RCBO-oriented fault location distribution box control method and system

Abstract

The RCBO-oriented fault positioning distribution box control method and system of the present application belongs to the intelligent power distribution and fault diagnosis technical field of electrical engineering field, comprising the following steps: S1, collecting three-phase current signals and residual current signals of the power distribution system; S2, calculating the harmonic distortion rate and load imbalance degree based on the three-phase current signals; S3, calculating the residual current waveform complexity index based on the residual current signals; S4, combining the harmonic distortion rate, load imbalance degree, residual current waveform complexity index and preset basic threshold to generate an adaptive threshold; S5, according to the value of the residual current waveform complexity index, selecting one of the multiple preset fault judgment modes; S6, using the selected fault judgment mode to process the residual current signals to generate a fault judgment result, and at least one fault judgment mode uses the adaptive threshold, which provides a solid data foundation for subsequent adaptive adjustment.

Description

Technical Field

[0001] This invention relates to the field of intelligent power distribution and fault diagnosis in electrical engineering, specifically to a fault location distribution box control method and system for RCBO. Background Technology

[0002] In industrial power distribution scenarios with multiple branches connected in parallel and high-power nonlinear loads, traditional fault location methods based on residual current devices mainly rely on fixed fault judgment thresholds.

[0003] This situation poses a serious challenge to the accuracy and reliability of the method in complex industrial power environments. Specifically, the current waveform distortion and three-phase load imbalance are common in the power grid. These disturbances seriously interfere with the fault judgment mechanism of the traditional RCBO, thereby reducing the accuracy of fault location and frequently causing misjudgments, non-fault tripping or missed judgments, and failure to detect real faults.

[0004] The root cause of the above problems lies in the limitations of traditional technical methods. The fixed threshold judgment mechanism used cannot dynamically adapt to the complex electromagnetic environment caused by factors such as harmonics and load imbalance. When the intensity of background noise and interference increases and the waveform becomes more complex, the judgment method based solely on current amplitude will lose its reliability. As a result, in the current industrial power distribution scenario, traditional methods are unable to simultaneously ensure sensitive response to real faults and effective suppression of various electrical noises. This directly affects the overall accuracy of fault location and reduces the reliability of power distribution system operation.

[0005] The information disclosed in the background section above is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a fault location control method and system for RCBO-oriented distribution boxes to solve the problems mentioned in the background art.

[0007] The technical solution of the present invention includes the following steps:

[0008] S1. Acquire the three-phase current signal and residual current signal of the power distribution system;

[0009] S2. Calculate the harmonic distortion rate and load imbalance based on the three-phase current signal;

[0010] S3. Calculate the residual current waveform complexity index based on the residual current signal;

[0011] S4. Combine the harmonic distortion rate, load imbalance, residual current waveform complexity index and preset basic thresholds to generate an adaptive threshold.

[0012] S5. Select one of several preset fault judgment modes based on the value of the residual current waveform complexity index.

[0013] S6. Using the selected fault judgment mode, process the residual current signal to generate a fault judgment result. At least one fault judgment mode uses an adaptive threshold.

[0014] S7. Based on the fault diagnosis results, output control commands.

[0015] Preferably, S4 specifically includes:

[0016] S41. Calculate the foundation compensation coefficient based on harmonic distortion rate and load imbalance.

[0017] S42. Calculate the dynamic adjustment coefficient based on the basic compensation coefficient and the residual current waveform complexity index;

[0018] S43. Multiply the base threshold by the dynamic adjustment coefficient to generate the adaptive threshold.

[0019] Preferably, S3 specifically includes:

[0020] S31. Apply wavelet packet decomposition to the residual current signal to obtain the energy distribution in multiple frequency sub-bands;

[0021] S32. Calculate the Shannon entropy of the energy distribution;

[0022] S33. Normalize the Shannon entropy to obtain the residual current waveform complexity index.

[0023] Preferably, S5 and S6 specifically include:

[0024] When the residual current waveform complexity index is lower than the first preset threshold, the low complexity mode is selected, and in this mode, the real-time residual current amplitude is compared with the basic threshold.

[0025] When the residual current waveform complexity index is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the medium complexity mode is selected. In this mode, the real-time residual current amplitude is compared with the adaptive threshold.

[0026] When the residual current waveform complexity index is higher than the second preset threshold, the high complexity mode is selected, and wavelet feature matching is performed in this mode.

[0027] Preferably, wavelet feature matching specifically includes:

[0028] S61. Extract the wavelet packet energy spectrum of the residual current signal as the feature vector to be measured;

[0029] S62. Calculate the similarity between the feature vector to be tested and each reference vector in the preset real fault wavelet feature library;

[0030] S63. When the highest value in the similarity score exceeds the preset matching confidence threshold, a real fault is determined to exist.

[0031] Preferably, the basic compensation coefficient is calculated by using a preset basic error compensation model. The basic compensation coefficient is linearly positively correlated with the harmonic distortion rate and the load imbalance.

[0032] Preferably, the dynamic adjustment coefficient is calculated by multiplying the basic compensation coefficient by the amplification term driven by the residual current waveform complexity index.

[0033] The fault location distribution box control system for RCBO includes:

[0034] The data acquisition module is used to acquire the three-phase current signal and residual current signal of the power distribution system;

[0035] The feature quantization module, connected to the data acquisition module, is used to calculate the harmonic distortion rate and load imbalance based on the three-phase current signal, and to calculate the residual current waveform complexity index based on the residual current signal.

[0036] The dynamic compensation module, connected to the feature quantization module, is used to combine harmonic distortion rate, load imbalance, residual current waveform complexity index and preset basic threshold to generate an adaptive threshold.

[0037] The fault diagnosis module, connected to the feature quantization module and the dynamic compensation module, is used to select the fault diagnosis mode based on the residual current waveform complexity index, and to process the residual current signal using the selected mode to output control commands.

[0038] This invention provides an improved method and system for fault location distribution box control for RCBO, which, compared with the prior art, has the following improvements and advantages:

[0039] 1. This scheme achieves comprehensive quantification of power grid disturbance status through the acquisition and analysis of three-phase current signals and residual current signals. By introducing the calculation of harmonic distortion rate and load imbalance, this scheme, for the first time, incorporates the degree of disturbance in the macroscopic power grid environment into the fault judgment considerations. Furthermore, by introducing a residual current waveform complexity index based on wavelet packet energy entropy, this scheme can accurately assess the distortion and contamination degree of the residual current waveform from a signal morphology perspective. This multi-dimensional feature quantification enables the system to transform from a black-box state to a white-box state with deep insight into the power grid environment, providing a solid data foundation for subsequent adaptive adjustments.

[0040] 2. This solution constructs a complete adaptive threshold generation logic, which changes the rigid mode of existing technologies that rely on fixed thresholds. It can significantly improve the judgment accuracy under complex working conditions, greatly reduce the false alarm rate caused by environmental disturbances, and at the same time ensure a reliable response to real faults.

[0041] 3. This solution innovatively designs a multi-mode fault judgment strategy based on the residual current waveform complexity index, realizing a leap from signal processing to pattern recognition. This method can accurately identify the real fault signal characteristics submerged by noise under extremely harsh and strong interference backgrounds, solving the technical problem that existing technologies completely fail under such extreme conditions. Attached Figure Description

[0042] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0043] Figure 1 This is a flowchart of the fault location distribution box control method for RCBO according to the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0045] Example 1:

[0046] Please see Figure 1 This invention provides a fault location control method for distribution boxes for RCBO (Responsive Control Box), comprising the following steps:

[0047] S1. Acquire the three-phase current signal and residual current signal of the power distribution system;

[0048] S2. Calculate the harmonic distortion rate and load imbalance based on the three-phase current signal;

[0049] S3. Calculate the residual current waveform complexity index based on the residual current signal;

[0050] S4. Combine the harmonic distortion rate, load imbalance, residual current waveform complexity index and preset basic thresholds to generate an adaptive threshold.

[0051] S5. Select one of several preset fault judgment modes based on the value of the residual current waveform complexity index.

[0052] S6. Using the selected fault judgment mode, process the residual current signal to generate a fault judgment result. At least one fault judgment mode uses an adaptive threshold.

[0053] S7. Based on the fault diagnosis result, output control commands;

[0054] A fault location control method for distribution boxes based on RCBO (Responsive Control Box) is applied to industrial power distribution scenarios with multiple parallel branches and high-power nonlinear loads. In such scenarios, grid current waveform distortion and three-phase load imbalance are common. These factors severely interfere with the traditional RCBO fault judgment mechanism that relies on fixed thresholds, thereby reducing the accuracy of fault location and causing misjudgments or missed judgments. To solve this technical problem, this method quantifies the key disturbance factors of the power distribution system in real time and constructs a dynamic compensation model and multi-mode judgment strategy based on the quantification results to achieve high adaptability to complex electromagnetic environments. The achieved technical effect is that while ensuring sensitive response to real faults, it effectively suppresses non-faulty tripping caused by background noise, harmonic interference, and load imbalance, thereby significantly improving the overall accuracy of fault location and operational reliability of industrial power distribution systems.

[0055] This solution achieves comprehensive quantification of grid disturbance status through the acquisition and analysis of three-phase current signals and residual current signals. Existing RCBO fault judgment technologies typically rely solely on monitoring the residual current amplitude, lacking effective identification and quantification methods for the root causes of abnormal residual current fluctuations, such as grid background harmonics and three-phase load imbalance. This solution, by introducing the calculation of harmonic distortion rate and load imbalance, is the first to incorporate the degree of disturbance in the macroscopic grid environment into the fault judgment considerations. Furthermore, by introducing a residual current waveform complexity index based on wavelet packet energy entropy, this solution can accurately assess the distortion and contamination degree of the residual current waveform from a signal morphology perspective. This multi-dimensional feature quantification enables the system to transform from a black-box state to a white-box state with deep insight into the grid environment, providing a solid data foundation for subsequent adaptive adjustments—a fundamental difference unmatched by existing technologies.

[0056] Example 2:

[0057] S3 specifically includes:

[0058] S31. Apply wavelet packet decomposition to the residual current signal to obtain the energy distribution in multiple frequency sub-bands;

[0059] S32. Calculate the Shannon entropy of the energy distribution;

[0060] S33. Normalize the Shannon entropy to obtain the residual current waveform complexity index.

[0061] In this embodiment, the calculation of the residual current waveform complexity index aims to introduce a quantitative index to accurately characterize the degree to which the residual current waveform deviates from the ideal shape. The calculation process applies wavelet packet decomposition to the acquired residual current signal, mapping the total signal energy to a series of discrete frequency sub-bands, forming an energy spectrum distribution. Based on this energy spectrum distribution, its Shannon entropy, i.e., wavelet packet energy entropy, is calculated. A spectrum with energy uniformly dispersed across all sub-bands corresponds to a high entropy value, indicating mixed signal components and a complex waveform shape. Conversely, a spectrum with energy highly concentrated in a few sub-bands corresponds to a low entropy value, indicating a relatively pure waveform shape. To eliminate the influence of dimensions and provide a unified benchmark for subsequent mode switching, this entropy value is normalized to generate a dimensionless index between 0 and 1, i.e., the residual current waveform complexity index. The generation of this index provides a quantitative basis for subsequent adaptive threshold dynamic adjustment and fault judgment mode selection, and is a key link in realizing the upgrade from amplitude judgment to shape recognition.

[0062] Example 3:

[0063] S4 specifically includes:

[0064] S41. Calculate the foundation compensation coefficient based on harmonic distortion rate and load imbalance.

[0065] S42. Calculate the dynamic adjustment coefficient based on the basic compensation coefficient and the residual current waveform complexity index;

[0066] S43. Multiply the base threshold by the dynamic adjustment coefficient to generate an adaptive threshold;

[0067] The basic compensation coefficient is calculated by using a preset basic error compensation model. The basic compensation coefficient is linearly positively correlated with the harmonic distortion rate and the load imbalance.

[0068] The dynamic adjustment coefficient is calculated by multiplying the basic compensation coefficient by the amplification term driven by the residual current waveform complexity exponent.

[0069] In this embodiment, the generation of the adaptive threshold is implemented as a logically progressive dynamic compensation calculation process, with the goal of producing a dynamic fault judgment threshold that can reflect the real-time state of power grid disturbances. ;

[0070] The first stage of this process is based on the compensation coefficient. The calculation; the technical motivation lies in the intensity of macroscopic disturbances, i.e., harmonic distortion rate. With load imbalance Preliminary baseline corrections are performed because these two disturbances are the main factors that reduce the system's signal-to-noise ratio and cause misjudgments; the calculation follows a preset basic error compensation model:

[0071] Basic compensation coefficient, harmonic distortion rate, load unbalance weighting coefficient:

[0072] ;

[0073] in, , Load imbalance Weighting coefficient;

[0074] in, and The values ​​were obtained through Fourier transform and the ratio of the range to the mean of the effective values ​​of the three-phase currents, respectively. The calculated results were directly substituted into the numerical values ​​corresponding to the percentages; weighting coefficients. and These are dimensionless preset parameters. The method for determining their values ​​is based on statistical regression analysis of a large amount of historical operating data. For example, if statistical patterns show that for every 10 percentage point increase in harmonic distortion rate, the fault location accuracy decreases by approximately 15%, then a parameter can be set. The initial value is 0.015;

[0075] To provide a clear, non-inventive calibration method for those skilled in the art to determine weighting coefficients. and The steps may include:

[0076] Data preparation: Construct a data set containing... Test set of historical data samples:

[0077] ;

[0078] in, and The first Harmonic distortion rate and load imbalance of each sample; Given a binary label, when traditional RCBO fails to detect misclassification or false negatives under this sample condition, ,otherwise ; This represents the total number of historical data samples. The index of the sample, from 1 to ;

[0079] Model fitting: Standard statistical models such as logistic regression are used to assess the probability of misclassification. Perform a fitting; the model can be represented as:

[0080] ;

[0081] in, Let be the probability of misjudgment under a given harmonic distortion rate and load imbalance. is the base of the natural logarithm. The optimal fit coefficients are obtained from the test set using methods such as maximum likelihood estimation. Harmonic distortion rate, To determine the load imbalance, the optimal fit coefficients are obtained from the test set using methods such as maximum likelihood estimation. and ;

[0082] Coefficient mapping: mapping the fitted coefficients As a weighting coefficient The benchmark; for example, it can be directly set as follows. , This method transforms the process of determining coefficients into a standard, repeatable statistical analysis task, thereby avoiding excessive experimentation.

[0083] Based on the basic compensation, a waveform complexity index is introduced. A second fine-tuning process is performed to generate dynamic adjustment coefficients. The technical motivation for this step is that... It only reflects the overall energy intensity of the disturbance, failing to capture the waveform structure of the disturbance. This precisely compensates for this deficiency, enabling the quantification of the degree to which fault characteristics are obscured by complex noise; the calculation formula is defined as:

[0084] Dynamic adjustment coefficient, basic compensation coefficient, waveform complexity index, complexity influencing factor:

[0085] ;

[0086] in, , , Complexity influencing factor;

[0087] in, This is the output of the previous stage of calculation. The normalized index generated in the preceding steps; influence factor This is a dimensionless preset parameter used for adjustment. The contribution to the overall compensation magnitude; its setting logic is based on experimental data, for example, if experiments show that when If the traditional amplitude determination method fails, then It needs to be set to be sufficient at this height Values ​​that exhibit a significant amplification effect within an interval, such as The introduction of this coefficient makes It can more sensitively reflect the true difficulty of signal identification and achieves two-dimensional perturbation compensation from energy to form;

[0088] To specifically calibrate the complexity impact factor The following experimental steps can be used:

[0089] Constructing a test signal: Prepare a set of signals with known amplitudes. Typical real fault current signals are generated; white noise or harmonic interference of different intensities are superimposed onto this set of signals by a signal generator to generate a series of signals with different waveform complexity indices. For example, the signal to be measured can be in increments of 0.1, ranging from 0.4 to 0.8.

[0090] Determine the minimum detectable threshold: without employing the compensation mechanism of this invention, i.e., using only the basic threshold. Adjust the overall amplitude of the signal under test to find the value of each... At this level, the minimum current amplitude that can be reliably detected is denoted as... ;

[0091] calculate Value: According to the adaptive threshold formula of the present invention:

[0092] ;

[0093] Theoretically, it should have In the experimental environment, Given, It can be calculated based on the current background interference; therefore, for each A solution can be found. :

[0094] ;

[0095] For all calculated By calculating the average value or selecting the maximum value based on the worst-case scenario, a reasonable value supported by experimental data can be obtained. As a preset parameter; : in each The minimum current amplitude that can be reliably detected at this level; Basic compensation coefficient; : No. The waveform complexity index of a signal under test;

[0096] Using dynamic adjustment coefficients Generate adaptive threshold :

[0097] Adaptive threshold base threshold dynamic adjustment coefficient:

[0098] ;

[0099] in, ;

[0100] in, It is a fundamental threshold with a clear physical meaning, such as the rated residual operating current set according to safety regulations, whose dimension is milliampere; this formula combines this fundamental threshold with the dimensionless comprehensive disturbance coefficient calculated in the previous steps. Multiplication, product It is a with A dynamic threshold with the same physical dimensions that changes in real time with the power grid environment; the generation of this adaptive threshold makes the fault judgment benchmark no longer rigid, but can dynamically adapt to the power grid operating conditions, thereby achieving accurate fault location in a high-interference environment.

[0101] This solution constructs a complete adaptive threshold generation logic, changing the rigid mode of existing technologies that rely on fixed thresholds; the core of this logic lies in its dynamic compensation mathematical model, and its derivation process has clear physical meaning and technical rationality.

[0102] Formula for calculating the basic compensation coefficient ; Basic compensation coefficient, Harmonic distortion rate Load imbalance The practical significance of the weighting coefficient lies in the fact that it establishes a compensation benchmark that adjusts according to the intensity of macroscopic disturbances; when the power grid environment deteriorates, i.e. or When it increases, The coefficient increases linearly; this increase will raise the final judgment threshold in subsequent calculations, thereby actively avoiding misjudgments caused by strong background noise.

[0103] Formula for calculating dynamic adjustment coefficient ; Dynamic adjustment coefficient Basic compensation coefficient, Residual current waveform complexity index The practical significance of the complexity influencing factor lies in the fact that, based on macroscopic compensation, it introduces fine-tuning of the quality of specific signal waveforms; Amplification terms of the drive It can amplify the compensation level a second time according to the complexity of the waveform; a noise waveform with a complex shape (high...) This will be more effective than a simpler form of noise with the same amplitude (lower noise level). This leads to a more significant threshold adjustment;

[0104] Adaptive threshold formula ; Adaptive threshold Baseline threshold The dynamic adjustment coefficient is a dimensionless coefficient that changes in real time with the operating conditions. Acting on a fundamental threshold with a clear physical meaning Above, a dynamic judgment benchmark with physical meaning, measured in mA, and capable of adapting to the power grid environment in real time was generated. Compared to the fixed threshold of existing technologies, this adaptive threshold can significantly improve the accuracy of judgment under complex working conditions, greatly reduce the false alarm rate caused by environmental disturbances, and ensure a reliable response to real faults.

[0105] Example 4:

[0106] S5 and S6 specifically include:

[0107] When the residual current waveform complexity index is lower than the first preset threshold, the low complexity mode is selected, and in this mode, the real-time residual current amplitude is compared with the basic threshold.

[0108] When the residual current waveform complexity index is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the medium complexity mode is selected. In this mode, the real-time residual current amplitude is compared with the adaptive threshold.

[0109] When the residual current waveform complexity index is higher than the second preset threshold, the high complexity mode is selected, and wavelet feature matching is performed in this mode.

[0110] Wavelet feature matching specifically includes:

[0111] S61. Extract the wavelet packet energy spectrum of the residual current signal as the feature vector to be measured;

[0112] S62. Calculate the similarity between the feature vector to be tested and each reference vector in the preset real fault wavelet feature library;

[0113] S63. When the highest value in the similarity score exceeds the preset matching confidence threshold, it is determined that there is a real fault.

[0114] The method for determining the preset matching confidence threshold aims to maximize the identification rate of real faults while minimizing the misjudgment rate of non-faulty, high-interference signals; a standard method for determining it is as follows:

[0115] Prepare a validation dataset: Prepare a validation dataset that is independent of the feature library and consists of two parts: one part is real fault signals of various known types, positive samples, and the other part is non-fault operation signals under high harmonics and high imbalance, negative samples.

[0116] Calculate the similarity distribution: Take the feature vectors of all samples in the validation dataset and calculate the similarity with all reference vectors in the feature library, and record the highest similarity value obtained for each sample; this will result in two sets of similarity score distributions: one set from positive samples and the other set from negative samples;

[0117] Determine the optimal threshold: Analyze the receiver operation characteristic curve by plotting it. The curve has the false positive rate (the proportion of negative samples that are misclassified as faulty) on the horizontal axis and the true positive rate (the proportion of positive samples that are correctly identified) on the vertical axis. Select the similarity value corresponding to the point on the curve closest to the top left corner, or select a similarity value that achieves the best balance between false positives and false negatives based on the tolerance for false negatives in actual applications, as the matching confidence threshold.

[0118] In this embodiment, based on waveform complexity index The multi-mode fault diagnosis logic is based on the following: The signal quality it represents automatically switches between three preset modes. The determination of the first and second preset thresholds is based on statistical analysis of a large amount of experimental data. For example, they can be set to 0.3 and 0.6 to define three ranges of signal quality from pure to moderate interference to severe contamination.

[0119] when When the value is below 0.3, the system enters a low-complexity mode. In this mode, because the residual current waveform is considered relatively pure with minimal interference, the system employs high-response-speed judgment logic to directly compare the real-time residual current amplitude with a fixed baseline threshold. Comparisons are made; this strategy ensures that faults can be detected as quickly as possible when the power grid environment is favorable.

[0120] when When the value is between 0.3 and 0.6, the system switches to a medium complexity mode; this range indicates that there is significant background noise and interference in the waveform; at this time, the system activates the aforementioned dynamic compensation mechanism, which combines the real-time residual current amplitude with a dynamically generated adaptive threshold. Compare; because The noise level has been moderately raised to compensate for moderate disturbances, and this mode can effectively avoid false alarms caused by background noise.

[0121] when When the amplitude is above 0.6, the system enters a high-complexity mode. Under this condition, the waveform is considered severely contaminated, and the amplitude-based judgment method loses its reliability. The judgment logic then switches to wavelet feature matching based on pattern recognition. The workflow of this mode is as follows: extract the wavelet packet energy spectrum of the current residual current signal and construct it as the feature vector to be tested; compare it with an internally preset real fault wavelet feature library, which stores wavelet packet energy spectrum reference vectors representing various typical leakage fault waveforms; calculate the similarity between the vector to be tested and each reference vector in the library, such as cosine similarity, and determine it as a real fault only when the highest similarity exceeds a preset matching confidence threshold. This method of identification based on waveform morphology rather than amplitude exhibits high robustness in high-noise backgrounds, ensuring accurate identification of real faults even in harsh electromagnetic environments.

[0122] The construction process of the pre-defined real fault wavelet feature library is crucial to the feasibility of this invention. The construction method includes the following steps:

[0123] Define the fault type: Identify the typical leakage fault types that need to be identified, such as: high-resistivity grounding fault, intermittent arcing grounding fault, and capacitive leakage fault caused by equipment insulation aging;

[0124] Obtain fault waveform samples: Obtain current waveform data for the above fault types through one or a combination of the following two methods:

[0125] Simulation modeling: Using circuit simulation software such as MATLAB / Simulink, a power distribution system model containing nonlinear loads, such as frequency converters and switching power supplies, is built. By setting different grounding resistances, arc models, and ground capacitances in the model, residual current waveform data under various typical faults are simulated and generated.

[0126] Experimental platform setup: Establish a physical experimental platform, reproduce various faults under real load conditions through a controllable fault injection device, such as a resistor grounding circuit controlled by a MOSFET, and use a high-precision data acquisition card to record the residual current waveform;

[0127] Extract and store feature vectors: For each acquired fault waveform sample, perform the same processing as in step S31, i.e., apply wavelet packet decomposition to obtain its energy distribution in multiple frequency subbands; normalize this energy distribution vector to form a feature vector of standard length; store this feature vector in pairs with its corresponding fault type label, such as high resistance fault or arc fault. All these paired data together constitute the real fault wavelet feature library; this library should contain at least dozens of samples for each typical fault type to ensure its diversity and representativeness.

[0128] This solution innovatively designs a multi-mode fault judgment strategy based on the residual current waveform complexity index; the advancement of this strategy lies in its recognition that the optimal fault judgment method is different under different signal quality.

[0129] when When the value is low, the system selects a mode that compares the real-time residual current amplitude with the base threshold; this mode has a fast response speed and is suitable for operating conditions with a clean power grid environment, ensuring rapid isolation of interference-free faults;

[0130] when When the value is within a moderate range, the system switches to a mode that compares the real-time residual current amplitude with an adaptive threshold; this mode is the core operating mode of this scheme, which utilizes the aforementioned generated... It effectively filters out moderate levels of noise interference, achieving the best balance between ensuring stable system operation and achieving accurate fault location;

[0131] when When the value is extremely high, the system determines that any comparison based on amplitude is unreliable and switches to wavelet feature matching mode. In this mode, the system no longer focuses on the amplitude of the residual current, but on its shape. By extracting the wavelet packet energy spectrum of the signal under test as a feature vector and performing similarity calculation with a preset real fault wavelet feature library, a leap from signal processing to pattern recognition is achieved. This method can accurately identify the real fault signal features submerged in noise under extremely harsh and strong interference backgrounds, solving the technical problem that existing technologies completely fail under such extreme conditions.

[0132] Example 5:

[0133] The fault location distribution box control system for RCBO includes:

[0134] The data acquisition module is used to acquire the three-phase current signal and residual current signal of the power distribution system;

[0135] The feature quantization module, connected to the data acquisition module, is used to calculate the harmonic distortion rate and load imbalance based on the three-phase current signal, and to calculate the residual current waveform complexity index based on the residual current signal.

[0136] The dynamic compensation module, connected to the feature quantization module, is used to combine harmonic distortion rate, load imbalance, residual current waveform complexity index and preset basic threshold to generate an adaptive threshold.

[0137] The fault judgment module, connected to the feature quantization module and the dynamic compensation module, is used to select the fault judgment mode based on the residual current waveform complexity index, and process the residual current signal using the selected mode to output control commands.

[0138] A fault location distribution box control system for RCBO (Residual Current Box) is a physical implementation of the aforementioned control method, consisting of a series of functionally coupled modules. The data acquisition module, as the system's signal input, is responsible for acquiring the instantaneous values ​​of the three-phase current and residual current of the distribution system in real time. The feature quantization module, connected to the data acquisition module, performs signal processing and calculations, converting the raw current signal into three key quantification indicators: harmonic distortion rate... Load imbalance and residual current waveform complexity index The dynamic compensation module receives these three metrics from the feature quantization module and, based on its built-in compensation model, calculates and outputs an adaptive threshold in real time. The fault diagnosis module, as the system's control and decision-making unit, is connected to both the feature quantization module and the dynamic compensation module. Based on the received data... Value selection judgment mode, and call The built-in wavelet feature library can complete the final fault assessment and then output control commands to the RCBO actuator; the data flow and logical connection between the modules together constitute a closed-loop adaptive control system, realizing the automation of fault diagnosis logic;

[0139] This solution introduces multi-dimensional quantification of the power grid environment and signal patterns, and establishes a dynamic adaptive threshold and intelligent multi-mode judgment mechanism based on this, forming a complete closed-loop adaptive control system. Its beneficial effects are that, compared with existing technologies, this solution has achieved a substantial improvement in the accuracy, reliability and environmental adaptability of fault location. Especially in industrial application scenarios with concentrated high-power nonlinear loads, it can significantly reduce the false positive and false negative rates of RCBO, bringing huge economic benefits and safety value.

[0140] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A fault location distribution box control method for RCBO, characterized in that, Includes the following steps: S1. Acquire the three-phase current signal and residual current signal of the power distribution system; S2. Calculate the harmonic distortion rate and load imbalance based on the three-phase current signal; S3. Calculate the residual current waveform complexity index based on the residual current signal; S4. Combine the harmonic distortion rate, load imbalance, residual current waveform complexity index and preset basic thresholds to generate an adaptive threshold. S5. Select one of several preset fault judgment modes based on the value of the residual current waveform complexity index. S6. Using the selected fault judgment mode, process the residual current signal to generate a fault judgment result. At least one fault judgment mode uses an adaptive threshold. S7. Based on the fault diagnosis result, output control commands; S5 and S6 specifically include: When the residual current waveform complexity index is lower than the first preset threshold, the low complexity mode is selected, and in this mode, the real-time residual current amplitude is compared with the basic threshold. When the residual current waveform complexity index is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the medium complexity mode is selected. In this mode, the real-time residual current amplitude is compared with the adaptive threshold. When the residual current waveform complexity index is higher than the second preset threshold, the high complexity mode is selected, and wavelet feature matching is performed in this mode.

2. The fault location distribution box control method for RCBO according to claim 1, characterized in that, S4 specifically includes: S41. Calculate the foundation compensation coefficient based on harmonic distortion rate and load imbalance. S42. Calculate the dynamic adjustment coefficient based on the basic compensation coefficient and the residual current waveform complexity index; S43. Multiply the base threshold by the dynamic adjustment coefficient to generate the adaptive threshold.

3. The fault location distribution box control method for RCBO according to claim 1, characterized in that, S3 specifically includes: S31. Apply wavelet packet decomposition to the residual current signal to obtain the energy distribution in multiple frequency sub-bands; S32. Calculate the Shannon entropy of the energy distribution; S33. Normalize the Shannon entropy to obtain the residual current waveform complexity index.

4. The fault location distribution box control method for RCBO according to claim 1, characterized in that, Wavelet feature matching specifically includes: S61. Extract the wavelet packet energy spectrum of the residual current signal as the feature vector to be measured; S62. Calculate the similarity between the feature vector to be tested and each reference vector in the preset real fault wavelet feature library; S63. When the highest value in the similarity score exceeds the preset matching confidence threshold, a real fault is determined to exist.

5. The fault location distribution box control method for RCBO according to claim 2, characterized in that, The basic compensation coefficient is calculated by using a preset basic error compensation model. The basic compensation coefficient is linearly positively correlated with the harmonic distortion rate and the load imbalance.

6. The fault location distribution box control method for RCBO according to claim 2, characterized in that, The dynamic adjustment coefficient is calculated by multiplying the basic compensation coefficient by the amplification term driven by the residual current waveform complexity index.

7. A fault location distribution box control system for RCBO, based on the fault location distribution box control method for RCBO as described in any one of claims 1-6, characterized in that, include: The data acquisition module is used to acquire the three-phase current signal and residual current signal of the power distribution system; The feature quantization module, connected to the data acquisition module, is used to calculate the harmonic distortion rate and load imbalance based on the three-phase current signal, and to calculate the residual current waveform complexity index based on the residual current signal. The dynamic compensation module, connected to the feature quantization module, is used to combine harmonic distortion rate, load imbalance, residual current waveform complexity index and preset basic threshold to generate an adaptive threshold. The fault diagnosis module, connected to the feature quantization module and the dynamic compensation module, is used to select the fault diagnosis mode based on the residual current waveform complexity index, and to process the residual current signal using the selected mode to output control commands.

Images