed as: I T =I0+k T (T-T0), where I T : Current value at current temperature I0: Current value at reference temperature, k T : Temperature coefficient (current change caused by each degree of temperature change), T: current temperature, T0: reference temperature, Determine whether the temperature change causes a false trip by calculating the current change caused by the temperature change: ΔI T =k T (T-T0), by analyzing ΔI T The phase coefficient of the false operation event is used to determine the impact of temperature changes on the relay protection device.

5. The intelligent analysis system for relay protection output signals according to claim 3 is characterized in that: The electromagnetic interference analysis includes electromagnetic interference correlation analysis. Assume that the electromagnetic interference signal is E(t), the current signal after filtering is I(t), and the cross-correlation function R EI (τ) to analyze the correlation between the two: Among them, R EI (τ): cross-correlation function between electromagnetic interference signal and current signal, E(t): electromagnetic interference signal, I(t): current signal, τ: time delay parameter, when τ = 0, if R EI (0) The value is large, indicating that there is a significant correlation between the electromagnetic interference signal and the current signal, which is the cause of the false operation event.

6. The intelligent analysis system for relay protection output signals according to claim 3 is characterized in that: The analysis of the impact of equipment aging on malfunction events includes aging degree assessment. The change of equipment performance over time is described by an exponential decay model. Assuming that the equipment performance parameter is P(t), its change relationship is: P(t)=P0·e -λt , where P(t): performance parameter of the device at time t, P0: initial performance parameter, λ: aging rate constant, t: time; By comparing the current performance parameter P(t) with historical data, the impact of equipment aging on false trip events can be evaluated, including the calculation of the attenuation of the performance parameter: ΔP=P0-P(t)=P0·(1-e -λt )。 7. The intelligent analysis system for relay protection output signals according to claim 3 is characterized in that: The threshold rationality analysis unit includes setting a threshold comparison, comparing the actual operation data after filtering i + and set threshold I set Compare and calculate the probability of malfunction Pmalfunction: Among them, P 误动 : Probability of false operation, N 误动 : The number of times the actual current value exceeds the set threshold, N 总 : Total number of sampling times; If the probability of malfunction P 误动 If it is too high, it means that the threshold setting is unreasonable and needs to be adjusted and the reasonable threshold range needs to be recalculated: I new =μ I +k·σ I , where I new : New threshold value, μ I : average value of current data, σ I : standard deviation of current data, k: constant.

8. The intelligent analysis system for relay protection output signals according to claim 1 is characterized in that: The refusal event analysis module includes: Fault type identification unit: Based on the filtered data, the fault type of the power system is analyzed through pattern recognition and feature matching algorithms to identify the fault type identification error. The pattern recognition and feature matching include using a support vector machine model to identify the fault type of the filtered data. Assuming the current data is I(t), the identified fault type is F, and the training model is obtained: F = Model (I(t)), where Model represents the trained support vector machine model. The identification result is compared with the actual fault type, and the fault type identification error rate is calculated: Among them, II is the indicator function, when F i ≠F actual,i When , it takes 1, otherwise it takes 0, and N is the total number of samples; Protection device status monitoring unit: real-time monitoring of the operating status and performance parameters of the relay protection device, combined with the filtered data, analyzes the probability of protection device failure, monitors the response time, contact status, internal voltage and current of the protection device, and calculates the changes in performance parameters through the filtered data. The changes in performance parameters include the response time T of the protection device. r It is expressed as: T r =T measured -T expected , where T measured is the actual measured response time, T expected is the expected response time, if T r If it deviates from 0, there is a protection device failure; Communication delay detection unit: monitors the communication delay between the relay protection device and the control center, identifies the refusal to operate event caused by communication delay by analyzing the filtered data and communication log. Let the communication delay be D(t), calculate the distribution and statistical characteristics of the delay through the filtered data and communication log, and calculate the average and standard deviation of the communication delay: Among them, μ D is the average value of communication delay, σ D is the standard deviation of communication delay, N is the total number of communication samples, and if the communication delay significantly exceeds the preset threshold, a refusal event occurs.

9. The intelligent analysis system for relay protection output signals according to claim 1, characterized in that: The feature extraction module extracts the current waveform, harmonic components and action time distribution from the filtered data, specifically; Current waveform extraction: The current signal I(t) after filtering is decomposed by discrete wavelet transform to extract the characteristics of each frequency band: Among them, c j,k is the wavelet coefficient, ψ j,k (t) is the wavelet basis function; Harmonic component extraction: The filtered current signal I(t) is converted to the frequency domain through fast Fourier transform to extract the harmonic components: Where I(f) is the current signal in the frequency domain, and f is the frequency; Action time distribution extraction: Count the action time of the relay protection device, generate an action time distribution histogram, and extract the statistical characteristics of the action time, namely the mean and variance: Among them, T i is the i-th action time, μ T is the mean action time, is the action time variance, and N is the number of actions.

10. The intelligent analysis system for relay protection output signals according to claim 9, characterized in that: The model training module uses a machine learning algorithm to train a misoperation and refusal event recognition model based on historical misoperation and refusal event data, specifically including: The historical misoperation and refusal to operate event data are normalized to ensure the uniformity and validity of data input, and the extracted key feature parameters are combined into a feature vector X: X=[c j,k ,I(f),μ T ,s T ]; Based on the support vector machine SVM, the model training is carried out, and the model training process is as follows: Among them, w and b are model parameters, C is the regularization parameter, and y i is the label of the i-th sample, including false action or refusal to act, X i is the feature vector of the i-th sample; Model Validation: Use cross-validation methods to evaluate the performance of the model and adjust the model parameters to achieve the best performance.

Citation Information

Patent Citations

  • On-line monitoring method of state of relay protection equipment

    CN102005736A

  • Multi-mode hidden-fault risk analysis method of relaying protection system

    CN103488873A

  • Relay protection judgment method based on dynamic tendency analysis

    CN104915777A

  • Protection motion behavior intelligent analysis method based on multisource wave recording files

    CN105353753A

  • Method and device for determining operation state of relay protection equipment

    CN117791525A

Cited By

  • Relay protection fault identification and response system

    CN119852950A

  • Method for preventing relay protection misoperation

    CN121192606A

  • A method for preventing maloperation of a relay protection

    CN121192606B