Method for testing unbalance current of ac filter capacitor bank

Abstract

The application relates to the technical field of current testing, and particularly discloses a method for testing unbalanced current of an AC filter capacitor bank, which uses a voltage regulator to apply a test voltage smaller than a normal operating voltage to an AC filter capacitor bank to be tested, filters and reconstructs signal components of an unbalanced test current signal between bridge legs of the AC filter capacitor bank based on signal distribution orientation under the test voltage condition, obtains a more accurate unbalanced current measurement value, and converts an unbalanced current protection setting value based on a ratio between the normal operating voltage and the test voltage of the AC filter capacitor bank, and then generates a test result according to a current difference value between the unbalanced current measurement value and the unbalanced current protection setting value. In this way, electromagnetic interference, harmonic noise components and the like in the unbalanced test current signal can be effectively filtered out, the accuracy and reliability of the test result are improved, and the maintenance quality of the AC filter capacitor bank is effectively guaranteed.

Description

Technical Field

[0001] The present application relates to the technical field of current testing, and more specifically, to a method for testing unbalanced current of an AC filter capacitor group. Background Art

[0002] AC filters, as key equipment within converter stations in UHVDC transmission projects, play a crucial role in compensating for the reactive power consumed by converter valves and filtering out DC system harmonics. Capacitors, core components of AC filters, have a direct impact on system stability. However, capacitor banks, exposed to high voltages and strong electromagnetic interference, are prone to abnormal unbalanced currents due to long-term exposure to high voltages and harsh electromagnetic interference. These faults, such as bird damage, bushing leakage, and excessive capacitance, can lead to these abnormalities.

[0003] Currently, unbalance protection is commonly deployed in projects to detect capacitor failures. This includes unbalanced stage I alarm, unbalanced stage II long-delay tripping, and unbalanced stage III short-delay tripping. After replacing or repairing defective or faulty capacitors, the unbalanced current must be less than 20% of the operating protection value. Therefore, arm unbalanced current testing of the corresponding AC filter capacitor bank is necessary to verify the repair quality and provide a basis for equipment commissioning.

[0004] In the early days, converter station maintenance sites lacked dedicated test equipment to accurately measure capacitor bank unbalanced currents, requiring them to build their own circuits for testing. This method required disconnecting multiple equipment lead connection points, resulting in high-altitude work risks and lengthy maintenance. Furthermore, when testing unbalanced current in AC filter capacitor banks, the measured current signal (lub_measured) is often not a pure power-frequency sine wave. It may contain electromagnetic interference from surrounding operating equipment, noise from the test instrument itself, transient interference from switching operations, and the start-up and shutdown of nearby equipment. This results in significant test errors, often requiring multiple tests to reach a more accurate conclusion, and low test efficiency.

[0005] At present, some existing digital filters (such as Butterworth and Chebyshev) or Fourier transform (FFT)-based methods are effective in dealing with signal interference problems, but they have requirements for signal stability and are easily affected by spectrum leakage and fence effects. They are not effective for non-stationary noise (noise characteristics change over time). When the useful signal (fundamental wave) and noise frequencies are close or overlapping, it is difficult to separate them perfectly.

[0006] Therefore, an optimized method for testing unbalanced current of AC filter capacitor banks is expected. Summary of the Invention

[0007] In order to solve the above technical problems, the present application is proposed. An embodiment of the present application provides an unbalanced current test method for an AC filter capacitor group, which uses a voltage regulator to apply a test voltage less than the normal operating voltage to the AC filter capacitor group to be tested. Under the test voltage condition, the unbalanced test current signal between the bridge hips of the AC filter capacitor group is filtered and reconstructed based on the signal distribution guidance to filter out the noise and interference components in the current signal, obtain a more accurate unbalanced current measurement value, and convert the unbalanced current protection constant based on the ratio between the normal operating voltage and the test voltage of the AC filter capacitor group, and then generate a test result based on the current difference between the unbalanced current measurement value and the converted unbalanced current protection constant. In this way, the noise components such as electromagnetic interference and harmonics in the unbalanced test current signal can be effectively filtered out, the accuracy and reliability of the test results can be improved, and a strong guarantee can be provided for the maintenance quality of the AC filter capacitor group.

[0008] According to one aspect of the present application, a method for testing unbalanced current of an AC filter capacitor bank is provided, comprising:

[0009] Use a voltage regulator to apply a test voltage to the AC filter capacitor bank to be tested, where the test voltage is lower than the normal operating voltage;

[0010] Under the condition of applying the test voltage, measuring the unbalanced current between the bridge hips of the AC filter capacitor group;

[0011] generating a test result based on a comparison between the unbalanced current measurement value and the converted unbalanced current protection setting value;

[0012] Among them, under the condition of applying the test voltage, the unbalanced current measurement value between the bridge hips of the AC filter capacitor group is tested, including: filtering and reconstructing the unbalanced test current signal based on the signal distribution orientation to obtain an unbalanced test current enhancement signal; based on the unbalanced test current enhancement signal, the unbalanced current value is obtained.

[0013] Compared with the prior art, the unbalanced current test method of the AC filter capacitor group provided by the present application uses a voltage regulator to apply a test voltage less than the normal operating voltage to the AC filter capacitor group to be tested. Under the test voltage condition, the unbalanced test current signal between the bridge hips of the AC filter capacitor group is filtered and reconstructed based on the signal distribution orientation to filter out the noise and interference components in the current signal, thereby obtaining a more accurate unbalanced current measurement value, and based on the ratio between the normal operating voltage and the test voltage of the AC filter capacitor group, the unbalanced current protection constant is converted, and then the test result is generated based on the current difference between the unbalanced current measurement value and the converted unbalanced current protection constant. In this way, the noise components such as electromagnetic interference and harmonics in the unbalanced test current signal can be effectively filtered out, the accuracy and reliability of the test results can be improved, and a strong guarantee can be provided for the maintenance quality of the AC filter capacitor group. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0015] Figure 1 This is the electrical principle wiring diagram of the AC filter.

[0016] Figure 2 Schematic diagram of the AC filter capacitor bank.

[0017] Figure 3 Schematic diagram of heating defect in capacitor bushing.

[0018] Figure 4 This is a schematic diagram of the waveform recording of the unbalanced protection stage III tripping.

[0019] Figure 5 This is the wiring diagram for the capacitor bank unbalanced current test principle.

[0020] Figure 6 Schematic diagram of disconnection point for unbalanced current measurement.

[0021] Figure 7 Flowchart of a method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application.

[0022] Figure 8 This is the wiring diagram for the unbalanced current test principle.

[0023] Figure 9 This is a schematic diagram of the disconnect point for unbalanced current measurement in the new wiring method.

[0024] Figure 10 For the new wiring method, there is no need to disconnect the marked area and the lead diagram on the opposite side.

[0025] Figure 11 Flowchart of sub-step S2 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application.

[0026] Figure 12 Flowchart of sub-step S21 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application.

[0027] Figure 13 Schematic diagram of data flow in sub-step S21 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application.

[0028] Figure 14 Flowchart of sub-step S212 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application.

[0029] Figure 15 4 is a flowchart of sub-step S213 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application.

[0030] Figure 16 Flowchart of sub-step S2132 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0032] Although the present application makes various references to certain modules in the system according to embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.

[0033] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0034] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0035] It is worth noting that in this application, all actions to obtain data are carried out in compliance with the relevant data protection laws and policies of the country where they are located and with the authorization given by the owner of the corresponding device.

[0036] In the AC filter capacitor bank unbalanced current test:

[0037] 1. AC filter capacitor bank wiring method

[0038] The typical configuration of UHV converter station filters mainly includes SC shunt capacitor banks, HP3 AC filter banks, HP12 / 24 AC filter banks, etc.

[0039] Among them, capacitor (see the principle wiring Figure 1 As one of the important components of AC filters, there are many of them. Taking the receiving-end converter station of a certain UHVDC project as an example, its AC filter field is equipped with 20 groups of filters (including parallel capacitor groups), and a total of 19,272 capacitors are in operation. The AC filter capacitor group generally adopts "H" type wiring, with a total of 4 bridge arms, and each bridge arm contains several capacitors. Current transformers are installed between the bridge arms. These current transformers are different from general AC current transformers and have a smaller ratio, generally 1 / 1 or 1 / 2. They are used to detect unbalanced currents between the bridge arms. The on-site equipment layout is as follows: Figure 2 As shown in the figure, 1-high voltage capacitor bank; 2-current transformer; 3-anchor bolt; 4-grounding flat iron.

[0040] 2. AC filter capacitor bank protection configuration

[0041] The structure of the AC filter capacitor bank is relatively complex and the operating environment is relatively harsh. Common faults include tripping caused by bird damage, oil leakage from the casing, heating of the joints, bulging of the device body, and excessive capacitance. The overall fault types are diverse. Figure 3 shown.

[0042] According to statistics, from 2016 to 2021, the AC filter of domestic converter stations was shut down more than 80 times due to bird damage, capacitor defects and faults. Figure 4 shown.

[0043] At present, unbalance protection is generally configured in projects to detect capacitor faults, including unbalance I stage alarm, unbalance II stage long delay tripping and unbalance III stage short delay tripping.

[0044] The basic principle of unbalanced I and III protection is:

[0045] I ub >I ubqd

[0046]

[0047] Where Iub is the unbalanced current of the capacitor bank, that is, Figure 2 The current flowing through the T1 current transformer; Iubqd is the unbalanced starting setting value, which is fixed at 10mA inside the device; Itro is the through current of the AC filter, that is, Figure 2 The current flowing through the T2 current transformer, Kubzd is the unbalanced proportional coefficient protection setting, the I stage delay is 10s to alarm, and the III stage delay is 20ms to trip.

[0048] The basic principle of unbalanced II stage protection:

[0049]

[0050] Among them, ΔIub is the transient unbalanced current of the AC filter, that is, Figure 2 The current mutation in the T1 current transformer, Itro is the transient through-current of the AC filter, that is, Figure 2 The current flowing through the T2 current transformer, ΔKubzd is the unbalanced proportional coefficient protection setting, and the output is delayed for 120 minutes.

[0051] 3. Principle and method of capacitor bank unbalanced current test

[0052] After replacing and repairing defective or faulty capacitors, the unbalanced current should be less than 20% of the operating protection value. The bridge arm unbalanced current test of the corresponding AC filter capacitor bank is required to verify the quality of the repair work and provide a basis for the equipment to be put into operation. The principle of capacitor bank unbalanced current test is as follows: Figure 5 shown.

[0053] During the test, a test voltage of a certain amplitude is input through the voltage regulator to measure the unbalanced current value between the bridge arms. At the same time, the operating protection setting is converted to the test voltage and compared with the test value and the protection setting to determine the balance of the capacitor bank.

[0054] In the early days, there was no dedicated complete set of test equipment that could accurately measure the unbalanced current of capacitor banks. The converter station maintenance site used self-built circuit testing. Using the above test circuit, it was necessary to disconnect multiple equipment lead connection points, such as Figure 6 As shown in the red box, there are too many disconnection points on the equipment, which makes climbing operations risky and time-consuming.

[0055] Since the stability and anti-interference performance of the self-built circuit are insufficient, and other filters running in the surrounding area generate large electromagnetic interference in the environment, the test error is large during the actual test process, and multiple tests are often required to draw a more accurate conclusion. The data conversion has a certain error probability and the test efficiency is low.

[0056] In response to the above technical problems, the present application proposes a method for testing unbalanced current of an AC filter capacitor group, which uses a voltage regulator to apply a test voltage less than the normal operating voltage to the AC filter capacitor group to be tested. Under the test voltage condition, the unbalanced test current signal between the bridge hips of the AC filter capacitor group is filtered and reconstructed based on the signal distribution orientation to filter out the noise and interference components in the current signal, thereby obtaining a more accurate unbalanced current measurement value. Based on the ratio between the normal operating voltage and the test voltage of the AC filter capacitor group, the unbalanced current protection constant is converted, and then the test result is generated according to the current difference between the unbalanced current measurement value and the converted unbalanced current protection constant. In this way, the noise components such as electromagnetic interference and harmonics in the unbalanced test current signal can be effectively filtered out, the accuracy and reliability of the test results can be improved, and a strong guarantee can be provided for the maintenance quality of the AC filter capacitor group.

[0057] Figure 7 FIG. 1 is a flow chart of a method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 7 As shown, the unbalanced current testing method of the AC filter capacitor group includes the following steps: S1, using a voltage regulator to apply a test voltage to the AC filter capacitor group to be tested, wherein the test voltage is lower than the normal operating voltage; S2, under the condition of applying the test voltage, measuring the unbalanced current between the bridge hips of the AC filter capacitor group; S3, generating a test result based on the comparison between the unbalanced current measurement value and the converted unbalanced current protection set value.

[0058] In the above-mentioned AC filter capacitor group unbalanced current test method, in step S1, a test voltage is applied to the AC filter capacitor group to be tested using a voltage regulator, and the test voltage is lower than the normal operating voltage. It should be understood that the balance of the capacitor group needs to be verified after replacement and maintenance, but there are safety risks in directly applying the normal operating voltage, and the on-site conditions are difficult to meet the high-voltage testing requirements. According to the "Power Transmission and Transformation Equipment Status Maintenance Test Procedure", it is allowed to verify the matching of protection settings through an equivalent test with reduced voltage. Therefore, the present application applies a test voltage lower than the normal operating voltage to the AC filter capacitor group to be tested, simulates the operating conditions in a safe and controllable low-voltage environment, and obtains a current signal reflecting the true unbalanced state of the capacitor group, which helps to avoid potential hazards to equipment and personnel caused by high-voltage operation.

[0059] In particular, without departing from the basic principles of the test, this application optimizes the wiring method. The new wiring method only requires the removal of the leads on the low-voltage side of the capacitor bank. There is no need to remove the leads on the high-voltage side of the capacitor bank, nor is there any need to open the high-voltage side grounding switch. The grounding switch at the high-voltage end of the AC filter is used to connect part of the earth operation test wiring loop. The test voltage application method is similar to the reverse connection method for measuring the dielectric loss and capacitance of the equipment. During the test, the test voltage is applied to the low-voltage side of the capacitor bank, and a complete circuit is formed through the capacitor tower, the high-voltage side grounding switch, and the instrument grounding. The test wiring is as follows Figure 8 shown.

[0060] The new test wiring method does not require the removal of the high-voltage side leads of the capacitor, nor does it require the connection of high-voltage side test lines during the test, significantly reducing the need for special vehicles and the intensity and danger of high-altitude operations. Figure 9 、 Figure 10 shown.

[0061] Furthermore, since the high-voltage side of the capacitor bank remains connected, the capacitance of the high-voltage bridge arm can be measured using the ground loop. For example, to measure the capacitance of bridge arm 2, one end of the capacitance meter can be connected to ground and the other end can be connected between bridge arms 1 and 2. Compared to the previous improvement, this eliminates the need for dedicated test leads on the high-voltage side.

[0062] In the above-mentioned AC filter capacitor group unbalanced current test method, the step S2 tests the unbalanced current measurement value between the bridge hips of the AC filter capacitor group under the condition of applying the test voltage. It should be understood that due to the presence of interference such as broadband harmonics and transient pulses generated by the operation of adjacent filters in the converter station environment, the collected unbalanced test current signal contains a large amount of non-power frequency noise, and the signal-to-noise ratio is low, which cannot accurately reflect the true unbalanced state of the capacitor group. Traditional filtering methods have significant limitations in processing such non-stationary and nonlinear noise signals, and the filtering effect is poor. In this regard, the present application introduces a signal processing algorithm based on deep learning technology, which automatically identifies and extracts the effective components in the signal by learning the characteristic distribution of the unbalanced test current signal, suppresses high-frequency harmonics and transient interference, and improves the signal-to-noise ratio and purity of the signal. Among them, Figure 11 FIG. 1 is a flow chart of sub-step S2 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 11 As shown, the step S2 includes the steps of: S21, filtering and reconstructing the signal components of the unbalanced test current signal based on signal distribution guidance to obtain an unbalanced test current enhanced signal; S22, obtaining the unbalanced current value based on the unbalanced test current enhanced signal.

[0063] Figure 12 Flowchart of sub-step S21 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 13 FIG. 1 is a data flow diagram of sub-step S21 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 12 and Figure 13 As shown, the step S21 includes the following steps: S211, capturing the unbalanced test current signal; S212, performing current signal feature extraction based on modal decomposition on the unbalanced test current signal to obtain a set of current signal component feature coding vectors; S213, performing feature filtering based on cluster distribution learning on the set of current signal component feature coding vectors to obtain a filtered set of current signal component feature coding vectors; S214, inputting the filtered set of current signal component feature coding vectors into a current signal reconstruction model to obtain an unbalanced test current enhanced signal.

[0064] More specifically, step S211 captures the unbalanced test current signal. Specifically, the unbalanced test current signal generated by the AC filter capacitor bank under the condition of applying the test voltage is collected in real time by the bridge arm current transformer, and the analog signal is converted into a digital signal using a high-precision ADC module to ensure the integrity and timeliness of the signal acquisition.

[0065] During implementation, the first consideration is how to efficiently and accurately acquire the original signal. By using a voltage regulator to apply a test voltage lower than the normal operating voltage to the AC filter capacitor bank under test, the impact on the equipment itself can be effectively reduced while ensuring that subsequent tests can be conducted within a safe range. Furthermore, high-precision bridge arm current transformers can be used to acquire the unbalanced test current signal generated by the AC filter capacitor bank in real time under the applied test voltage. These transformers typically possess high sensitivity and accuracy, enabling them to accurately detect even subtle current signal changes in complex electromagnetic environments. To further enhance signal acquisition quality, the placement of the transformers is often optimized based on actual site conditions to ensure comprehensive coverage of all areas where unbalanced current may occur. Furthermore, given the complex internal structure of the capacitor bank, electrical characteristics may vary between different parts. Therefore, when deploying multiple transformers, these variations must be fully accounted for, and targeted measures must be taken to achieve optimal data acquisition.

[0066] After the bridge arm current transformer successfully captures the unbalanced test current signal, the next step is to convert this analog signal into a digital form. This step primarily relies on a high-precision ADC module. The ADC module's function is to convert the analog signal received from the transformer into a digital form that can be processed by a computer, laying the foundation for further data analysis and processing. During this conversion process, to ensure the integrity and timeliness of signal conversion, the ADC module's operating parameters, including key indicators such as sampling rate and resolution, must be strictly controlled. The sampling rate determines the number of samples collected per unit time, which directly affects whether the resulting digital signal can accurately reflect the changing trends of the original analog signal. The resolution, on the other hand, is related to the signal quantization accuracy, that is, the ability to distinguish between two adjacent values, which is crucial for improving signal conversion quality.

[0067] It's important to note that while converting analog signals to digital, attention must also be paid to signal integrity during transmission. Because capacitor banks operate in complex and variable environments, they may be subject to strong electromagnetic interference sources, which can impact the signal at the receiving end to varying degrees. Therefore, when designing the signal transmission path, cables with excellent shielding should be used whenever possible, and the wiring scheme should be carefully planned to avoid parallel installation with other strong interference sources, thereby minimizing external noise interference with the received signal. Furthermore, signal attenuation, which can occur over long transmission distances, can be addressed by adding signal amplifiers, ensuring good signal quality and stability throughout the entire transmission link.

[0068] To gain a deeper understanding of the characteristics of unbalanced test current signals, comprehensive analysis sometimes requires the use of other complementary methods. For example, sensors specifically designed to monitor environmental conditions (such as temperature and humidity) can be installed near the data collection point. Simultaneously recording these parameters can help identify specific patterns or anomalies in the signal under certain conditions. This approach not only helps technicians better interpret the collected data but also provides valuable insights for subsequent optimization of test methods.

[0069] Figure 14 FIG. 1 is a flow chart of sub-step S212 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 14 As shown, the step S212 includes the steps of: S2121, performing CEEMDAN signal decomposition on the unbalanced test current signal to obtain a set of current signal components; S2122, extracting current signal features of each current signal component in the set of current signal components to obtain a set of current signal component feature coding vectors.

[0070] In a specific example of the present application, the step S2121 performs CEEMDAN signal decomposition on the unbalanced test current signal to obtain a set of current signal components. It should be understood that since the unbalanced test current signal is a non-stationary mixed signal, and the traditional Fourier transform is not effective in processing time-varying noise (such as harmonic components that vary with time), the present application uses CEEMDAN (Complete Ensemble Empirical Mode Decomposition) technology to decompose the unbalanced test current signal to form a set of current signal components. Specifically, CEEMDAN signal decomposition is an adaptive signal time-frequency analysis method. By adding adaptive white noise to the signal and iterating multiple times, it overcomes the modal aliasing problem of traditional EMD and can achieve high-precision decomposition of the signal. In the present application, the CEEMDAN signal decomposition technology is used to decompose the unbalanced test current signal from a complex non-stationary signal into a series of intrinsic mode components (IMFs). Each IMF represents an inherent oscillation mode of the signal with different frequencies and time scales. In this way, the power frequency fundamental component, harmonic component and noise component in the unbalanced test current signal can be effectively separated, providing a basis for subsequent signal reconstruction and filtering.

[0071] In a specific example of the present application, the step S2122 extracts the current signal features of each current signal component in the set of current signal components to obtain a set of current signal component feature encoding vectors. It should be understood that each IMF component (i.e., the current signal component) contains different physical meanings (such as noise, fundamental wave, harmonics, etc.). In order to further quantitatively describe its characteristics through feature extraction so as to accurately identify effective components in the subsequent feature selection process, the present application adopts a multi-dimensional feature combination to extract the current signal features of each current signal component (such as mean, variance, peak factor, main frequency, frequency band energy ratio, sample entropy, permutation entropy, etc.) from three dimensions: time domain, frequency domain, and entropy value. A normalization algorithm (such as Z-Score) is used to eliminate dimensional differences to construct a vectorized feature representation of each IMF component, obtain a set of current signal component feature encoding vectors, and provide a quantitative basis for subsequent feature filtering operations.

[0072] More specifically, in step S213, the set of current signal component feature coding vectors is subjected to feature filtering based on cluster distribution learning to obtain a filtered set of current signal component feature coding vectors. Specifically, since most traditional feature selection algorithms analyze each component feature independently, while ignoring the correlation between features and the group distribution characteristics of signal components, it is easy to mistakenly delete useful features or retain redundant noise features in a complex noise environment. To address this problem, the present application proposes a feature filtering method based on cluster distribution learning, which constructs a semantic association between individual current signal component features and signal component group distribution characteristics by learning the distribution characteristics of each current signal component feature coding vector and the entire component feature set, and then screens out the feature subset that contributes most to the current unbalanced current test task, that is, the filtered set of current signal component feature coding vectors, providing a more accurate information basis for subsequent signal reconstruction. Among them, Figure 15 FIG. 1 is a flow chart of sub-step S213 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 15 As shown, the step S213 includes the steps of: S2131, extracting the i-th current signal component feature coding vector from the set of current signal component feature coding vectors as the current signal component individual feature coding vector; S2132, based on the semantic association distribution pattern of the set of current signal component feature coding vectors, performing feature modulation on the current signal component individual feature coding vector to obtain a current signal component individual feature modulation coding vector; S2133, determining whether to delete the i-th current signal component feature coding vector based on the information gain of the current signal component individual feature modulation coding vector relative to the set of current signal component feature coding vectors.

[0073] In a specific example of the present application, step S2131 is expressed as follows:

[0074] X={x1,x2,...,x i ,...,x n}

[0075] Among them, X represents the set of characteristic coding vectors of current signal components, x1, x2, x i and x n They represent the first, second, i-th and n-th current signal component feature coding vectors in the set of current signal component feature coding vectors, n is the number of current signal component feature coding vectors, and x i As the individual feature encoding vector of the current signal component.

[0076] That is, by focusing on the unique properties of a single signal component, such as the amplitude fluctuation of a specific frequency component and the time-varying characteristics of noise interference, a more refined analysis basis is provided for subsequent signal component filtering.

[0077] Figure 16 FIG. 1 is a flow chart of sub-step S2132 of the method for testing unbalanced current of an AC filter capacitor bank according to an embodiment of the present application. Figure 16 As shown, the step S2132 includes the steps of: S21321, calculating the group distribution semantic map of the set of current signal component feature coding vectors to obtain the current signal component feature group distribution semantic map; S21322, mapping the current signal component individual feature coding vector to the feature space of the current signal component feature group distribution semantic map to obtain the current signal component individual feature modulation coding vector.

[0078] In a specific example of the present application, step S21321 includes: calculating the semantic association between any two current signal component feature coding vectors in the set of current signal component feature coding vectors to obtain the current signal component feature group distribution semantic map composed of multiple semantic associations, which is expressed as follows:

[0079]

[0080] Where R(·,·) represents the semantic correlation between the feature encoding vectors of the corresponding two current signal components, [·;·] represents feature cascade, and x j represents the jth current signal component feature coding vector in the set of current signal component feature coding vectors, W r and b represent the weight parameter matrix and bias term respectively, sigmoid is the activation function, r i,j Represents x i and x jThe semantic association interaction encoding vector of the current signal components between them, T is r i,j The characteristic scale value of h is the vector r i,j The position index of the eigenvalue in , M represents the semantic map of the current signal component feature group distribution.

[0081] That is, by constructing a graph-structured representation, the feature encoding vectors of each current signal component are placed in the overall context environment for examination, and the semantic dependencies and functional complementary relationships beyond simple linear correlations are explored, providing group context information containing prior knowledge such as feature cluster structure and component status differences for subsequent signal processing, and obtaining a semantic graph of the current signal component feature group distribution. This provides a global perspective guidance for the precise design of signal filtering and reconstruction strategies, avoiding the problem of incomplete signal filtering caused by ignoring nonlinear correlations between components.

[0082] In a specific example of the present application, step S21322 is expressed as follows:

[0083]

[0084] Among them, v i Represents x i The corresponding individual characteristic modulation coding vector of the current signal component.

[0085] That is, by embedding the individual feature coding vector of the current signal component into the feature space of the semantic graph of the feature group distribution of the current signal component, and utilizing the global correlation information implicit in the graph structure to contextually enhance the individual features, the modulated individual feature modulation coding vector of the current signal component not only retains the independent attributes of the individual feature coding vector of a single current signal component, but also incorporates its structural position and functional association in the overall signal, providing a more comprehensive feature expression for subsequent signal filtering and reconstruction, thereby improving the ability to distinguish useful signals and noise components in subsequent signal processing.

[0086] In a preferred example of the present application, step S2133 includes: first, performing global semantic conjugate distribution universal optimization on the individual feature modulation coding vector of the current signal component to obtain an optimized individual feature modulation coding vector of the current signal component, which is expressed as:

[0087]

[0088] x′ i =ln(v i )⊙[ln(x i ) ⊙-1 ]

[0089]

[0090] in,(·) T represents the transpose of a vector, represents vector multiplication, (·) ⊙-1 represents the inverse of each eigenvalue in the calculation vector, M′ represents the semantic map of the optimized current signal component feature population distribution, ln(·) represents the logarithmic function with the natural constant e as the base, ⊙ represents the dot product, and x′ i Represents x i The corresponding optimized current signal component individual feature encoding vector, v′ i Indicates v i The corresponding optimized current signal component individual characteristic modulation coding vector.

[0091] Here, the current signal component feature group distribution semantic map is obtained by (x i ,x j ) is calculated pair by pair to capture the feature space topology and other information of the local correlation of the feature coding vectors of each current signal component, there is inevitably a problem of global universal unevenness of local correlation defects, which affects the accuracy of the overall-local difference significance measurement of the individual feature modulation coding vectors of the current signal components. In this regard, the present application introduces the universal optimization of the global semantic conjugate distribution to correct the energy distribution deviation of the local correlation defects in the global semantic background, so as to enhance the unified characterization capability of the individual feature modulation coding vectors of the current signal components for the signal components in a complex electromagnetic environment. The generated optimized individual feature modulation coding vectors of the current signal components not only retain the local unique properties of the signal components, but also accurately map their energy scaling relationship in the global signal space, thereby solving the problem of feature expression deviation caused by the limitations of local correlation calculations.

[0092] Then, the current signal component individual feature semantic increment operator of the optimized current signal component individual feature modulation coding vector relative to the set of current signal component feature coding vectors is calculated, which is expressed as follows:

[0093]

[0094] Where L represents v i The vector scale, v i,k Indicates v i The eigenvalue of the kth position in the equation, λ represents v i The corresponding semantic incremental intermediate parameter, arctan(·) represents the inverse tangent function, Indicates v i The corresponding semantic increment operator of individual features of current signal components.

[0095] Specifically, counterfactual reasoning is used to quantify the causal effect of each optimized current signal component's individual feature modulation coding vector, assessing its impact on the signal's overall semantic expressiveness, thereby providing a quantifiable priority basis for the subsequent selective filtering and reconstruction of signal components. This generates a quantitative indicator reflecting the importance of each signal component, namely the current signal component individual feature semantic increment operator. A high value indicates that the optimized current signal component's individual feature modulation coding vector is a core element in maintaining the signal's effective characteristics, and its removal would significantly damage the signal's semantics. A low value corresponds to a secondary or interfering component, whose removal would have a minimal impact on the signal's overall characteristics, thereby achieving the key retention of useful signals and the precise removal of noise components.

[0096] Finally, based on the comparison between the semantic increment operator of the individual feature of the current signal component and the preset threshold, it is determined whether to delete the feature coding vector of the i-th current signal component, which is expressed as follows:

[0097]

[0098] Where θ represents the preset mask threshold and mask(·) is the mask function.

[0099] That is, by constructing an objective decision boundary through preset thresholds, noise or secondary components that have a weak impact on the overall semantic expression ability of the signal are eliminated, while retaining the core useful components, achieving denoising and retaining the truth in the feature dimension, and retaining high-value components that carry key information such as fundamental wave amplitude and phase, thereby significantly improving the purity of the signal components, reducing the interference of non-stationary noise on the measurement results, and laying the foundation for the accurate calculation of the unbalanced current value.

[0100] More specifically, in step S214, the filtered set of the current signal component feature encoding vectors is input into the current signal reconstruction model to obtain the unbalanced test current enhancement signal. Specifically, in the current signal reconstruction stage, the present application adopts an autoencoder model under a deep learning architecture as the current signal reconstruction model. The autoencoder model can learn the low-dimensional representation of the data from the input high-dimensional feature space through unsupervised learning, and reconstruct the high-dimensional features of the original data based on this low-dimensional representation, thereby achieving effective dimensionality reduction of features and information retention. In the present application, the autoencoder model receives the filtered set of the feature encoding vectors of the current signal component after feature filtering as input, maps the high-dimensional features to the low-dimensional latent space through its encoding layer, and then reconstructs the high-dimensional feature representation of the current signal from the low-dimensional latent space through the decoding layer, i.e., the unbalanced test current enhancement signal. This process can not only effectively remove the noise component, but also retain the key information reflecting the unbalanced state of the capacitor bank, thereby greatly improving the signal-to-noise ratio and interpretability of the signal.

[0101] Specifically, step S22 involves obtaining the unbalanced current value based on the unbalanced test current enhancement signal. It should be understood that the unbalanced test current enhancement signal has been filtered out of non-stationary noise and interference components, and its amplitude and phase can truly reflect the degree of imbalance between the capacitor bank bridge arms. Therefore, by performing an effective value calculation on the unbalanced test current enhancement signal, the unbalanced current value between the capacitor bank bridge arms of the AC filter can be obtained.

[0102] In the above-mentioned AC filter capacitor bank unbalanced current testing method, step S3 generates a test result based on a comparison between the unbalanced current measurement value and the converted unbalanced current protection setting value. More specifically, step S3 includes: obtaining the unbalanced current protection setting value of the AC filter capacitor bank under normal operating voltage; converting the unbalanced current protection setting value based on the ratio between the normal operating voltage and the test voltage to obtain the converted unbalanced current protection setting value; calculating the current difference between the unbalanced current measurement value and the converted unbalanced current protection setting value; and generating the test result based on a comparison between the current difference and a preset threshold.

[0103] It should be understood that based on the principle of electromagnetic induction, the voltage regulator generates a low-voltage current proportional to the operating voltage between the capacitor bank bridge arms by changing the input voltage amplitude. Since the capacitor's capacitive reactance is inversely proportional to the voltage frequency, under power frequency conditions, the low-voltage test current and the high-voltage operating current are linearly proportional. Based on this, the unbalanced current protection setting can be equivalently converted using the voltage ratio. The conversion formula is as follows:

[0104]

[0105] Among them, I2 represents the unbalanced I-segment protection current value under the test voltage, that is, the converted unbalanced current protection constant value, U1 represents the test applied voltage value, I1 represents the unbalanced I-segment protection current constant value, U n Indicates the rated line voltage value of the filter.

[0106] The calculated unbalanced current value is then compared with the converted unbalanced current protection setting. By quantifying the deviation between the current difference and the preset threshold, the capacitor bank balance state is assessed, providing a reliable basis for commissioning decisions. For example, according to standard requirements, the unbalanced current measurement value must be less than 20% of the protection setting value. Based on this, a threshold can be dynamically set to determine whether the capacitor bank is in a normal state or requires maintenance. This allows for real-time monitoring and early warning of the capacitor bank's unbalanced state, improving the maintenance efficiency and operational reliability of the power grid.

[0107] In summary, the unbalanced current test method of the AC filter capacitor group based on the embodiment of the present application is explained, which uses a voltage regulator to apply a test voltage less than the normal operating voltage to the AC filter capacitor group to be tested. Under the test voltage condition, the unbalanced test current signal between the bridge hips of the AC filter capacitor group is filtered and reconstructed based on the signal distribution guidance to filter out the noise and interference components in the current signal, obtain a more accurate unbalanced current measurement value, and based on the ratio between the normal operating voltage and the test voltage of the AC filter capacitor group, convert the unbalanced current protection constant value, and then generate the test result based on the current difference between the unbalanced current measurement value and the converted unbalanced current protection constant value. In this way, the noise components such as electromagnetic interference and harmonics in the unbalanced test current signal can be effectively filtered out, the accuracy and reliability of the test results can be improved, and a strong guarantee can be provided for the maintenance quality of the AC filter capacitor group.

[0108] The basic principles of the present invention have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, and effects mentioned in the present invention are merely illustrative and non-limiting, and should not be construed as necessarily possessed by each embodiment of the present invention. Furthermore, the specific details of the above embodiments are provided for illustrative purposes and to facilitate understanding, and are not intended to be limiting. These details do not necessarily limit the present invention to being implemented using these specific details.

[0109] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0110] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be encompassed therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0111] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0112] Finally, it should be noted that the above description has been provided for purposes of illustration and description. Furthermore, the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to be limiting. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art will appreciate that the technical solutions of the present invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for testing unbalanced current of an AC filter capacitor bank, characterized in that: include: Use a voltage regulator to apply a test voltage to the AC filter capacitor bank to be tested, where the test voltage is lower than the normal operating voltage; Under the condition of applying the test voltage, measuring the unbalanced current between the bridge hips of the AC filter capacitor group; generating a test result based on a comparison between the unbalanced current measurement value and the converted unbalanced current protection setting value; Wherein, under the condition of applying the test voltage, testing the unbalanced current measurement value between the bridge hips of the AC filter capacitor group includes: filtering and reconstructing the unbalanced test current signal based on signal distribution guidance to obtain an unbalanced test current enhanced signal; and obtaining the unbalanced current value based on the unbalanced test current enhanced signal; The unbalanced test current signal is filtered and reconstructed based on signal distribution guidance to obtain an unbalanced test current enhanced signal, including: Capture unbalanced test current signal; Performing current signal feature extraction based on modal decomposition on the unbalanced test current signal to obtain a set of current signal component feature coding vectors; Performing feature filtering based on cluster distribution learning on the set of current signal component feature coding vectors to obtain a filtered set of current signal component feature coding vectors; Inputting the filtered set of the current signal component feature coding vectors into a current signal reconstruction model to obtain the unbalanced test current enhancement signal; Performing feature filtering based on cluster distribution learning on the set of current signal component feature coding vectors to obtain a filtered set of current signal component feature coding vectors, including: Extracting the i-th current signal component feature coding vector from the set of current signal component feature coding vectors as the current signal component individual feature coding vector; Based on the semantic association distribution pattern of the set of current signal component feature coding vectors, feature modulating the individual feature coding vectors of the current signal components to obtain individual feature modulation coding vectors of the current signal components; Based on the information gain of the individual characteristic modulation coding vector of the current signal component relative to the set of the current signal component characteristic coding vectors, it is determined whether to delete the i-th current signal component characteristic coding vector.

2. The method for testing unbalanced current of an AC filter capacitor bank according to claim 1, characterized in that: Based on the comparison between the unbalanced current measurement value and the converted unbalanced current protection setting value, a test result is generated, including: Obtaining an unbalanced current protection setting value of the AC filter capacitor bank under normal operating voltage; Based on the ratio between the normal operating voltage and the test voltage, the unbalanced current protection constant is converted to obtain the converted unbalanced current protection constant.

3. The method for testing unbalanced current of an AC filter capacitor bank according to claim 2, characterized in that: Generating a test result based on a comparison between the unbalanced current measurement value and the converted unbalanced current protection setting value further includes: Calculating a current difference between the unbalanced current measurement value and the converted unbalanced current protection set value; The test result is generated based on a comparison between the current difference and a preset threshold.

4. The method for testing unbalanced current of an AC filter capacitor bank according to claim 1, characterized in that: Performing current signal feature extraction based on modal decomposition on the unbalanced test current signal to obtain a set of current signal component feature coding vectors, including: performing CEEMDAN signal decomposition on the unbalanced test current signal to obtain a set of current signal components; The current signal feature of each current signal component in the set of current signal components is extracted to obtain a set of current signal component feature coding vectors.

5. The method for testing unbalanced current of an AC filter capacitor bank according to claim 4, characterized in that: Based on the semantic association distribution pattern of the set of current signal component feature coding vectors, feature modulating the individual feature coding vectors of the current signal components to obtain the individual feature modulation coding vectors of the current signal components, including: Calculating a group distribution semantic map of the set of current signal component feature coding vectors to obtain a current signal component feature group distribution semantic map; The individual feature coding vector of the current signal component is mapped to the feature space of the current signal component feature group distribution semantic map to obtain the individual feature modulation coding vector of the current signal component.

6. The method for testing unbalanced current of an AC filter capacitor bank according to claim 5, characterized in that: Calculating a group distribution semantic map of the set of current signal component feature coding vectors to obtain a current signal component feature group distribution semantic map includes: The semantic association between any two current signal component feature coding vectors in the set of the current signal component feature coding vectors is calculated to obtain the current signal component feature group distribution semantic graph composed of multiple semantic associations.

7. The method for testing unbalanced current of an AC filter capacitor bank according to claim 6, characterized in that: Determining whether to delete the i-th current signal component feature coding vector based on the information gain of the individual feature modulation coding vector of the current signal component relative to the set of current signal component feature coding vectors includes: Performing global semantic conjugate distribution universal optimization on the individual feature modulation coding vector of the current signal component to obtain an optimized individual feature modulation coding vector of the current signal component; Calculating a current signal component individual feature semantic increment operator of the optimized current signal component individual feature modulation coding vector relative to the set of current signal component feature coding vectors; Based on the comparison between the semantic increment operator of the individual feature of the current signal component and a preset threshold, it is determined whether to delete the i-th current signal component feature coding vector.

Images