Circuit fault feature active enhancement detection method for distribution box

Abstract

The invention discloses a line fault characteristic active enhancement detection method for a distribution box, and relates to the technical field of power system detection, and the method comprises the following steps: carrying out the real-time collection of an electrical signal in the distribution box, carrying out the processing of the signal through an active enhancement strategy, highlighting the fault characteristic, and obtaining the data of the enhanced electrical signal. Through real-time signal acquisition, enhancement, feature extraction, frequency overlap analysis, machine learning evaluation and digital signal processing (DSP) frequency shift operation, it is ensured that an enhanced signal frequency spectrum is not overlapped with a system working frequency, and fault signal distortion or misjudgment caused by frequency spectrum interference is avoided. A machine learning model and a DSP algorithm are combined, signal processing is dynamically adjusted, the frequency overlapping problem is automatically identified and corrected, the performance and stability of a fault detection system are improved, an intelligent and accurate fault detection solution is provided, and the safety and reliability of a power system are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of power system detection, and in particular to a method for actively enhancing line fault feature detection for a distribution box. Background Art

[0002] "Active enhanced detection of line fault features in distribution boxes" refers to the process of actively extracting and enhancing the features of the collected electrical signals (such as current, voltage, temperature rise, harmonics, etc.) using certain technical means during fault detection of the internal lines of the distribution box, thereby amplifying or highlighting the key characteristic signals related to the fault and improving the sensitivity and accuracy of fault detection. Unlike traditional passive detection methods, active enhancement not only includes the suppression of noise interference, but may also combine signal processing algorithms, feature transformation, pattern recognition or artificial intelligence models to enhance potential fault features in the original data, so that even minor, early or hidden fault features can be effectively identified, thereby improving the response capability and diagnostic efficiency to abnormal conditions. This method is particularly suitable for multi-type fault detection scenarios under complex load environments in low-voltage distribution systems.

[0003] The existing technology has the following deficiencies: In the process of fault detection in the distribution box, active enhancement strategies (such as nonlinear transformation or signal superposition) are usually used to process the collected electrical signals in order to highlight and amplify the features related to the fault. However, these enhancement strategies may, in some cases, resonate with the spectrum or interfere with the natural operating frequency of the distribution system (such as the 50Hz grid frequency). In particular, processing methods such as nonlinear transformation and signal superposition may change the frequency components of the signal, causing the enhanced signal spectrum to overlap with the operating frequency band of the power system. At this time, the normal operating frequency components of the system may interfere or resonate with the enhanced signal frequency, resulting in distortion of the signal characteristics or deformation of the morphology.

[0004] This spectrum resonance or mutual interference will not only affect the accuracy of signal characteristics, but also make it difficult to distinguish fault signals from normal working signals, resulting in the fault detection system being unable to effectively identify potential faults, or misjudging normal signals as fault signals, affecting the stability and safety of the system.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a method for actively enhancing the detection of line fault features for distribution boxes, which ensures that the enhanced signal spectrum does not overlap with the system operating frequency through real-time signal acquisition, enhancement, feature extraction, frequency overlap analysis, machine learning evaluation and digital signal processing (DSP) frequency shift operations, avoiding distortion or misjudgment of fault signals caused by spectrum interference. Combined with machine learning models and DSP algorithms, signal processing is dynamically adjusted, frequency overlap problems are automatically identified and corrected, the performance and stability of the fault detection system are improved, intelligent and accurate fault detection solutions are provided, and the safety and reliability of the power system are enhanced to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a method for actively enhancing line fault feature detection for a distribution box, comprising the following steps: Collect electrical signals in the distribution box in real time, process the signals through active enhancement strategies, highlight fault characteristics, and obtain enhanced electrical signal data; The enhanced electrical signal data acquired in real time is stored in a data set, and key features reflecting the overlap between the signal frequency and the normal working frequency of the power distribution system are extracted from the data set through feature engineering technology. The extracted key features are comprehensively analyzed and the degree of frequency overlap is quantified. The key features after comprehensive analysis are input into the pre-trained machine learning model for further evaluation to determine whether the enhanced signal overlaps with the normal working frequency; When it is identified that the frequency of the enhanced signal overlaps with the operating frequency of the distribution system, a digital signal processing algorithm is used to introduce a frequency shift operation in the enhanced signal to shift the frequency bandwidth of the signal to a frequency band outside the operating frequency of the distribution system to avoid frequency overlap; subsequently, the frequency-shifted signal is further analyzed to confirm that its frequency no longer overlaps with the operating frequency of the distribution system, and on this basis, the fault signal is identified.

[0008] Preferably, the electrical signals in the distribution box are collected in real time, and the signals are processed by an active enhancement strategy to highlight the fault characteristics. The specific steps are as follows: Collect electrical signals from electrical equipment in the distribution box in real time to provide raw data for subsequent processing; Apply active enhancement strategies to acquired electrical signals to highlight potential fault characteristics; After applying the enhancement strategy, enhanced electrical signal data is obtained to provide optimized signal input for subsequent analysis and processing.

[0009] Preferably, key features reflecting the overlap of signal frequency and normal operating frequency of the power distribution system are extracted from the data set through feature engineering technology, and the extracted features include the amplitude and phase changes of the signal and the power distribution of the enhanced signal in the spectrum. The amplitude and phase changes of the signal and the power distribution of the enhanced signal in the spectrum are comprehensively analyzed under the detection window to generate amplitude and phase error reference values ​​and power spectrum density fluctuation reference values, respectively, and the degree of frequency overlap is quantified by the amplitude and phase error reference values ​​and the power spectrum density fluctuation reference values.

[0010] Preferably, the specific steps of comprehensively analyzing the amplitude and phase changes of the signal under the detection window to generate the amplitude and phase error reference value are as follows: The amplitude change of the signal is quantified, the amplitude error between the enhanced signal and the ideal signal is analyzed, the amplitude distortion factor is defined, and the amplitude difference of the signal in the frequency space is measured to quantify the amplitude change degree. The calculation expression of the amplitude distortion factor is as follows: , where is the amplitude distortion factor, The enhanced signal is The frequency at the frequency point Amplitude value, The ideal signal is The frequency at the frequency point Amplitude value, is the number of frequencies, is a very small positive number used to avoid division by zero. is an exponential parameter used to perform nonlinear amplification of the amplitude difference; Analyze the phase error of the signal, quantify the phase change of the signal, and determine whether the phase of the signal has changed by calculating the phase difference between the enhanced signal and the ideal signal. Define the phase distortion factor. The calculation expression of the phase distortion factor is as follows: , where is the phase distortion factor, is the adjustment parameter, The enhanced signal is The frequency at the frequency point The phase angle at The ideal signal is The frequency at the frequency point The phase angle at is the phase index parameter, is the exponential change of the phase error; Finally, the amplitude distortion factor and phase distortion factor Combined, the amplitude and phase error reference value is obtained to quantify the amplitude and phase changes of the enhanced signal. The calculation expression is as follows: , where is the amplitude phase error reference value.

[0011] Preferably, the specific steps of comprehensively analyzing the power distribution of the enhanced signal in the spectrum under the detection window to generate a power spectrum density fluctuation reference value are as follows: In the process of signal spectrum analysis, the power spectrum density of the enhanced electrical signal is first calculated to understand the energy distribution of the signal in different frequency ranges and calculate the power of each frequency band. The calculation expression is as follows: , where It is The power spectral density of the frequency band, It is The center frequency of the frequency band, is the frequency bandwidth, The signals are at different frequencies The energy distribution on After obtaining the power spectrum density of each frequency band, the fluctuation degree of power in the spectrum is further analyzed to quantify the change of signal frequency. The calculation expression of the reference value of power spectrum density fluctuation is as follows: , where is the power spectrum density fluctuation reference value, is the power spectral density of the previous frequency band, is the weight factor, indicating The contribution of each frequency band to power fluctuation, is the total number of frequency bands.

[0012] Preferably, the amplitude phase error reference value and power spectrum density fluctuation reference value that have undergone comprehensive analysis are input into a pre-trained machine learning model for further evaluation, and a frequency overlap risk coefficient is generated by the machine learning model. The frequency overlap risk coefficient is used to determine whether the enhanced signal overlaps with the normal working frequency.

[0013] Preferably, the frequency overlap risk coefficient generated when the pre-trained machine learning model is used to determine whether the enhanced signal overlaps with the normal working frequency is compared with a preset frequency overlap risk coefficient reference threshold to determine whether the enhanced signal overlaps with the normal working frequency. The judgment logic is as follows: If the frequency overlap risk coefficient is greater than the pre-set frequency overlap risk coefficient reference threshold, it is judged that the enhanced signal frequency overlaps with the normal operating frequency of the distribution system; if the frequency overlap risk coefficient is less than or equal to the pre-set frequency overlap risk coefficient reference threshold, it is judged that the enhanced signal frequency does not overlap with the normal operating frequency of the distribution system.

[0014] Preferably, when it is identified that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a digital signal processing algorithm is used to introduce a frequency shift operation in the enhanced signal to shift the frequency bandwidth of the signal to a frequency band outside the operating frequency of the power distribution system; then, the specific steps of further analyzing the frequency-shifted signal are as follows: After identifying the frequency overlap, the digital signal processing algorithm is used to frequency shift the enhanced signal, shifting its frequency components from the operating frequency bandwidth of the power distribution system to the external frequency band to avoid frequency overlap. The calculation expression is as follows: When it is determined that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a frequency shift operation is introduced through digital signal processing technology to shift the frequency bandwidth of the signal out of the operating frequency band of the power distribution system. The frequency shift calculation expression is as follows: , where is the frequency domain representation of the frequency-shifted signal, is the frequency domain representation of the original enhanced signal, that is, the enhanced signal spectrum, is the complex exponential function, is an imaginary unit, is the frequency shift, is the natural base; After the frequency shift operation is completed, the frequency-shifted signal needs to be further analyzed to ensure that the frequency of the signal has been completely decoupled and no longer overlaps with the operating frequency of the power distribution system. The calculation expression is as follows: , where is the frequency overlap risk coefficient recalculated after frequency shift, It is a frequency overlap risk coefficient calculation model, which is used to analyze the frequency characteristics of the signal after frequency shift. is the feature extraction function; like ,in, is the reference threshold of the frequency overlap risk coefficient.

[0015] In the above technical solution, the technical effects and advantages provided by the present invention are: The present invention ensures that the enhanced signal spectrum does not overlap with the operating frequency of the system through a series of steps such as real-time signal acquisition and enhancement, feature extraction and frequency overlap analysis, machine learning model evaluation, and frequency shift operations in digital signal processing (DSP). This process avoids the problem of fault signal distortion or misjudgment caused by spectrum interference, allowing fault signals to be clearly and accurately identified. In addition, through the combination of machine learning models and DSP algorithms, the signal processing process can be dynamically adjusted, and the impact of frequency overlap can be automatically identified and corrected, thereby improving the overall performance and stability of the fault detection system. Ultimately, this solution provides a more intelligent and accurate solution for distribution system fault detection, which helps to improve the safety and reliability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 The present invention is a method flow chart of the method for actively enhancing detection of line fault characteristics of a distribution box. DETAILED DESCRIPTION

[0018] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0019] The present invention provides Figure 1 The line fault feature active enhanced detection method for a distribution box shown includes the following steps: Collect electrical signals in the distribution box in real time, process the signals through active enhancement strategies, highlight fault characteristics, and obtain enhanced electrical signal data; The specific steps of real-time acquisition of electrical signals in the distribution box and processing of the signals through active enhancement strategies to highlight fault characteristics are as follows: Collect electrical signals from electrical equipment in the distribution box in real time to provide raw data for subsequent processing; In the distribution box, sensors (such as current and voltage sensors) need to be installed first to collect electrical signals in real time. Common electrical signals include current, voltage, power, frequency, etc. These signals usually reflect the working status of the distribution system. Sensors can continuously monitor the current and voltage waveforms in the power grid and obtain electrical signal data at different time points. These data usually include signals under normal working conditions, and may also contain abnormal waveforms caused by faults. Through this real-time signal acquisition, the system can track the status of the power system in real time and provide raw materials for subsequent fault diagnosis.

[0020] Apply active enhancement strategies to acquired electrical signals to highlight potential fault characteristics; After acquiring the real-time electrical signal, active enhancement strategies are applied to the signal processing process. Common active enhancement strategies include nonlinear transformation (such as using high-order harmonic extraction methods) and signal superposition (such as enhancing certain frequency components by superimposing multiple signals). The purpose of these strategies is to make the fault signal more prominent in the background noise by increasing the characteristic frequency components in the signal. For example, specific frequency components may be added to the current signal to amplify the fluctuations related to the fault. This enhancement can not only help the detection system identify potential faults more keenly, but also make the characteristics of the fault more obvious. However, this process may also cause overlap with the normal operating frequency, which needs to be processed in subsequent steps.

[0021] After applying the enhancement strategy, enhanced electrical signal data is obtained to provide optimized signal input for subsequent analysis and processing.

[0022] After being processed by the active enhancement strategy, the enhanced electrical signal contains more obvious fault characteristics, especially highlighting the fluctuations related to the fault in frequency. At this time, it is necessary to further record and store these enhanced signal data through the data acquisition module in the system. These enhanced data will serve as the basis for subsequent analysis, spectrum analysis, feature extraction and further signal processing. The enhanced signal data is more diagnostically valuable than the original signal, so they become the key input data for subsequent fault detection and fault diagnosis models. Through this process, the system can track the enhanced signal in real time and ensure accurate signal data support for fault detection.

[0023] Fault signals in power distribution systems are usually drowned out by background noise, so active enhancement strategies are used to process the signals in order to amplify potential fault features and thus enhance the sensitivity of fault identification. Common enhancement methods include using nonlinear transformations (such as adding harmonic components) or signal superposition (such as adding multiple signals to enhance certain frequency components). These enhancement strategies make fault feature signals outside the normal operating frequency range more obvious, but may also introduce frequency overlap or interference problems.

[0024] The purpose of enhancement is to highlight and amplify the features related to the fault through specific signal processing methods (such as nonlinear transformation, signal superposition, etc.). By enhancing the signal, the system's ability to identify weak fault signals can be improved, making the fault features more prominent in the noise background.

[0025] The enhanced electrical signal data acquired in real time is stored in a data set, and key features reflecting the overlap between the signal frequency and the normal working frequency of the power distribution system are extracted from the data set through feature engineering technology. The extracted key features are comprehensively analyzed and the degree of frequency overlap is quantified. The data set is a structured data set that stores all enhanced electrical signals. By collecting enhanced signal data at different time points, different equipment locations, and different fault conditions, the data set provides rich information for subsequent signal analysis. This step ensures that all enhanced signal data can be effectively compared and mined during the analysis process. The purpose of establishing a data set is to facilitate the extraction and quantitative analysis of enhanced signal features, while providing sufficient samples for the training and real-time application of machine learning models.

[0026] Through feature engineering technology, key features reflecting the overlap between the signal frequency and the normal working frequency of the distribution system are extracted from the data set. The extracted features include the amplitude and phase changes of the signal and the power distribution of the enhanced signal in the spectrum. The amplitude and phase changes of the signal and the power distribution of the enhanced signal in the spectrum are comprehensively analyzed under the detection window to generate amplitude and phase error reference values ​​and power spectrum density fluctuation reference values, respectively. The degree of frequency overlap is quantified by the amplitude and phase error reference values ​​and the power spectrum density fluctuation reference values.

[0027] Abnormal changes in the amplitude and phase of the signal may indeed indicate that the frequency of the enhanced signal overlaps with the normal operating frequency of the distribution system. The reason is that the operating frequency of the power system (such as 50Hz or 60Hz) is the baseband signal in the system and usually occupies the main component in the signal. If the enhancement strategy (such as nonlinear transformation or signal superposition) changes the frequency components of the signal, and the frequency components of these enhanced signals are close to or overlap with the normal operating frequency, it will cause interference or resonance in frequency. Frequency overlap may cause abnormal changes in the amplitude and phase of the signal. First, the amplitude of the signal may be abnormally enhanced or suppressed, because frequency overlap may cause resonance, that is, the amplitude of a certain frequency component is over-amplified or weakened. Second, abnormal changes in phase usually mean that there is interference between the frequency components of the signal and the operating frequency, resulting in a shift or distortion in the phase of the signal. Generally, when the amplitude and phase of the signal fluctuate significantly, this reflects that the signal frequency interferes with or overlaps with the normal operating frequency of the system, thereby affecting the authenticity and accuracy of the signal. This phenomenon is extremely unfavorable for further processing and fault detection of the signal, and may lead to misjudgment or missed judgment.

[0028] The specific steps for comprehensively analyzing the amplitude and phase changes of the signal under the detection window to generate the amplitude and phase error reference value are as follows: The amplitude change of the signal is quantified, the amplitude error between the enhanced signal and the ideal signal (i.e. the signal at the normal working frequency) is analyzed, the amplitude distortion factor is defined, and the amplitude difference of the signal in the frequency space is measured to quantify the amplitude change degree. The calculation expression of the amplitude distortion factor is as follows: , where is the amplitude distortion factor, The enhanced signal is The frequency at the frequency point Amplitude value, The ideal signal is The frequency at the frequency point Amplitude value, is the number of frequencies, is a very small positive number used to avoid division by zero. is an exponential parameter used to perform nonlinear amplification of the amplitude difference; The above steps quantify the amplitude difference between the enhanced signal and the ideal signal, which is expressed by the amplitude distortion factor To identify abnormal changes in the signal in the frequency space. Detect whether the signal has a significant change in amplitude due to frequency overlap or interference, thus providing a basis for subsequent processing.

[0029] Analyze the phase error of the signal, quantify the phase change of the signal, and determine whether the phase of the signal has changed by calculating the phase difference between the enhanced signal and the ideal signal. Define the phase distortion factor. The calculation expression of the phase distortion factor is as follows: , where is the phase distortion factor, is a tuning parameter used to control the sensitivity of phase changes to the exponent. The enhanced signal is The frequency at the frequency point The phase angle at The ideal signal is The frequency at the frequency point The phase angle at is the phase index parameter, which is used to nonlinearly amplify the phase error. It is the exponential change of the phase error, which reflects the change of the phase deviation between the enhanced signal and the ideal signal; Through this formula, we can get a phase distortion factor with a value between 0 and 1. The closer the value is to 1, the greater the phase change of the signal, which means that the frequency of the signal may overlap or interfere with the normal operating frequency of the system.

[0030] Finally, the amplitude distortion factor and phase distortion factor Combined, the amplitude and phase error reference value is obtained to quantify the amplitude and phase changes of the enhanced signal. The calculation expression is as follows: , where is the amplitude phase error reference value.

[0031] The amplitude and phase error reference value can effectively integrate the amplitude and phase changes of the signal and reflect whether the signal overlaps with the normal working frequency of the system. When the value is large, it means that the amplitude and phase of the signal have undergone large abnormal changes, which may interfere with or resonate with the normal working frequency. A smaller value indicates that the signal does not overlap with the normal frequency and the signal status is relatively normal.

[0032] Through the above steps and formulas, the amplitude and phase changes of the signal can be systematically analyzed, so as to accurately determine whether the enhanced signal frequency overlaps with the normal operating frequency of the power distribution system.

[0033] After comprehensively analyzing the amplitude and phase changes of the signal under the detection window, an amplitude and phase error reference value is generated. By analyzing the amplitude and phase changes of the signal through the amplitude and phase error reference value, it is indeed possible to determine whether the enhanced signal frequency overlaps with the normal operating frequency of the power distribution system. When the signal frequency overlaps with the system operating frequency, it usually causes abnormal changes in the amplitude and phase of the signal, because frequency interference or resonance may cause drastic changes in the amplitude and obvious phase shift. The amplitude and phase error reference value can reflect the extent of these changes by quantifying the amplitude error and phase error of the signal. When the amplitude and phase error reference value has a large performance value, it means that the amplitude and phase changes of the signal are more significant, which usually indicates that the enhanced signal frequency overlaps with the normal operating frequency of the system (such as 50Hz or 60Hz), and this overlap will cause spectrum resonance and signal distortion. When the amplitude and phase error reference value is small, it indicates that the amplitude and phase changes of the signal are small, that is, the enhanced signal frequency does not overlap with the normal operating frequency, the signal is relatively stable, and is not affected by frequency interference. Therefore, the larger the amplitude and phase error reference value, the more it can indicate the possibility of frequency overlap, and vice versa, it means that the signal does not overlap.

[0034] When the power distribution of the enhanced signal in the spectrum fluctuates abnormally, it often indicates that the enhanced signal frequency overlaps with the normal operating frequency of the distribution system. This is because in spectrum analysis, the power spectrum density of the signal is usually expressed as the energy distribution of each frequency component, and the operating frequency of the distribution system (such as 50Hz or 60Hz) corresponds to the stable fundamental energy position. When the signal is actively enhanced, if the frequency enhancement strategy (such as harmonic amplification, signal superposition, etc.) amplifies certain frequency components and overlaps with the system fundamental frequency region, the power spectrum in this frequency band will increase significantly, forming unnatural peaks or fluctuations. This fluctuation is called "abnormal fluctuation". Its essence is that the energy of the enhanced signal is abnormally concentrated near the system operating frequency, breaking the balance of the normal spectrum energy distribution, indicating that the enhanced signal frequency overlaps with the system frequency. This overlap may cause spectrum resonance, energy leakage or signal feature distortion, which seriously interferes with the identification of fault features. Therefore, spectrum power fluctuation is an important indirect feature for determining frequency overlap.

[0035] The specific steps of comprehensively analyzing the power distribution of the enhanced signal in the spectrum under the detection window to generate the power spectrum density fluctuation reference value are as follows: In the process of signal spectrum analysis, the power spectrum density of the enhanced electrical signal is first calculated to understand the energy distribution of the signal in different frequency ranges. The power spectrum density is an important indicator of the signal frequency component. It is extracted from the time domain signal through fast Fourier transform (FFT) or other frequency domain conversion methods to calculate the power of each frequency band. The calculation expression is as follows: , where It is The power spectral density of the frequency band, It is The center frequency of the frequency band, is the bandwidth, that is, the bandwidth of each frequency band, The signals are at different frequencies The energy distribution on The purpose of the above steps is to calculate the power value in each frequency band, reflecting the energy concentration of the signal in the frequency band. The key to this step is to capture the fluctuation characteristics of the signal in different frequency bands by dividing the signal into frequency bands and calculating the power value in each frequency band.

[0036] After obtaining the power spectrum density of each frequency band, the fluctuation degree of power in the spectrum is further analyzed to quantify the change of signal frequency. The calculation expression of the reference value of power spectrum density fluctuation is as follows: , where is the power spectrum density fluctuation reference value, is the power spectral density of the previous frequency band, is the weight factor, indicating The contribution of each frequency band to power fluctuation, is the total number of frequency bands.

[0037] The above steps can quantify the fluctuation of the spectrum by calculating the power difference between adjacent frequency bands and weighted summing them. If the power difference in some frequency bands is significant, it means that the frequency components of the signal have changed significantly in these frequency bands, which may be caused by the overlap of the signal frequency with the normal operating frequency of the system (such as 50Hz or 60Hz). Ultimately, the larger the value of the power spectrum density fluctuation reference value, the more obvious the overlap between the signal frequency and the normal operating frequency of the system, and the more significant the spectrum fluctuation.

[0038] The larger the power spectrum density fluctuation reference value generated after comprehensive analysis of the power distribution of the enhanced signal in the spectrum under the detection window, the more it indicates that the enhanced signal frequency overlaps with the normal operating frequency of the power distribution system. Specifically, the power spectrum density fluctuation reference value is used to measure whether the power distribution in the signal spectrum is stable. When the signal frequency overlaps with the normal operating frequency of the power system (such as 50Hz or 60Hz), the enhanced signal spectrum may fluctuate abnormally, which means that the frequency components accumulate in an area close to the operating frequency, causing the power spectrum density to show an abnormal increase or fluctuation in this frequency band. The increase in the reference value means that the energy distribution fluctuation of the spectrum on these frequency components increases, indicating that the degree of frequency overlap is high. When the enhanced signal frequency does not overlap with the operating frequency, the power spectrum distribution of the signal is relatively stable and the fluctuation reference value is low. Therefore, this reference value can be used to effectively determine whether the enhanced signal overlaps with the system frequency.

[0039] The key features after comprehensive analysis are input into the pre-trained machine learning model for further evaluation to determine whether the enhanced signal overlaps with the normal working frequency; The amplitude phase error reference value and power spectrum density fluctuation reference value that have undergone comprehensive analysis are input into a pre-trained machine learning model for further evaluation. The frequency overlap risk coefficient is generated by the machine learning model, and the frequency overlap risk coefficient is used to determine whether the enhanced signal overlaps with the normal working frequency.

[0040] A pre-trained machine learning model is a machine learning model that has been trained with historical data, samples, and annotations, and is used in practical applications to process new input data and make predictions or classifications. During the training process, the model learns the relationship between various features in the signal and the target output (such as frequency overlap risk) by learning from a large amount of annotated data. Therefore, the pre-trained machine learning model has the ability to automatically identify key patterns, regularities, and anomalies in the signal, so that it can accurately determine whether the enhanced signal overlaps with the normal operating frequency of the system. The model training process includes selecting appropriate features (such as amplitude phase error, power spectral density fluctuation, etc.) and fitting the data through supervised learning algorithms (such as support vector machines, decision trees, neural networks, etc.). The training data set usually contains different states of multiple electrical signals (including normal working states and fault states with frequency overlap), and each sample is equipped with a label to indicate whether the sample has frequency overlap.

[0041] After training is completed, the machine learning model will continuously adjust its internal parameters through the optimization algorithm so that the model can predict new, unseen data as accurately as possible. In practical applications, after the newly collected signal is feature extracted and analyzed, these feature values ​​(such as amplitude phase error reference value and power spectrum density fluctuation reference value) will be passed as input to the pre-trained model. The model will evaluate the input features based on the previously learned rules and patterns and calculate a frequency overlap risk coefficient. This risk coefficient represents the possibility and severity of the overlap between the enhanced signal and the normal working frequency. According to the value of the frequency overlap risk coefficient, the system can determine whether the signal overlaps with the normal working frequency of the system, and further decide whether measures (such as frequency shift or signal adjustment) need to be taken. By using the trained model, the fault detection system can realize intelligent and automated signal analysis and judgment, improve the accuracy and response speed of fault detection, and avoid the shortcomings of manual intervention or traditional methods.

[0042] The machine learning model is not limited here and can be used to calculate the amplitude phase error reference value and power spectrum density fluctuation reference value Perform comprehensive analysis to generate frequency overlap risk coefficients The machine learning model can be used. To implement the technical solution of the present invention, the present invention provides a specific implementation method: Frequency Overlap Risk Factor The generation formula is as follows: , where and They are the amplitude and phase error reference values ​​respectively and power spectrum density fluctuation reference value The preset scaling factor of and Both are greater than 0.

[0043] The preset proportionality coefficient refers to the coefficient set in the model through prior knowledge or experience, which is used to weight the importance of different features, thereby affecting the model's learning and calculation of these features. and are used to respectively compare the amplitude and phase error reference values and power spectrum density fluctuation reference value The contribution of each feature to the frequency overlap risk is weighted. The degree of influence on the final calculation results.

[0044] In practical applications, and The size of may be set based on historical data, experimental results or expert experience. These proportional coefficients can help the model more accurately adjust and evaluate the contribution of different factors in the signal to the frequency overlap risk while ensuring the stability and reliability of the model. Ideally, and They should be adjusted according to the characteristics of the actual signal, and their sum will usually be equal to 1, thus ensuring that the weighted sum of the signal will not be distorted during the calculation process.

[0045] It can be seen from the frequency overlap risk coefficient that the larger the amplitude phase error reference value generated after comprehensive analysis of the amplitude and phase changes of the signal under the detection window, the larger the power spectrum density fluctuation reference value generated after comprehensive analysis of the power distribution of the enhanced signal in the spectrum under the detection window. This indicates that the larger the frequency overlap risk coefficient generated when the pre-trained machine learning model is used to judge whether the enhanced signal overlaps with the normal working frequency, the greater the probability that the enhanced signal frequency overlaps with the normal working frequency of the distribution system. Conversely, the smaller the probability that the enhanced signal frequency overlaps with the normal working frequency of the distribution system.

[0046] The frequency overlap risk coefficient generated when the pre-trained machine learning model is used to determine whether the enhanced signal overlaps with the normal working frequency is compared with the pre-set frequency overlap risk coefficient reference threshold to determine whether the enhanced signal overlaps with the normal working frequency. The judgment logic is as follows: If the frequency overlap risk coefficient is greater than the pre-set frequency overlap risk coefficient reference threshold, it is judged that the enhanced signal frequency overlaps with the normal operating frequency of the distribution system; if the frequency overlap risk coefficient is less than or equal to the pre-set frequency overlap risk coefficient reference threshold, it is judged that the enhanced signal frequency does not overlap with the normal operating frequency of the distribution system.

[0047] When it is identified that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a digital signal processing (DSP) algorithm is used to introduce a frequency shift operation in the enhanced signal to shift the frequency bandwidth of the signal to a frequency band outside the operating frequency of the power distribution system to avoid frequency overlap; then, the frequency-shifted signal is further analyzed to confirm that its frequency no longer overlaps with the operating frequency of the power distribution system, and on this basis, the fault signal is identified; The frequency shift operation is introduced through the digital signal processing (DSP) algorithm to effectively solve the problem of overlap between the enhanced signal frequency and the normal operating frequency of the distribution system (such as 50Hz or 60Hz), thereby avoiding fault detection errors caused by spectrum resonance or frequency interference. In the fault detection of the distribution system, when the active enhancement strategy is used to process the electrical signal, the signal frequency may overlap with the operating frequency of the system, causing the enhanced signal to be distorted or interfered, making it difficult to accurately identify the fault characteristics. This overlap will affect the spectral structure of the signal, causing the system to be unable to correctly distinguish between normal operating signals and potential fault signals, thereby increasing the risk of misjudgment or missed judgment.

[0048] To effectively solve this problem, digital signal processing (DSP) algorithms are used for frequency shifting. The frequency shifting operation ensures that the frequency of the signal no longer overlaps with the system operating frequency by shifting the frequency components of the enhanced signal. Specifically, the DSP algorithm shifts the frequency bandwidth of the enhanced signal to an area outside the operating frequency of the power system, thereby avoiding spectrum resonance and eliminating interference between signals. The signal after the frequency shifting operation will be in an interference-free frequency band, thereby maintaining the characteristics and stability of the enhanced signal.

[0049] Subsequently, by further analyzing the frequency-shifted signal, it can be confirmed that the signal's spectrum no longer overlaps with the operating frequency, ensuring the authenticity and accuracy of the signal. This process helps the system identify fault signals more accurately and ensures the reliability of fault detection results. The frequency shift operation not only effectively avoids interference caused by frequency overlap, but also ensures the effectiveness of the enhanced signal, providing a clearer and more accurate signal basis for subsequent fault diagnosis.

[0050] When it is identified that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a digital signal processing (DSP) algorithm is used to introduce a frequency shift operation in the enhanced signal to shift the frequency bandwidth of the signal to a frequency band outside the operating frequency of the power distribution system; then, the specific steps for further analyzing the frequency-shifted signal are as follows: After identifying the frequency overlap, the digital signal processing (DSP) algorithm is used to frequency shift the enhanced signal, shifting its frequency components from the operating frequency bandwidth of the power distribution system to the external frequency band to avoid frequency overlap. The calculation expression is as follows: When it is determined that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a frequency shift operation is introduced through digital signal processing (DSP) technology to shift the frequency bandwidth of the signal out of the operating frequency band of the power distribution system. The frequency shift operation is to apply a phase rotation factor in the frequency domain to achieve signal translation to avoid overlapping with the operating frequency (such as 50Hz). The frequency shift calculation expression is as follows: , where It is the frequency domain representation of the signal after frequency shift, that is, the enhanced signal spectrum after frequency shift. is the frequency domain representation of the original enhanced signal, that is, the enhanced signal spectrum, is the complex exponential function, is an imaginary unit, is the frequency shift (offset), is the natural base; Through the above steps, the signal frequency is shifted to an area far away from the working frequency band to avoid resonance or interference with the power distribution system frequency. This ensures that the characteristics of the enhanced signal are not interfered by the power distribution system frequency, and subsequent signal analysis and fault diagnosis can continue.

[0051] After the frequency shift operation is completed, the frequency-shifted signal needs to be further analyzed to ensure that the frequency of the signal has been completely decoupled and no longer overlaps with the operating frequency of the power distribution system. The specific operation is to recalculate the frequency overlap risk coefficient and determine whether it is lower than the reference threshold, so as to confirm the signal processing effect and identify the fault signal. The calculation expression is as follows: , where It is the frequency overlap risk coefficient recalculated after frequency shift, which is used to evaluate whether the signal after frequency shift still overlaps with the operating frequency of the power distribution system. It is a frequency overlap risk coefficient calculation model, which is used to analyze the frequency characteristics of the signal after frequency shift. It is a feature extraction function, which is used to extract new frequency domain features from the frequency-shifted signal in order to calculate the risk coefficient; like ,in, is the reference threshold of the frequency overlap risk coefficient.

[0052] Through this step, the frequency overlap risk coefficient of the frequency-shifted signal is re-evaluated and the fault signal is identified. , it means that the signal frequency has successfully left the system operating frequency band and there is no overlap anymore. Fault diagnosis and location can be performed based on the frequency-shifted signal.

[0053] The present invention ensures that the enhanced signal spectrum does not overlap with the operating frequency of the system through a series of steps such as real-time signal acquisition and enhancement, feature extraction and frequency overlap analysis, machine learning model evaluation, and frequency shift operations in digital signal processing (DSP). This process avoids the problem of fault signal distortion or misjudgment caused by spectrum interference, allowing fault signals to be clearly and accurately identified. In addition, through the combination of machine learning models and DSP algorithms, the signal processing process can be dynamically adjusted, and the impact of frequency overlap can be automatically identified and corrected, thereby improving the overall performance and stability of the fault detection system. Ultimately, this solution provides a more intelligent and accurate solution for distribution system fault detection, which helps to improve the safety and reliability of the power system.

[0054] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0055] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0056] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0057] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0058] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0059] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0060] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0061] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0062] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0063] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for actively enhancing detection of line fault characteristics of a distribution box, characterized in that: The following steps are involved: Collect electrical signals in the distribution box in real time, process the signals through active enhancement strategies, highlight fault characteristics, and obtain enhanced electrical signal data; The enhanced electrical signal data acquired in real time is stored in a data set, and key features reflecting the overlap between the signal frequency and the normal working frequency of the power distribution system are extracted from the data set through feature engineering technology. The extracted key features are comprehensively analyzed and the degree of frequency overlap is quantified. The key features after comprehensive analysis are input into the pre-trained machine learning model for further evaluation to determine whether the enhanced signal overlaps with the normal working frequency; When it is identified that the frequency of the enhanced signal overlaps with the operating frequency of the distribution system, a digital signal processing algorithm is used to introduce a frequency shift operation in the enhanced signal to shift the frequency bandwidth of the signal to a frequency band outside the operating frequency of the distribution system to avoid frequency overlap; subsequently, the frequency-shifted signal is further analyzed to confirm that its frequency no longer overlaps with the operating frequency of the distribution system, and on this basis, the fault signal is identified.

2. The method for actively enhancing line fault feature detection for a distribution box according to claim 1, characterized in that: The specific steps of real-time acquisition of electrical signals in the distribution box and processing of the signals through active enhancement strategies to highlight fault characteristics are as follows: Collect electrical signals from electrical equipment in the distribution box in real time to provide raw data for subsequent processing; Apply active enhancement strategies to acquired electrical signals to highlight potential fault characteristics; After applying the enhancement strategy, enhanced electrical signal data is obtained to provide optimized signal input for subsequent analysis and processing.

3. The line fault feature active enhancement detection method for a distribution box according to claim 1 is characterized in that: Through feature engineering technology, key features reflecting the overlap between the signal frequency and the normal working frequency of the distribution system are extracted from the data set. The extracted features include the amplitude and phase changes of the signal and the power distribution of the enhanced signal in the spectrum. The amplitude and phase changes of the signal and the power distribution of the enhanced signal in the spectrum are comprehensively analyzed under the detection window to generate amplitude and phase error reference values ​​and power spectrum density fluctuation reference values, respectively. The degree of frequency overlap is quantified by the amplitude and phase error reference values ​​and the power spectrum density fluctuation reference values.

4. The line fault feature active enhancement detection method for a distribution box according to claim 3 is characterized in that: The specific steps for comprehensively analyzing the amplitude and phase changes of the signal under the detection window to generate the amplitude and phase error reference value are as follows: The amplitude change of the signal is quantified, the amplitude error between the enhanced signal and the ideal signal is analyzed, the amplitude distortion factor is defined, and the amplitude difference of the signal in the frequency space is measured to quantify the amplitude change degree. The calculation expression of the amplitude distortion factor is as follows: , where is the amplitude distortion factor, Is the enhanced signal in The frequency at the frequency point Amplitude value, The ideal signal is The frequency at the frequency point Amplitude value, is the number of frequencies, is a very small positive number used to avoid division by zero. is an exponential parameter used to perform nonlinear amplification of the amplitude difference; Analyze the phase error of the signal, quantify the phase change of the signal, and determine whether the phase of the signal has changed by calculating the phase difference between the enhanced signal and the ideal signal. Define the phase distortion factor. The calculation expression of the phase distortion factor is as follows: , where is the phase distortion factor, is the adjustment parameter, The enhanced signal is The frequency at the frequency point The phase angle at The ideal signal is The frequency at the frequency point The phase angle at is the phase index parameter, is the exponential change of the phase error; Finally, the amplitude distortion factor and phase distortion factor Combined, the amplitude and phase error reference value is obtained to quantify the amplitude and phase changes of the enhanced signal. The calculation expression is as follows: , where is the amplitude phase error reference value.

5. The line fault feature active enhancement detection method for a distribution box according to claim 3 is characterized in that: The specific steps of comprehensively analyzing the power distribution of the enhanced signal in the spectrum under the detection window to generate the power spectrum density fluctuation reference value are as follows: In the process of signal spectrum analysis, the power spectrum density of the enhanced electrical signal is first calculated to understand the energy distribution of the signal in different frequency ranges and calculate the power of each frequency band. The calculation expression is as follows: , where It is The power spectral density of the frequency band, It is The center frequency of the frequency band, is the frequency bandwidth, The signals are at different frequencies The energy distribution on After obtaining the power spectrum density of each frequency band, the fluctuation degree of power in the spectrum is further analyzed to quantify the change of signal frequency. The calculation expression of the reference value of power spectrum density fluctuation is as follows: , where is the power spectrum density fluctuation reference value, is the power spectral density of the previous frequency band, is the weight factor, indicating The contribution of each frequency band to power fluctuation, is the total number of frequency bands.

6. The line fault feature active enhancement detection method for a distribution box according to claim 3 is characterized in that: The amplitude phase error reference value and power spectrum density fluctuation reference value that have undergone comprehensive analysis are input into a pre-trained machine learning model for further evaluation. The frequency overlap risk coefficient is generated by the machine learning model, and the frequency overlap risk coefficient is used to determine whether the enhanced signal overlaps with the normal working frequency.

7. The line fault feature active enhancement detection method for a distribution box according to claim 6, characterized in that: The frequency overlap risk coefficient generated when the pre-trained machine learning model is used to determine whether the enhanced signal overlaps with the normal working frequency is compared with the pre-set frequency overlap risk coefficient reference threshold to determine whether the enhanced signal overlaps with the normal working frequency. The judgment logic is as follows: If the frequency overlap risk coefficient is greater than the pre-set frequency overlap risk coefficient reference threshold, it is judged that the enhanced signal frequency overlaps with the normal operating frequency of the distribution system; if the frequency overlap risk coefficient is less than or equal to the pre-set frequency overlap risk coefficient reference threshold, it is judged that the enhanced signal frequency does not overlap with the normal operating frequency of the distribution system.

8. The line fault feature active enhancement detection method for a distribution box according to claim 7, characterized in that: When it is identified that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a digital signal processing algorithm is used to introduce a frequency shift operation in the enhanced signal to shift the frequency bandwidth of the signal to a frequency band outside the operating frequency of the power distribution system; then, the specific steps for further analyzing the frequency-shifted signal are as follows: After identifying the frequency overlap, the digital signal processing algorithm is used to frequency shift the enhanced signal, shifting its frequency components from the operating frequency bandwidth of the power distribution system to the external frequency band to avoid frequency overlap. The calculation expression is as follows: When it is determined that the frequency of the enhanced signal overlaps with the operating frequency of the power distribution system, a frequency shift operation is introduced through digital signal processing technology to shift the frequency bandwidth of the signal out of the operating frequency band of the power distribution system. The frequency shift calculation expression is as follows: , where is the frequency domain representation of the frequency-shifted signal, is the frequency domain representation of the original enhanced signal, that is, the enhanced signal spectrum, is the complex exponential function, is an imaginary unit, is the frequency shift, is the natural base; After the frequency shift operation is completed, the frequency-shifted signal needs to be further analyzed to ensure that the frequency of the signal has been completely decoupled and no longer overlaps with the operating frequency of the power distribution system. The calculation expression is as follows: , where is the frequency overlap risk coefficient recalculated after frequency shift, It is a frequency overlap risk coefficient calculation model, which is used to analyze the frequency characteristics of the signal after frequency shift. is the feature extraction function; like ,in, is the reference threshold of the frequency overlap risk coefficient.

Images