Cascade module non-intrusive monitoring method and system

Abstract

The invention discloses a cascade module non-intrusive monitoring method and system, and the method comprises the steps: obtaining a voltage signal of a DC bus in a preset time window length, obtaining a voltage fluctuation time domain signal, carrying out the signal conversion of the voltage fluctuation time domain signal through employing fast Fourier transform, carrying out the extraction, obtaining a plurality of main frequency components, and carrying out the extraction of the main frequency components; obtaining capacitance attenuation degree data according to each main frequency component, then judging whether the coupling strength index is related to the capacitance attenuation degree or not according to voltage fluctuation data of adjacent sub-modules of the direct current bus, and if yes, performing feature extraction on fluctuation coupling information by using a prediction neural network to obtain a monitoring feature vector, the optimized non-intrusive monitoring model is obtained through the monitoring feature vector, the optimized non-intrusive monitoring model is used for analyzing the voltage fluctuation data of the direct-current bus, the capacitance attenuation degree is obtained, and the accuracy and reliability of capacitance attenuation monitoring are improved through the method.

Description

Technical Field

[0001] The present invention relates to the technical field of cascade converter monitoring, and in particular to a non-intrusive monitoring method and system for a cascade module. Background Art

[0002] In a cascaded converter, multiple submodules are connected in series to form a DC bus voltage. When the electrolytic capacitor of a submodule is attenuated, the DC side voltage of the submodule will fluctuate, and the output voltage will produce a ripple of characteristic frequency. However, the ripple will couple with the adjacent normal submodule through the DC bus, causing its voltage to also fluctuate with the corresponding frequency. This ripple coupling effect between adjacent modules will seriously interfere with the method based on single submodule monitoring, thereby affecting the monitoring accuracy. Specifically, the submodule with parameter attenuation will produce characteristic frequency components with large amplitudes in its output voltage, but after the coupling effect of the DC bus, these characteristic frequency components will be absorbed by the adjacent normal submodules to a certain extent and show similar spectral characteristics. Therefore, it is difficult to accurately determine the location and attenuation degree of the faulty module only through the spectrum analysis of the voltage of a single submodule. At the same time, with the increase in the number of cascaded submodules, the propagation and coupling process of the ripple in the DC bus becomes more complicated, which further reduces the monitoring accuracy. This requires that in the non-intrusive monitoring method, the ripple coupling effect between sub-modules must be comprehensively considered, and the accurate positioning and quantitative analysis of the faulty sub-modules must be achieved through reasonable mathematical models and algorithms to ensure the reliable operation of the converter. Summary of the invention

[0003] In order to solve the above technical problems, the embodiments of the present invention provide a non-intrusive monitoring method and system for cascade modules to solve the technical problem of low accuracy in monitoring the degree of capacitance attenuation due to the ripple coupling effect between factor modules.

[0004] A first aspect of an embodiment of the present invention provides a non-intrusive monitoring method for a cascade module, the method comprising: The voltage signal of the DC bus in a preset time window length is obtained to obtain a voltage fluctuation time domain signal, and the voltage fluctuation time domain signal is converted and extracted by fast Fourier transform to obtain multiple main frequency components, and a harmonic characteristic sequence is obtained according to each main frequency component; A voltage waveform of the submodule is acquired according to a first preset sampling frequency, and the voltage waveform is converted into a time-frequency domain by Fourier transform to obtain a frequency domain spectrum line, and characteristic frequencies are extracted according to the frequency domain spectrum line to obtain a plurality of characteristic frequency points, and a capacitance attenuation characteristic frequency point is extracted based on the amplitude corresponding to each characteristic frequency point and a capacitance attenuation characteristic library, and an amplitude change trend sequence is obtained according to the amplitude corresponding to the capacitance attenuation characteristic frequency point; Obtain the fitting curve of the coupling strength index and the capacitance attenuation degree from the capacitance attenuation characteristic frequency points, the amplitude change trend sequence, and the voltage fluctuation data of adjacent sub-modules of the DC bus. According to the fitting curve, obtain the correlation between the coupling strength index and the capacitance attenuation degree. Based on the correlation, determine whether the coupling strength index is related to the capacitance attenuation degree. If it is related, calculate the error prediction sequence and the confidence level of the monitoring process. According to the error prediction sequence and the confidence level, analyze the voltage fluctuation data of the DC bus to obtain the monitoring result, where the monitoring result is the capacitance attenuation degree.

[0005] In a possible implementation manner of the first aspect, use the fast Fourier transform to perform signal conversion and extraction on the voltage fluctuation time-domain signal to obtain multiple main frequency components. According to each main frequency component, obtain the harmonic characteristic sequence, including: Obtain the voltage signal of the DC bus within the preset time window length to obtain multiple voltage data. Calculate the fluctuation amplitude sequence according to the preset reference voltage value and each voltage data to obtain the voltage fluctuation time-domain signal; Perform segmented windowing processing on the voltage fluctuation time-domain signal, and use the preset window function to smooth the signal to obtain the smoothed voltage fluctuation sequence; Perform Fourier transform on the smoothed voltage fluctuation sequence to obtain the frequency-domain representation sequence. Extract spectral lines from the frequency-domain representation sequence according to the preset frequency-domain analysis range to obtain multiple main frequency components. Group according to the frequency values of each main frequency component to determine the fundamental frequency component and multiple sub-harmonic frequency components; Calculate the ratio of the amplitude of each sub-harmonic frequency component to the fundamental frequency component respectively to obtain the harmonic component content ratio sequence. Determine the sub-harmonic frequency components greater than the preset harmonic limit value in the harmonic component content ratio sequence as the over-limit harmonic components to obtain the harmonic characteristic sequence.

[0006] In a possible implementation manner of the first aspect, extract characteristic frequency points according to the frequency-domain spectral lines to obtain multiple characteristic frequency points. Based on the amplitude corresponding to each characteristic frequency point and the capacitance attenuation characteristic library, extract the capacitance attenuation characteristic frequency points. According to the capacitance attenuation characteristic frequency points, obtain the corresponding amplitude change trend sequence, including: Extract the frequency-domain spectral lines to obtain the harmonic frequency sequence and the harmonic amplitude sequence; Classify the harmonic frequency sequence according to the preset frequency interval to obtain multiple harmonic frequency interval sequences. Calculate the amplitude ratio of the harmonic frequency components in each harmonic frequency interval sequence to obtain the harmonic frequency distribution characteristic sequence. Select the frequency points with an amplitude ratio greater than the preset ratio in the harmonic frequency distribution characteristic sequence as the characteristic frequency points to obtain multiple characteristic frequency points; Based on the amplitude corresponding to each characteristic frequency point, a fixed-length sliding window is used to calculate the amplitude change rate to obtain an amplitude change rate sequence, and the amplitude in the amplitude change rate sequence that is greater than a preset change rate threshold is taken as an amplitude mutation point; Calculate the spectrum similarity of the amplitude mutation point within the preset range, judge the similarity threshold, obtain the spectrum change type, match the spectrum change type with the capacitance attenuation feature library, and obtain the spectrum correlation sequence through matching calculation. Extract the capacitance attenuation characteristic frequency points according to the spectrum correlation sequence, and record the amplitude change trend sequence corresponding to the capacitance attenuation characteristic frequency points.

[0007] The capacitance attenuation characteristic frequency points, amplitude change trend sequence and voltage fluctuation data of adjacent submodules of the DC bus are used to obtain a fitting curve of the coupling strength index and the capacitance attenuation data. The correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve, including: Normalize the characteristic frequency points and amplitude change trend sequence of capacitance attenuation to obtain a normalized characteristic matrix, and use the principal component dimensionality reduction method to extract features from the normalized characteristic matrix to obtain a reduced dimensionality feature vector; The dimension-reduced feature vector is input into the trained vector machine model for regression prediction to obtain capacitance attenuation degree data, wherein the capacitance attenuation degree data includes a capacitance loss prediction value and a remaining life prediction value; According to the voltage fluctuation data and capacitance attenuation degree data of the adjacent submodules of the DC bus, a fitting curve of the coupling strength index and the capacitance attenuation data is obtained, and the correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve.

[0008] In a possible implementation of the first aspect, a fitting curve between a coupling strength index and capacitance attenuation data is obtained according to voltage fluctuation data and capacitance attenuation degree data of adjacent submodules of a DC bus, and a correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve, including: Collecting voltage fluctuation data of adjacent sub-modules of the DC bus, aligning the voltage fluctuation data using a synchronous sampling clock to obtain aligned fluctuation data, and collecting the aligned fluctuation data according to a second preset sampling frequency to obtain complete cycle waveform data; A fixed-length sliding window is used to segment the complete periodic waveform data to obtain segmented data, and the segmented data is Fourier transformed to obtain a frequency domain representation sequence. Multiple characteristic harmonic frequency points are extracted from the frequency domain representation sequence. According to each characteristic harmonic frequency point, the frequency offset between adjacent submodules is calculated to obtain a frequency deviation sequence. The characteristic harmonic frequency point in the frequency deviation sequence whose frequency offset is greater than a preset frequency threshold is taken as an abnormal frequency point; According to the abnormal frequency points, the characteristic harmonic amplitude ratio of adjacent sub-modules is calculated, and the waveform correlation coefficient is calculated using the adjacent window data to obtain the coupling characteristic sequence. The average value of the waveform correlation coefficient sequence is calculated to obtain the coupling strength index. The coupling strength index and the capacitance attenuation degree data are polynomially fitted to obtain the fitting curve. The fitting curve is extracted to obtain the correlation between the coupling strength index and the capacitance attenuation degree.

[0009] In a possible implementation manner of the first aspect, calculating an error prediction sequence and a confidence level of a monitoring process includes: The prediction neural network is used to extract the characteristics of the wave coupling information and obtain the monitoring characteristic vector; The monitoring feature vector is input into the constructed coupling relationship and the correlation model of monitoring accuracy for processing, and the error prediction sequence and confidence level of the monitoring process are obtained.

[0010] In a possible implementation of the first aspect, using a prediction neural network to extract features from the fluctuation coupling information to obtain a monitoring feature vector includes: Normalize the wave coupling information to obtain a standardized characteristic matrix; A processor is used to perform convolution operation on the standardized feature matrix to obtain a coupling feature map; A second layer of convolution operation is performed on the coupling feature map to obtain a propagation feature map. A pooling operation is used to reduce the dimension of the propagation feature map to obtain reduced-dimensional features. The fully connected layer is used to map and transform the reduced-dimensional features to obtain a monitoring feature vector.

[0011] In a possible implementation manner of the first aspect, the prediction neural network is obtained by training using sample fluctuation coupling information, including: Obtain sample fluctuation coupling information, input the sample fluctuation coupling information into the initial prediction neural network for prediction, and obtain the initial monitoring result; Calculate the mean square error of the initial monitoring results to obtain the monitoring error sequence, perform probability distribution fitting on the monitoring error sequence, and obtain the standard deviation and mean parameters; The monitoring confidence interval is calculated based on the standard deviation and mean parameters to obtain the monitoring accuracy assessment result; According to the monitoring accuracy evaluation results and the monitoring error sequence, the parameters of the initial prediction neural network are optimized, the feature extraction parameters and prediction parameters are updated, and the prediction neural network is obtained.

[0012] In a possible implementation of the first aspect, voltage fluctuation data of the DC bus is analyzed according to the error prediction sequence and the confidence level to obtain a monitoring result, including: Calculate the filter center frequency and bandwidth parameters based on the error prediction sequence and confidence level; The bus voltage fluctuation data is filtered by using the filter center frequency parameter and bandwidth parameter to obtain the filtered data, and the filtered data is extracted to obtain a multi-band fluctuation feature sequence; The multi-band fluctuation feature sequence is decomposed by wavelet multi-scale to obtain the harmonic feature sequence, and the support vector machine is used to classify the harmonic feature sequence to obtain the attenuation feature vector; Calculate the capacitance attenuation quantitative index according to the attenuation characteristic vector, normalize the capacitance attenuation index to obtain a correction result sequence, compare the correction result sequence with the original capacitance attenuation degree to obtain a correction deviation, update the filter parameters according to the correction deviation, and obtain an optimized non-intrusive monitoring model; The optimized non-intrusive monitoring model is used to analyze the voltage fluctuation data of the DC bus to obtain the monitoring results.

[0013] In order to solve the same technical problem, a second aspect of an embodiment of the present invention provides a cascade module non-intrusive monitoring system, the system comprising: An acquisition module is used to acquire a voltage signal of a DC bus in a preset time window length, obtain a voltage fluctuation time domain signal, perform signal conversion and extraction on the voltage fluctuation time domain signal using a fast Fourier transform, obtain multiple main frequency components, and obtain a harmonic characteristic sequence according to each main frequency component; A first calculation module is used to obtain a voltage waveform of a submodule according to a first preset sampling frequency, perform a time-frequency domain conversion on the voltage waveform using Fourier transform to obtain a frequency domain spectrum, perform characteristic frequency extraction according to the frequency domain spectrum to obtain a plurality of characteristic frequency points, extract a capacitance attenuation characteristic frequency point based on the amplitude corresponding to each characteristic frequency point and a capacitance attenuation characteristic library, and obtain an amplitude change trend sequence according to the amplitude corresponding to the capacitance attenuation characteristic frequency point; The second calculation module is used to obtain a fitting curve of a coupling strength index and a capacitance attenuation degree by using the capacitance attenuation characteristic frequency point, the amplitude change trend sequence and the voltage fluctuation data of adjacent sub-modules of the DC bus, obtain the correlation between the coupling strength index and the capacitance attenuation degree according to the fitting curve, judge whether the coupling strength index is correlated with the capacitance attenuation degree based on the correlation, and if so, calculate the error prediction sequence and the confidence level of the monitoring process, analyze the voltage fluctuation data of the DC bus according to the error prediction sequence and the confidence level, and obtain a monitoring result, wherein the monitoring result is the capacitance attenuation degree.

[0014] The technical solution of the present invention has the following advantages: The non-intrusive monitoring method of the cascade module provided in the embodiment of the present invention obtains the voltage signal of the DC bus in a preset time window length to obtain a voltage fluctuation time domain signal, uses fast Fourier transform to convert and extract the voltage fluctuation time domain signal to obtain multiple main frequency components, obtains capacitance attenuation degree data according to each of the main frequency components, and then judges whether the coupling strength index is related to the capacitance attenuation degree according to the voltage fluctuation data of the adjacent sub-modules of the DC bus. If so, a predictive neural network is used to extract features of the fluctuation coupling information to obtain a monitoring feature vector, and an optimized non-intrusive monitoring model is obtained by using the monitoring feature vector. The voltage fluctuation data of the DC bus is analyzed by using the optimized non-intrusive monitoring model to obtain the capacitance attenuation degree. Through the above method, the accuracy and reliability of capacitance attenuation monitoring are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 Flow chart of a non-intrusive monitoring method for cascade modules in an embodiment of the present invention; Figure 2 4 is a structural block diagram of a non-intrusive monitoring system of cascade modules in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] In the description of the present invention, it should be noted that the terms "first", "second" and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0019] The non-intrusive monitoring method of the cascade module provided by the embodiment of the present invention is as follows: Figure 1 As shown, Figure 1 The non-intrusive monitoring flow chart of the cascade module includes steps S101 to S105, and each step is as follows: S101: Acquire a voltage signal of a DC bus in a preset time window length to obtain a voltage fluctuation time domain signal, use fast Fourier transform to convert and extract the voltage fluctuation time domain signal to obtain multiple main frequency components, and obtain a harmonic characteristic sequence based on each of the main frequency components.

[0020] In this embodiment, the voltage sampling points on the DC bus are collected, and a complete waveform segment is intercepted from the voltage sampling points according to a preset time window length to obtain a voltage fluctuation time domain signal. Specifically, the voltage sampling points at multiple time intervals are collected for the DC bus, and a complete waveform segment is intercepted from the sampling signal according to a preset time window length, and a fluctuation amplitude sequence is calculated by comparing with a preset reference voltage value to generate a voltage fluctuation time domain signal.

[0021] In one embodiment, the voltage fluctuation time domain signal is converted and extracted by fast Fourier transform to obtain multiple main frequency components. According to each main frequency component, a harmonic characteristic sequence is obtained, including: Acquire a voltage signal of the DC bus in a preset time window length to obtain a plurality of voltage data, calculate a fluctuation amplitude sequence according to a preset reference voltage value and each voltage data, and obtain a voltage fluctuation time domain signal; The voltage fluctuation time domain signal is processed by segmented windowing, and the signal is smoothed by using a preset window function to obtain a smoothed voltage fluctuation sequence; Perform Fourier transform on the smoothed voltage fluctuation sequence to obtain a frequency domain representation sequence, perform spectrum line extraction on the frequency domain representation sequence according to a preset frequency domain analysis range to obtain multiple main frequency components, group the main frequency components according to their frequency values, and determine the fundamental frequency component and multiple subharmonic frequency components; The ratios of the amplitudes of the various subharmonic frequency components to the fundamental frequency component are calculated to obtain a harmonic component content ratio sequence, and the subharmonic frequency components in the harmonic component content ratio sequence that are greater than the preset harmonic limit are determined as out-of-limit harmonic components to obtain a harmonic characteristic sequence.

[0022] In this embodiment, the voltage sampling points on the DC bus are collected, and the complete waveform segment is intercepted from the voltage sampling points according to the preset time window length to obtain the voltage fluctuation time domain signal, and the voltage fluctuation time domain signal is subjected to segmented windowing processing, and the signal is smoothed using a preset window function to generate a smoothed voltage fluctuation sequence. The smoothed voltage fluctuation sequence is subjected to Fourier transform to obtain a frequency domain representation sequence, and the frequency domain analysis range is determined according to the sampling theorem. The frequency domain representation sequence is subjected to spectral line extraction within the frequency domain analysis range, and the amplitude of each frequency point is calculated. The main frequency components of the first percentage are obtained by sorting the amplitude from large to small.

[0023] Then, the main frequency components are grouped according to the frequency numerical relationship to determine the fundamental frequency component and each harmonic frequency component. For each harmonic frequency component, the ratio of its amplitude to the amplitude of the fundamental component is calculated to obtain the harmonic component content ratio sequence. According to the content ratio sequence, it is compared with the preset harmonic limit, the over-limit harmonic components are marked, and the harmonic characteristic sequence is generated.

[0024] As an example of this embodiment, when sampling the DC bus voltage, the voltage signal is discretized by selecting a suitable sampling interval. For the original voltage value collected by the bus end measurement device, 8000 data points are collected per second. On the basis of forming a fluctuation sequence, a preset reference voltage value is set to 500 volts. When the fluctuation of the measured voltage value exceeds the range of plus or minus 5% of the reference value, the corresponding sampling point and its time are recorded.

[0025] In the process of processing the DC bus voltage, the time window width is set to 0.2 seconds for the measured voltage fluctuation time domain signal, and the window function is used to achieve smoothing. The selection of the window function directly affects the accuracy of the spectrum analysis. When the window function width is set too large, the signal after smoothing is distorted, and when the window function width is too small, the measurement noise interference cannot be suppressed. Fast Fourier transform is applied to the smoothed voltage fluctuation sequence, and the spectrum analysis range is determined to be 4000 Hz according to the sampling theorem. The time domain signal is converted to the frequency domain space through Fourier transform, and the frequency domain representation sequence obtained contains amplitude spectrum and phase spectrum information.

[0026] In the frequency domain, the spectrum obtained is identified and extracted, and the spectrum amplitude screening threshold is set to 3% of the fundamental amplitude to extract the main frequency components. The main frequency components obtained from the frequency domain analysis are grouped according to the frequency size relationship. In the AC power supply system, the fundamental frequency is 50 Hz, and the harmonic order is determined by the multiple relationship between other frequency components and the fundamental frequency, such as 100 Hz for the second harmonic component and 150 Hz for the third harmonic component. When non-integer multiple frequency components appear, it may be caused by subsynchronous oscillation or equipment failure in the power system. For the identified harmonic components, the content ratio is obtained by calculating the ratio of the harmonic amplitude to the fundamental amplitude. The amplitude of the fundamental component is 12 volts, the second harmonic amplitude is 0.6 volts, and the third harmonic amplitude is 0.36 volts. The second harmonic content is 5% and the third harmonic content is 3%. According to the power system harmonic limit standard, the second harmonic content limit is 4% and the third harmonic content limit is 3%. By comparison, it is found that the second harmonic exceeds the limit, and it needs to be marked and corresponding suppression measures are taken. In the bus voltage fluctuation spectrum analysis, the voltage fluctuation signal is segmented and windowed and frequency-domain converted to achieve the extraction and analysis of each harmonic component. From the original voltage fluctuation data obtained by sampling, the fundamental frequency and the second and third harmonic frequency components are identified, the proportion of each harmonic content is analyzed, and the degree of harmonic pollution is evaluated. Harmonic components that exceed the limit requirements are marked to provide a basis for subsequent harmonic suppression and power quality management.

[0027] S102: Acquire the voltage waveform of the submodule according to the first preset sampling frequency, perform time-frequency domain conversion on the voltage waveform using Fourier transform to obtain frequency domain spectrum lines, extract characteristic frequencies according to the frequency domain spectrum lines to obtain multiple characteristic frequency points, extract capacitance attenuation characteristic frequency points based on the amplitude corresponding to each characteristic frequency point and the capacitance attenuation characteristic library, and obtain an amplitude change trend sequence according to the amplitude corresponding to the capacitance attenuation characteristic frequency point.

[0028] In this embodiment, the capacitance attenuation characteristic harmonics of the submodule are obtained, the difference between the amplitude and frequency distribution of the main harmonic components and the capacitance attenuation characteristic harmonics is analyzed, and it is determined whether there are characteristic harmonic components related to capacitance attenuation. If so, the characteristic frequency and amplitude change trend of the corresponding main harmonic components are recorded. Specifically, the voltage waveform is obtained from the submodule using the first preset sampling frequency, and the voltage waveform is converted into the time-frequency domain by Fourier transform to obtain the frequency domain spectrum line, and the harmonic frequency sequence and the harmonic amplitude sequence are extracted from the frequency domain spectrum line. Classification is performed within the preset frequency range according to the harmonic frequency sequence, and the amplitude proportion of the harmonic component in each frequency range is calculated by the harmonic amplitude sequence to obtain the harmonic frequency distribution characteristic sequence.

[0029] The characteristic frequency is extracted from the harmonic frequency distribution feature sequence, sorted according to the amplitude proportion in the frequency interval, and the frequency points whose proportion exceeds the preset threshold are selected as characteristic frequency points. For the amplitude corresponding to the characteristic frequency points, a fixed-length sliding window is used to calculate the amplitude change rate sequence, and the amplitude mutation point is identified by the preset change rate threshold.

[0030] Compare the frequency characteristics before and after the amplitude mutation point, calculate the spectrum similarity before and after the mutation, and judge the spectrum change type according to the similarity threshold. Match the spectrum change type with the capacitance attenuation feature library, and obtain the spectrum correlation sequence through matching calculation. Extract the capacitance attenuation characteristic frequency points according to the spectrum correlation sequence, and record the amplitude change trend sequence corresponding to the characteristic frequency points.

[0031] In one embodiment, characteristic frequency extraction is performed according to the frequency domain spectrum to obtain multiple characteristic frequency points, and the capacitance attenuation characteristic frequency points are extracted based on the amplitude corresponding to each characteristic frequency point and the capacitance attenuation characteristic library. According to the capacitance attenuation characteristic frequency points, the corresponding amplitude change trend sequence is obtained, including: Extract the frequency domain spectrum to obtain the harmonic frequency sequence and harmonic amplitude sequence; Classifying the harmonic frequency sequence according to a preset frequency interval to obtain a plurality of harmonic frequency interval sequences, calculating the amplitude proportion of the harmonic frequency components in each harmonic frequency interval sequence to obtain a harmonic frequency distribution characteristic sequence, and selecting a frequency point whose amplitude proportion is greater than a preset proportion as a characteristic frequency point in the harmonic frequency distribution characteristic sequence to obtain a plurality of characteristic frequency points; Based on the amplitude corresponding to each characteristic frequency point, a fixed-length sliding window is used to calculate the amplitude change rate to obtain an amplitude change rate sequence, and the amplitude in the amplitude change rate sequence that is greater than a preset change rate threshold is taken as an amplitude mutation point; Calculate the spectrum similarity of the amplitude mutation point within the preset range, judge the similarity threshold, obtain the spectrum change type, match the spectrum change type with the capacitance attenuation feature library, and obtain the spectrum correlation sequence through matching calculation. Extract the capacitance attenuation characteristic frequency points according to the spectrum correlation sequence, and record the amplitude change trend sequence corresponding to the capacitance attenuation characteristic frequency points.

[0032] In this embodiment, voltage data is collected for the submodule, a complete cycle voltage waveform is obtained by a preset sampling frequency, and then the waveform is converted into time-frequency domain using Fourier transform, and the harmonic frequency sequence and harmonic amplitude sequence are extracted from the frequency domain spectrum.

[0033] The harmonic frequency sequence is classified using the preset frequency interval, and the amplitude proportion of the harmonic component in each frequency interval is calculated to obtain the harmonic frequency distribution characteristic sequence. The characteristic frequency is extracted from the harmonic frequency distribution characteristic sequence, sorted according to the amplitude proportion in the frequency interval, and the frequency points whose proportion exceeds the preset threshold are selected as characteristic frequency points. For the amplitude corresponding to the characteristic frequency point, a fixed-length sliding window is used to calculate the amplitude change rate sequence, and then the amplitude mutation point is identified by the preset change rate threshold. The frequency characteristics before and after the amplitude mutation point are then compared to obtain the spectrum similarity before and after the mutation, and the spectrum change type is judged by the similarity threshold.

[0034] In addition, the spectrum change type is matched with the capacitance attenuation feature library, and the spectrum correlation sequence is obtained by matching degree calculation. The capacitance attenuation characteristic frequency point is extracted according to the spectrum correlation sequence, and the amplitude change trend sequence corresponding to the capacitance attenuation characteristic frequency point is recorded.

[0035] As an example of this embodiment, a sampling frequency of 10 kHz is selected when sampling the submodule voltage, and the voltage waveform is digitally collected. The voltage data sequence collected under normal working conditions shows periodic change characteristics. The time domain waveform is converted to the frequency domain space by Fourier transform to obtain a harmonic frequency sequence with a frequency range of 0 to 5000 Hz. According to the size of the harmonic frequency, the frequency interval is divided into three frequency intervals: low frequency band 0 to 500 Hz, medium frequency band 500 to 2000 Hz, and high frequency band 2000 to 5000 Hz, and the amplitude proportion of the harmonic components in each frequency interval is calculated respectively. Under normal working conditions, the harmonic frequency is mainly distributed in the low frequency band, the fundamental frequency is 50 Hz, the second harmonic is 100 Hz, and the third harmonic is 150 Hz. The amplitude of each harmonic decreases with increasing frequency. When the submodule capacitance decays, a new characteristic frequency component appears in the mid-frequency band. The amplitude of the characteristic frequency component is tracked through a sliding window of 500 sampling points, and the mutation point with an amplitude change rate exceeding 5% per second is recorded. The spectrum characteristics before and after the amplitude mutation point are analyzed, and the spectrum similarity of the 100 sampling points before and after the mutation is calculated. When the similarity is less than 0.8, it is judged that the spectrum has changed significantly. The types of spectrum changes include abnormal increase in harmonic component amplitude, frequency drift, bandwidth broadening, etc. These characteristics correspond to the degradation of capacitance parameters.

[0036] Typical degradation samples are extracted from the capacitor attenuation database, and the relationship between spectrum features and capacitor parameters is established. The spectrum correlation calculation threshold is set to 0.75. When the correlation between the measured spectrum and the sample spectrum exceeds the threshold, the corresponding frequency point is extracted as the capacitor attenuation characteristic frequency. When the capacitor is working normally, the amplitude of the characteristic frequency point remains below 3% of the fundamental amplitude. When the capacitor is attenuated, the amplitude of the characteristic frequency point shows an upward trend, and the amplitude increases by more than 5% within 24 hours. The harmonic characteristics caused by capacitor attenuation are mainly reflected in the change of spectrum distribution. The health status of the capacitor is judged by monitoring the amplitude change trend of the characteristic frequency point. Under normal conditions, the harmonic amplitude of the low-frequency band is stable, and the harmonic amplitude of the medium- and high-frequency bands is small. When the capacitor parameters are degraded, a new characteristic frequency component appears in the medium-frequency band, and the amplitude shows an increasing trend. The selection of the characteristic frequency is based on the spectrum similarity calculation. The setting of the correlation threshold takes into account the influence of measurement error and environmental noise, and the degree of degradation is judged by the amplitude change rate.

[0037] S103: The capacitor attenuation characteristic frequency points, the amplitude change trend sequence and the voltage fluctuation data of the adjacent sub-modules of the DC bus are used to obtain a fitting curve of the coupling strength index and the capacitor attenuation degree, and the correlation between the coupling strength index and the capacitor attenuation degree is obtained according to the fitting curve. Based on the correlation, it is determined whether the coupling strength index is correlated with the capacitor attenuation degree. If so, the error prediction sequence and the confidence level of the monitoring process are calculated, and the voltage fluctuation data of the DC bus is analyzed according to the error prediction sequence and the confidence level to obtain a monitoring result, wherein the monitoring result is the capacitor attenuation degree.

[0038] In this embodiment, a mapping relationship model between capacitor attenuation and harmonic components is established based on the frequency and amplitude variation trends of characteristic harmonic components and historical data related to capacitor attenuation. The mapping relationship model is trained using a support vector machine algorithm. The mapping relationship model outputs the remaining life of the capacitor and the capacitor loss to reflect the degree of capacitor attenuation.

[0039] According to the capacitor attenuation sample database, characteristic harmonic frequency sequence, amplitude change curve, capacitance loss value and remaining life value are obtained. The characteristic harmonic frequency sequence and amplitude change curve constitute the capacitance attenuation characteristic data set. The harmonic frequency and amplitude data in the capacitance attenuation characteristic data set are normalized to the maximum and minimum values ​​to obtain a normalized feature matrix, which contains the harmonic frequency distribution characteristics and amplitude change trend characteristics; the normalized feature matrix is ​​extracted by the principal component dimensionality reduction method to obtain a reduced dimension feature vector, which contains the main feature components; the support vector machine is used to perform regression prediction on the reduced dimension feature vector to obtain the capacitance loss prediction value and the remaining life prediction value, and the regression predictor is optimized by five-fold cross validation.

[0040] In one embodiment, the capacitance attenuation characteristic frequency point, the amplitude change trend sequence and the voltage fluctuation data of the adjacent submodules of the DC bus are used to obtain a fitting curve of the coupling strength index and the capacitance attenuation data, and the correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve, including: Normalize the characteristic frequency points and amplitude change trend sequence of capacitance attenuation to obtain a normalized characteristic matrix, and use the principal component dimensionality reduction method to extract features from the normalized characteristic matrix to obtain a reduced dimensionality feature vector; The dimension-reduced feature vector is input into the trained vector machine model for regression prediction to obtain capacitance attenuation degree data, wherein the capacitance attenuation degree data includes a capacitance loss prediction value and a remaining life prediction value; According to the voltage fluctuation data and capacitance attenuation degree data of the adjacent submodules of the DC bus, a fitting curve of the coupling strength index and the capacitance attenuation data is obtained, and the correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve.

[0041] In this embodiment, according to the capacitor attenuation samples recorded in the historical database, characteristic harmonic frequency sequences, amplitude change curves, capacitance loss values, and remaining life values ​​are extracted from the sample data to generate a capacitor attenuation feature data set. The harmonic frequency and amplitude data in the feature data set are normalized to the maximum and minimum values ​​to generate a normalized feature matrix, and the harmonic frequency distribution characteristics and amplitude change trend characteristics are extracted from the normalized feature matrix. The principal component dimensionality reduction method is used to reduce the dimension of the feature matrix, extract the main characteristic components, and generate a reduced dimension feature vector. For the reduced dimension feature vector and the corresponding capacitance loss value and remaining life value, a support vector machine is used to construct a regression predictor, and the predictor parameters are optimized through five-fold cross validation. The predicted mean square error is calculated based on the cross-validation results, and the samples whose errors exceed the preset threshold are marked to generate abnormal sample records. Then, the optimized predictor is used to predict and calculate the harmonic feature data collected in real time, and the capacitance loss value and the remaining life value are output. According to the comparison between the predicted output value and the preset threshold, the degree of capacitance attenuation is judged, and a decay state indicator sequence is generated.

[0042] As an example of this embodiment, the capacitor attenuation sample data contains multiple dimensional features, the harmonic frequency range is between 0 and 5000 Hz, the amplitude change curve records the amplitude change trend within 720 hours, the capacitor loss value represents the attenuation ratio of the capacitor parameter relative to the rated value, and the remaining life value reflects the continued service life of the capacitor.

[0043] In the feature data normalization process, the frequency data is divided by 5000 Hz to obtain the normalized frequency value, and the amplitude data is divided by the base wave amplitude to obtain the normalized amplitude. The loss value and life value are already in percentage form. The feature matrix contains the amplitude data of 50 frequency points. The 50-dimensional features are reduced to 10-dimensional principal component features through principal component analysis. The principal component features retain 85% of the information of the original data. The reduced feature vector is used as the input data of the support vector machine, and the corresponding capacitor loss value and remaining life value are used as the output label. The radial basis kernel function is used to construct a nonlinear mapping relationship.

[0044] Cross-validation uses a five-fold cross-validation method, randomly dividing the sample data into five parts, of which four are used for training and one is used for validation, and the complete validation results are obtained after five cycles. Taking the loss value as an example, the prediction error of the normal sample is within 3%. When the prediction error exceeds 5%, the sample is marked as an abnormal sample, and the characteristic distribution law of the abnormal sample is analyzed. In practical applications, the rated life of a certain type of capacitor is 40,000 hours. The harmonic characteristics in the capacitor voltage are obtained through online monitoring. After feature extraction and normalization, the input predictor calculates the capacitor loss value of 12% and the remaining life of 28,000 hours. According to the preset loss threshold of 20% and the life threshold of 10,000 hours, it is judged that the capacitor is in normal working condition. The capacitor attenuation prediction adopts a data-driven method, and the mapping relationship between features and attenuation degree is established through the accumulation of historical data. The feature extraction process focuses on data noise reduction and dimensionality reduction to avoid the influence of redundant features on prediction accuracy. The construction of the predictor is based on the nonlinear regression principle of the support vector machine, and cross-validation ensures the rationality of model parameter selection.

[0045] In practical applications, the online evaluation of the capacitor status is realized by combining the preset threshold value, providing data support for preventive maintenance. When the capacitor is in normal working condition, the harmonic characteristics remain stable and the error of the prediction result is small. When the capacitor parameters change significantly, the harmonic characteristics change accordingly, and the prediction result can timely reflect the changing trend of the capacitor status.

[0046] In one embodiment, a fitting curve between a coupling strength index and capacitance attenuation data is obtained based on voltage fluctuation data and capacitance attenuation degree data of adjacent submodules of a DC bus, and a correlation between the coupling strength index and capacitance attenuation degree is obtained based on the fitting curve, including: Collecting voltage fluctuation data of adjacent sub-modules of the DC bus, aligning the voltage fluctuation data using a synchronous sampling clock to obtain aligned fluctuation data, and collecting the aligned fluctuation data according to a second preset sampling frequency to obtain complete cycle waveform data; A fixed-length sliding window is used to segment the complete periodic waveform data to obtain segmented data, and the segmented data is Fourier transformed to obtain a frequency domain representation sequence. Multiple characteristic harmonic frequency points are extracted from the frequency domain representation sequence. According to each characteristic harmonic frequency point, the frequency offset between adjacent submodules is calculated to obtain a frequency deviation sequence. The characteristic harmonic frequency point in the frequency deviation sequence whose frequency offset is greater than a preset frequency threshold is taken as an abnormal frequency point; According to the abnormal frequency points, the characteristic harmonic amplitude ratio of adjacent sub-modules is calculated, and the waveform correlation coefficient is calculated using the adjacent window data to obtain the coupling characteristic sequence. The average value of the waveform correlation coefficient sequence is calculated to obtain the coupling strength index. The coupling strength index and the capacitance attenuation degree data are polynomially fitted to obtain the fitting curve. The fitting curve is extracted to obtain the correlation between the coupling strength index and the capacitance attenuation degree.

[0047] In this embodiment, the voltage fluctuation data of the adjacent submodules of the DC bus are obtained to obtain the fluctuation coupling information between the adjacent submodules and the DC bus. By analyzing the propagation of the characteristic harmonic components in the adjacent submodules, the correlation between the coupling strength and the capacitance attenuation degree is determined.

[0048] The voltage fluctuation data of the adjacent submodules of the DC bus is collected, and the fluctuation data is aligned using a synchronous sampling clock, and the complete cycle waveform is obtained by the preset sampling frequency. The synchronous sampling data is segmented using a fixed-length sliding window, and the segmented data is Fourier transformed to obtain a frequency domain representation sequence. The characteristic harmonic frequency points are extracted from the frequency domain representation sequence, and the frequency offset between adjacent submodules is calculated to generate a frequency deviation sequence. The frequency deviation sequence is compared with the preset frequency threshold, and the frequency deviation exceeding the limit frequency point is marked to generate a frequency anomaly record. Then, the characteristic harmonic amplitude ratio of adjacent submodules is calculated, and the waveform correlation coefficient is calculated using the adjacent window data to generate a coupling characteristic sequence. The propagation loss between adjacent submodules is calculated based on the coupling characteristic sequence, and the coupling strength index is obtained from the propagation loss data. Then, the coupling strength index and the submodule capacitance attenuation data are polynomially fitted, and the variation law of the coupling strength index with the capacitance attenuation degree is extracted from the fitting curve to obtain the correlation between the coupling strength index and the capacitance attenuation degree.

[0049] As an example of this embodiment, the voltage fluctuations of adjacent submodules of the DC bus have coupled propagation characteristics. The adjacent submodules are synchronously sampled at a sampling frequency of 10 kHz. The time interval between adjacent sampling points is 0.1 milliseconds, and the sampling clock deviation is controlled within 1 microsecond to ensure the synchronization of the waveform data. By setting the sampling length to a data segment of 1 second, the complete cycle waveform of the adjacent submodule is recorded. In the waveform analysis, a sliding window with a length of 0.2 seconds is used to segment the data, the overlap rate of adjacent windows is set to 50%, and the data of each window is Fourier transformed. In the frequency domain space, the fundamental frequency is 50 Hz, and the main characteristic harmonics are distributed in the frequency band below 500 Hz. The frequency points in the frequency band with an amplitude exceeding 3% of the fundamental amplitude are recorded as characteristic harmonic frequency points. For the characteristic harmonic frequency points, the frequency offset between adjacent submodules is calculated. Under normal working conditions, the frequency offset is within 1 Hz. When the frequency offset exceeds 2 Hz, it is marked as an abnormal frequency point.

[0050] The harmonic amplitude ratio of adjacent submodules reflects the attenuation characteristics of the wave propagation process. The amplitude ratio between 0.9 and 1.1 indicates that the propagation loss is small and the coupling effect is strong. The waveform correlation coefficient calculation uses adjacent window data of 0.1 seconds in length, with a calculation interval of 0.02 seconds to obtain the waveform correlation coefficient sequence. A correlation coefficient greater than 0.8 indicates good waveform synchronization and high coupling strength, and a correlation coefficient less than 0.5 indicates severe waveform distortion and low coupling strength. The average value is calculated from the correlation coefficient sequence as the coupling strength indicator.

[0051] In actual applications, when the capacitance attenuation of a certain model of submodule reaches 15%, the coupling strength index between adjacent modules decreases from 0.85 to 0.65, and the propagation loss increases. The corresponding relationship between coupling strength and capacitance attenuation is obtained by polynomial fitting. The fitting curve shows that the coupling strength shows a decreasing trend with the increase of capacitance attenuation. When the capacitance attenuation exceeds 20%, the coupling strength index is lower than 0.5, and the wave propagation characteristics change significantly. The coupling effect between submodules is closely related to the capacitance parameters. The change of capacitance state can be reflected by analyzing the wave propagation characteristics. The coupling strength index comprehensively considers the characteristics of frequency offset, amplitude attenuation, waveform correlation, etc. to evaluate the capacitance attenuation. Within the normal range of capacitance parameters, the coupling effect of adjacent submodules remains stable, and the wave propagation loss is small. As the capacitance parameters deteriorate, the coupling effect weakens, the propagation loss increases, and the waveform distortion intensifies.

[0052] In one embodiment, calculating the error prediction sequence and confidence level of the monitoring process includes: The prediction neural network is used to extract the characteristics of the wave coupling information and obtain the monitoring characteristic vector; The monitoring feature vector is input into the constructed coupling relationship and the correlation model of monitoring accuracy for processing, and the error prediction sequence and confidence level of the monitoring process are obtained.

[0053] In this embodiment, whether the coupling strength index is related to the capacitance attenuation degree is determined based on the correlation. If so, a predictive neural network is used to extract features of the fluctuation coupling information, including coupling strength, propagation speed, etc., and then combined with the capacitance attenuation quantitative index, a correlation model of the coupling relationship and monitoring accuracy is constructed to obtain the error prediction sequence and confidence level of the monitoring process.

[0054] According to the wave coupling information matrix, a data set is constructed, and the coupling intensity sequence, propagation speed sequence, and attenuation quantization sequence are normalized to the maximum and minimum values ​​to generate a standardized feature matrix. A multi-layer convolution structure is constructed. The first layer uses a processor to perform convolution operations on the standardized feature matrix to generate a coupling feature map. The second layer of convolution operations is performed on the coupling feature map to extract the propagation feature map, and pooling operations are used to reduce the feature dimension. The reduced dimensionality features are mapped and transformed through the fully connected layer to generate a monitoring feature vector.

[0055] In one embodiment, a prediction neural network is used to extract features from the wave coupling information to obtain a monitoring feature vector, including: Normalize the wave coupling information to obtain a standardized characteristic matrix; A processor is used to perform convolution operation on the standardized feature matrix to obtain a coupling feature map; A second layer of convolution operation is performed on the coupling feature map to obtain a propagation feature map. A pooling operation is used to reduce the dimension of the propagation feature map to obtain reduced-dimensional features. The reduced-dimensional features are mapped and transformed using a fully connected layer to obtain a monitoring feature vector.

[0056] In this embodiment, the wave coupling information includes three types of feature data: coupling strength, propagation speed and attenuation quantization. The coupling strength value ranges from 0 to 1, the propagation speed value ranges from 0 to 1000 meters per second, and the attenuation quantization value ranges from 0 to 100%. All feature data are mapped to the interval of 0 to 1 through maximum and minimum value normalization to ensure the comparability of features of different dimensions. The convolution structure adopts a two-layer design. The first layer uses 16 3×3 convolution kernels to process the input features, the step size is set to 1, the filling method uses zero padding, and 16 feature maps are output. The feature map reflects the local features and spatial correlation of the coupling data, and enhances the feature expression ability through nonlinear activation functions. The second layer uses 32 3×3 convolution kernels for feature extraction, and the step size is set to 2 to achieve feature dimensionality reduction. The pooling layer uses a 2×2 maximum pooling window with a step size of 2 to extract the most significant features of the local area, reduce the data dimension, and improve the robustness of feature expression. The fully connected layer contains 128 neurons, uses a nonlinear activation function to map and transform the features, and outputs a monitoring feature vector.

[0057] Then, for the waveform data of adjacent submodules, characteristic harmonic frequency points are extracted, the amplitude ratio between adjacent submodules is calculated, and the harmonic propagation attenuation sequence is obtained. The harmonic propagation attenuation sequence is segmented into time windows, the time displacement of adjacent windows is calculated, and the propagation speed characteristic sequence is obtained according to the physical distance between submodules. The AC voltage waveform and current waveform at both ends of the capacitor are collected, the voltage peak and valley values ​​are calculated to obtain the ripple coefficient, and the loss angle parameter is obtained from the voltage and current phase difference. The propagation speed characteristic sequence, ripple coefficient, and loss angle parameter are standardized using a normalization processor to generate a characteristic data matrix. A three-layer neural network structure is constructed, that is, a correlation model of the coupling relationship and monitoring accuracy is constructed. The input layer receives the characteristic data matrix, the hidden layer extracts the characteristic mapping relationship, and the output layer generates an error prediction sequence. The error prediction sequence is subjected to probability statistics, the mean and standard deviation parameters are calculated, and the error distribution function is generated. The cumulative probability value is calculated according to the error distribution function, and the monitoring error range is obtained by pre-setting the confidence threshold. The monitoring error range is compared and verified with the measured data, the neural network parameters are updated, and the feature extraction method is optimized.

[0058] The harmonic propagation characteristics between adjacent submodules reflect the state of capacitor parameters. By measuring the attenuation of characteristic harmonics between adjacent modules, the amplitude ratio of harmonic frequencies at 50 Hz, 100 Hz, and 150 Hz is extracted. Under normal working conditions, the harmonic propagation attenuation ratio is between 0.9 and 1.1. When the capacitor parameters deteriorate, the propagation attenuation intensifies and the ratio drops below 0.7. In the time window analysis, a 0.1 second data segment is selected to calculate the propagation time difference. The physical distance between adjacent submodules is 0.5 meters. The propagation speed is calculated by the time difference and distance. Under normal conditions, the propagation speed is about 800 meters per second. When the capacitor parameters deteriorate, the propagation speed drops to less than 500 meters per second, reflecting changes in harmonic propagation characteristics. In the measurement of capacitor AC characteristics, the peak value of the voltage waveform is 550 volts, the valley value is 450 volts, and the ripple coefficient calculation value is 0.1, reflecting the degree of voltage smoothness. The voltage and current phase difference is 15 degrees, and the loss tangent value is 0.27, which characterizes the loss characteristics of the capacitor.

[0059] The ripple coefficient and loss angle data are mapped to the interval of 0 to 1 through normalization processing to ensure the comparability of data of different dimensions. The neural network adopts a three-layer structure. The input layer contains 12 neurons, corresponding to the three types of characteristics of propagation speed, ripple coefficient, and loss angle. The hidden layer is set with 24 neurons. The hyperbolic tangent activation function is used to enhance the nonlinear expression ability. The output layer is set with 1 neuron to generate error prediction values. The network training uses 1000 sets of historical data, of which 800 sets are used for training and 200 sets are used for verification. Error statistical analysis shows that the mean of the predicted sequence is 0.03 and the standard deviation is 0.015. The normality test shows that the error basically conforms to the normal distribution. At the 95% confidence level, the monitoring error range is the mean plus or minus 2 times the standard deviation, that is, 0 to 0.06. The measured data verification shows that when the capacitor parameters are normal, the prediction error is within 0.02, and the confidence level reaches 98%. As the capacitor parameters deteriorate, the prediction error increases to 0.05 and the confidence level decreases to 92%. By comparing the verification results, the neural network structure was optimized, the number of hidden layer neurons was adjusted to 32, the weight parameters were updated, and the feature extraction method was improved. The average error of the optimized predictor on the test data was reduced to 0.02, the standard deviation was reduced to 0.01, and the 95% confidence interval was reduced to 0 to 0.04. In particular, the prediction performance under the condition of capacitor degradation was improved, providing a more accurate evaluation basis for capacitor status monitoring.

[0060] In one embodiment, the prediction neural network is trained by using sample fluctuation coupling information, including: Obtain sample fluctuation coupling information, input the sample fluctuation coupling information into the initial prediction neural network for prediction, and obtain the initial monitoring result; Calculate the mean square error of the initial monitoring results to obtain the monitoring error sequence, perform probability distribution fitting on the monitoring error sequence, and obtain the standard deviation and mean parameters; The monitoring confidence interval is calculated based on the standard deviation and mean parameters to obtain the monitoring accuracy assessment result; According to the monitoring accuracy evaluation results and the monitoring error sequence, the parameters of the initial prediction neural network are optimized, the feature extraction parameters and prediction parameters are updated, and the prediction neural network is obtained.

[0061] In this embodiment, the convolutional neural network is trained using the sample fluctuation coupling information, and the sample fluctuation coupling information is input into the initial prediction neural network for prediction to obtain the initial monitoring results. The mean square error calculator is used to calculate the error of the initial monitoring results to generate a monitoring error sequence. The monitoring error sequence is fitted with a probability distribution to calculate the standard deviation and mean parameters. The monitoring confidence interval is calculated based on the standard deviation and mean parameters to generate a monitoring accuracy evaluation result. The monitoring accuracy evaluation result is used to optimize the predictor parameters and update the feature extraction parameters and prediction parameters.

[0062] The error calculation adopts the mean square error criterion, and 1000 sets of test data are verified to calculate the error sequence between the predicted value and the true value. The mean of the error sequence is 0.05, and the standard deviation is 0.02. The normality test shows that the error basically conforms to the normal distribution. At the 95% confidence level, the confidence interval is the mean plus or minus 2 times the standard deviation, that is, 0.01 to 0.09. The monitoring accuracy evaluation results show that for data with coupling strength greater than 0.8, the prediction error is within 0.03, and the confidence level reaches 98%. For data with coupling strength between 0.5 and 0.8, the prediction error is within 0.05, and the confidence level is 95%. When the coupling strength is lower than 0.5, the prediction error increases to 0.08, and the confidence level decreases to 90%. According to the evaluation results, the predictor parameters are optimized, the number and size of convolution kernels are adjusted, the pooling strategy is updated, and the number of neurons in the fully connected layer is optimized. The average error of the optimized predictor on the test data was reduced from 0.05 to 0.03, the standard deviation was reduced from 0.02 to 0.015, and the 95% confidence interval was narrowed to 0.01 to 0.07. The improvement in monitoring accuracy is reflected in the reduction of prediction error and the convergence of confidence intervals, especially the improvement of prediction performance when the coupling strength is low.

[0063] In one embodiment, the voltage fluctuation data of the DC bus is analyzed according to the error prediction sequence and the confidence level to obtain monitoring results, including: Calculate the filter center frequency and bandwidth parameters based on the error prediction sequence and confidence level; The bus voltage fluctuation data is filtered by using the filter center frequency parameter and bandwidth parameter to obtain the filtered data, and the filtered data is extracted to obtain a multi-band fluctuation feature sequence; The multi-band fluctuation feature sequence is decomposed by wavelet multi-scale to obtain the harmonic feature sequence, and the support vector machine is used to classify the harmonic feature sequence to obtain the attenuation feature vector; Calculate the capacitance attenuation quantitative index according to the attenuation characteristic vector, normalize the capacitance attenuation index to obtain a correction result sequence, compare the correction result sequence with the original capacitance attenuation degree to obtain a correction deviation, update the filter parameters according to the correction deviation, and obtain an optimized non-intrusive monitoring model; The optimized non-intrusive monitoring model is used to analyze the voltage fluctuation data of the DC bus to obtain the monitoring results.

[0064] In this embodiment, according to the error rate and confidence level, the filter parameters and sampling frequency of the non-intrusive monitoring algorithm are adjusted, the extraction accuracy of the characteristic harmonic components is optimized, the DC bus voltage fluctuation data is re-analyzed, the corrected characteristic harmonic components are extracted, and it is determined whether the capacitance attenuation degree is consistent with the correction result. If the correction result is consistent with the capacitance attenuation degree, the final monitoring result is output. If not, the model parameters are readjusted and the monitoring algorithm is iteratively optimized until the correction result is consistent with the capacitance attenuation degree.

[0065] First, the center frequency and bandwidth parameters of the filter are calculated using the error prediction sequence and confidence level, and the passband fluctuation and stopband attenuation characteristics are extracted from the frequency response curve to generate the filter optimization parameter sequence. A multi-stage filter bank is constructed, and the passband characteristics of each level of the filter are set according to the optimized parameter sequence. The filter passband range is subdivided to generate a frequency-divided filter bank.

[0066] The bus voltage fluctuation data is preprocessed by using a frequency division filter group, and the voltage fluctuation characteristics are extracted from each frequency band to obtain a multi-band fluctuation characteristic sequence. The multi-band fluctuation characteristic sequence is subjected to wavelet multi-scale decomposition, and the wavelet basis function and the number of decomposition layers are set to extract the frequency characteristics from each scale coefficient. The amplitude threshold is set according to the frequency characteristics, the characteristic harmonic components are identified, and the harmonic characteristic sequence is generated. The support vector machine is used to classify the harmonic characteristic sequence, and the kernel function parameters and penalty factors are set to generate the attenuation characteristic vector.

[0067] The capacitance attenuation quantitative index is calculated according to the attenuation feature vector, and the capacitance attenuation index is normalized to generate a correction result sequence. The correction result sequence is compared and verified with the original attenuation degree, the correction deviation is calculated, and the filter optimization parameters are updated. The filter optimization design is based on the error rate and confidence index. When the error rate is 0.05 and the confidence is 95%, the center frequency is set at 500 Hz, the passband bandwidth is 200 Hz, and the stopband attenuation is greater than 40 decibels. The filter optimization parameters include four dimensions: center frequency, bandwidth, passband fluctuation, and stopband attenuation. The optimal parameter combination is determined by analyzing the frequency response curve. The multi-stage filter group adopts a 4-stage structure, and the frequency range is divided into four sub-bands: 0 to 250 Hz, 250 to 500 Hz, 500 to 750 Hz, and 750 to 1000 Hz. The passband fluctuation of each level of the filter is controlled within 1 decibel, and the stopband attenuation increases step by step with the frequency. The fine division of the frequency band is achieved through multi-stage filtering, and the frequency resolution of feature extraction is improved. In the extraction of voltage fluctuation features, the sampling frequency is set to 10 kHz, and the fluctuation data of each frequency band is subjected to wavelet decomposition. The db4 wavelet basis function is selected for 4-layer decomposition. Frequency features are extracted from each scale coefficient. In the frequency band of 0 to 250 Hz, the fundamental component and low-order harmonics are extracted; in the frequency band of 250 to 500 Hz, the intermediate frequency harmonic features are focused on; in the frequency band of 500 to 1000 Hz, the changes in high-frequency components are monitored. The amplitude threshold of harmonic feature identification is set to 3% of the fundamental amplitude to screen the characteristic harmonic frequency points. The fundamental frequency is 50 Hz, the second harmonic is 100 Hz, and the third harmonic is 150 Hz. The significance of each harmonic is judged by the amplitude ratio. The support vector machine uses the radial basis kernel function, the kernel parameter is set to 0.1, the penalty factor is 100, and the feature space mapping relationship is established. The calculation of the capacitance attenuation quantification index shows that the capacitance parameter attenuation is within 10% under normal working conditions, and the deviation between the correction result and the original judgment is within 3%. When the capacitance parameter deteriorates to 20%, the deviation of the correction result increases to 5%, reflecting that the accuracy of feature extraction decreases with the increase of attenuation. By comparing and verifying the update of filter parameters, adjusting the passband characteristics and stopband attenuation, the correction deviation after optimization is controlled within 2%. The filter optimization design runs through the entire monitoring process, from the optimization of frequency response characteristics to the design of multi-stage filter structure, and then to the improvement of feature extraction methods, forming a complete optimization chain. Through the feedback of error rate and confidence index, the filter parameters are continuously adjusted to improve the accuracy of feature extraction. In practical applications, the selection of filter parameters needs to balance the frequency resolution and computational complexity, to ensure the accuracy of feature extraction and meet the real-time requirements of online monitoring.

[0068] The difference sequence is calculated based on the correction result and the capacitance attenuation degree, and the difference sequence is normalized to generate the correction deviation index. The harmonic feature extraction parameters are calculated based on the correction deviation index, and the optimization direction is extracted from the parameter update curve to generate the parameter adjustment sequence. The update step size is set for the parameter adjustment sequence, and the neural network parameters are updated according to the step size to obtain the optimized network parameter group.

[0069] The voltage fluctuation characteristics are re-extracted using the optimized network parameters, and the characteristic harmonic frequency and amplitude sequence are calculated to generate a revised characteristic sequence. The capacitance attenuation degree is calculated based on the revised characteristic sequence, and the attenuation degree value is quantified to generate a revised attenuation index.

[0070] Calculate the deviation between the corrected attenuation index and the original attenuation degree, and determine whether the deviation is less than the preset threshold. If the deviation is less than the preset threshold, output the monitoring result; if the deviation is greater than the preset threshold, return to the parameter optimization link. Determine the convergence status based on the number of optimization iterations. When the number of iterations exceeds the preset value, stop the optimization and output the latest monitoring result.

[0071] The difference between the correction result and the capacitance attenuation degree is calculated using the relative error method. When the correction result is 15% and the actual attenuation degree is 12%, the relative error is 25%. All differences are mapped to the range of 0 to 1 through normalization processing, and the correction deviation index of 0.25 is obtained. The correction deviation index reflects the accuracy of the monitoring results. The smaller the index value, the more accurate the monitoring.

[0072] During the parameter optimization process, the neural network contains 12 neurons in the input layer, 24 neurons in the hidden layer, and 1 neuron in the output layer. The parameter adjustment adopts the adaptive learning rate method, and the initial learning rate is set to 0.01. When the correction deviation is greater than 0.2, the learning rate increases to 0.02 to accelerate parameter convergence; when the correction deviation is less than 0.1, the learning rate decreases to 0.005 to improve the optimization accuracy. The feature extraction parameters include three dimensions: filter bandwidth, sampling frequency, and window length. The bandwidth is adjusted from 200 Hz to 300 Hz, the sampling frequency is increased from 8 kHz to 12 kHz, and the window length is extended from 0.1 second to 0.15 second. The parameter adjustment direction is determined by the gradient of the correction deviation. The parameter update step is proportional to the deviation size. The larger the deviation, the larger the step. The optimized parameters are used to re-extract the voltage fluctuation characteristics, identify the characteristic harmonic components in the frequency band of 0 to 1000 Hz, and extract the fundamental 50 Hz and its harmonic components. The corrected characteristic harmonic amplitude shows that the fundamental amplitude is 10 volts, the second harmonic is 0.8 volts, and the third harmonic is 0.5 volts. The harmonic distribution more accurately reflects the state of the capacitor parameters. The quantification of the capacitor attenuation degree adopts a piecewise linear method. When the harmonic content is less than 5%, the attenuation degree is within 10%; when the harmonic content reaches 8%, the attenuation degree increases to 15%; when the harmonic content exceeds 10%, the attenuation degree exceeds 20%. Through the correspondence between harmonic characteristics and attenuation degree, an accurate state assessment standard is established. The optimization iteration process sets the maximum number of iterations to 100, and convergence is determined when the correction deviation changes by less than 0.01 for 10 consecutive iterations. In actual optimization, the convergence condition is generally achieved after 20 to 30 iterations, and the correction deviation is reduced from the initial 0.25 to below 0.05.

[0073] For data under different working conditions, there are differences in convergence speed, but the expected optimization goals can be achieved in the end. The optimization process of capacitor state monitoring reflects the ability of adaptive adjustment. The parameters are continuously optimized through feedback of correction deviations to improve monitoring accuracy. Parameter optimization, feature extraction, and state evaluation constitute a complete optimization chain to ensure the reliability of monitoring results. In the optimization process, it is necessary to balance the optimization accuracy and convergence speed to ensure that the monitoring results are accurate and meet the real-time requirements of online monitoring.

[0074] The non-intrusive monitoring system of the cascade module provided by the embodiment of the present invention is as follows: Figure 2 As shown, Figure 2 A block diagram of a non-intrusive monitoring system 200 of cascaded modules, comprising: An acquisition module 201 is used to acquire a voltage signal of a DC bus in a preset time window length, obtain a voltage fluctuation time domain signal, perform signal conversion on the voltage fluctuation time domain signal using a fast Fourier transform and extract the signal, obtain a plurality of main frequency components, and obtain a harmonic characteristic sequence according to each main frequency component; A first calculation module 202 is used to obtain a voltage waveform of a submodule according to a first preset sampling frequency, perform a time-frequency domain conversion on the voltage waveform using Fourier transform to obtain a frequency domain spectrum, perform characteristic frequency extraction according to the frequency domain spectrum to obtain a plurality of characteristic frequency points, extract a capacitance attenuation characteristic frequency point based on the amplitude corresponding to each characteristic frequency point and a capacitance attenuation characteristic library, and obtain an amplitude change trend sequence according to the amplitude corresponding to the capacitance attenuation characteristic frequency point; The second calculation module 203 is used to obtain a fitting curve of the coupling strength index and the capacitance attenuation degree based on the capacitance attenuation characteristic frequency points, the amplitude change trend sequence and the voltage fluctuation data of the adjacent sub-modules of the DC bus, obtain the correlation between the coupling strength index and the capacitance attenuation degree according to the fitting curve, and judge whether the coupling strength index is correlated with the capacitance attenuation degree based on the correlation. If so, calculate the error prediction sequence and confidence level of the monitoring process, analyze the voltage fluctuation data of the DC bus according to the error prediction sequence and the confidence level, and obtain the monitoring result, wherein the monitoring result is the capacitance attenuation degree.

[0075] The specific implementation of the non-invasive monitoring system for cascade modules is substantially the same as the specific implementation of the non-invasive monitoring method for cascade modules described above, and will not be described in detail herein.

[0076] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0077] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A non-intrusive monitoring method for cascade modules, characterized in that: include: Acquire a voltage signal of a DC bus in a preset time window length to obtain a voltage fluctuation time domain signal, perform signal conversion and extraction on the voltage fluctuation time domain signal using a fast Fourier transform to obtain a plurality of main frequency components, and obtain a harmonic characteristic sequence according to each of the main frequency components; Acquire a voltage waveform of the submodule according to a first preset sampling frequency, perform time-frequency domain conversion on the voltage waveform using Fourier transform to obtain a frequency domain spectrum, perform characteristic frequency extraction according to the frequency domain spectrum to obtain a plurality of characteristic frequency points, extract capacitance attenuation characteristic frequency points based on the amplitude corresponding to each of the characteristic frequency points and a capacitance attenuation characteristic library, and obtain an amplitude change trend sequence according to the amplitude corresponding to the capacitance attenuation characteristic frequency point; The capacitance attenuation characteristic frequency point, the amplitude change trend sequence and the voltage fluctuation data of the adjacent submodules of the DC bus are used to obtain a fitting curve of the coupling strength index and the capacitance attenuation degree. The correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve. Whether the coupling strength index is correlated with the capacitance attenuation degree is determined based on the correlation. If so, the error prediction sequence and the confidence level of the monitoring process are calculated. According to the error prediction sequence and the confidence level, the voltage fluctuation data of the DC bus is analyzed to obtain a monitoring result, wherein the monitoring result is the capacitance attenuation degree.

2. The non-intrusive monitoring method of cascade modules according to claim 1, characterized in that: The voltage fluctuation time domain signal is converted and extracted by fast Fourier transform to obtain a plurality of main frequency components, and a harmonic characteristic sequence is obtained according to each of the main frequency components, including: Acquire a voltage signal of the DC bus in a preset time window length to obtain a plurality of voltage data, calculate a fluctuation amplitude sequence according to a preset reference voltage value and each of the voltage data, and obtain a voltage fluctuation time domain signal; Performing segmented windowing processing on the voltage fluctuation time domain signal, smoothing the signal using a preset window function, and obtaining a smoothed voltage fluctuation sequence; Performing Fourier transform on the smoothed voltage fluctuation sequence to obtain a frequency domain representation sequence, performing spectrum line extraction on the frequency domain representation sequence according to a preset frequency domain analysis range to obtain a plurality of main frequency components, grouping the main frequency components according to their frequency values, and determining a fundamental frequency component and a plurality of subharmonic frequency components; The ratios of the amplitudes of the subharmonic frequency components to the fundamental frequency component are calculated to obtain a harmonic component content ratio sequence, and the subharmonic frequency components in the harmonic component content ratio sequence that are greater than a preset harmonic limit are determined to be out-of-limit harmonic components to obtain a harmonic characteristic sequence.

3. The non-intrusive monitoring method of cascade modules according to claim 1, characterized in that: The characteristic frequency is extracted according to the frequency domain spectrum to obtain a plurality of characteristic frequency points, and the capacitance attenuation characteristic frequency points are extracted based on the amplitude corresponding to each of the characteristic frequency points and the capacitance attenuation characteristic library, and the corresponding amplitude change trend sequence is obtained according to the capacitance attenuation characteristic frequency points, including: Extract the frequency domain spectrum lines to obtain a harmonic frequency sequence and a harmonic amplitude sequence; Classifying the harmonic frequency sequence according to a preset frequency interval to obtain a plurality of harmonic frequency interval sequences, calculating the amplitude proportion of the harmonic frequency components in each of the harmonic frequency interval sequences to obtain a harmonic frequency distribution characteristic sequence, and selecting a frequency point whose amplitude proportion is greater than a preset proportion in the harmonic frequency distribution characteristic sequence as a characteristic frequency point to obtain a plurality of characteristic frequency points; Based on the amplitude corresponding to each of the characteristic frequency points, a fixed-length sliding window is used to calculate the amplitude change rate to obtain an amplitude change rate sequence, and an amplitude in the amplitude change rate sequence that is greater than a preset change rate threshold is used as an amplitude mutation point; Calculate the spectrum similarity of the amplitude mutation point within a preset range, judge the similarity threshold to obtain the spectrum change type, match the spectrum change type with the capacitance attenuation feature library, obtain the spectrum correlation sequence through matching calculation, extract the capacitance attenuation characteristic frequency point according to the spectrum correlation sequence, and record the amplitude change trend sequence corresponding to the capacitance attenuation characteristic frequency point.

4. The non-intrusive monitoring method of cascade modules according to claim 1, characterized in that: The method of obtaining a fitting curve between a coupling strength index and capacitance attenuation data by using the capacitance attenuation characteristic frequency point, the amplitude change trend sequence and the voltage fluctuation data of the adjacent submodules of the DC bus, and obtaining the correlation between the coupling strength index and the capacitance attenuation degree according to the fitting curve includes: Normalizing the capacitance attenuation characteristic frequency points and the amplitude change trend sequence to obtain a normalized characteristic matrix, and extracting features from the normalized characteristic matrix using a principal component dimensionality reduction method to obtain a reduced dimensionality characteristic vector; Inputting the dimension-reduced feature vector into a trained vector machine model for regression prediction to obtain capacitance attenuation degree data, wherein the capacitance attenuation degree data includes a capacitance loss prediction value and a remaining life prediction value; A fitting curve between a coupling strength index and the capacitance attenuation data is obtained according to the voltage fluctuation data of adjacent submodules of the DC bus and the capacitance attenuation degree data, and a correlation between the coupling strength index and the capacitance attenuation degree is obtained according to the fitting curve.

5. The non-intrusive monitoring method of cascade modules according to claim 4, characterized in that: The method of obtaining a fitting curve between a coupling strength index and capacitance attenuation data according to the voltage fluctuation data of adjacent submodules of the DC bus and the capacitance attenuation degree data, and obtaining a correlation between the coupling strength index and the capacitance attenuation degree according to the fitting curve, includes: Collecting voltage fluctuation data of adjacent sub-modules of the DC bus, aligning the voltage fluctuation data using a synchronous sampling clock to obtain aligned fluctuation data, and collecting the aligned fluctuation data according to a second preset sampling frequency to obtain complete cycle waveform data; The complete cycle waveform data is segmented using a fixed-length sliding window to obtain segmented data, the segmented data is Fourier transformed to obtain a frequency domain representation sequence, a plurality of characteristic harmonic frequency points are extracted from the frequency domain representation sequence, and the frequency offset between adjacent submodules is calculated according to each of the characteristic harmonic frequency points to obtain a frequency deviation sequence, and the characteristic harmonic frequency point in the frequency deviation sequence whose frequency offset is greater than a preset frequency threshold is taken as an abnormal frequency point; According to the abnormal frequency point, the characteristic harmonic amplitude ratio of adjacent submodules is calculated, the waveform correlation coefficient is calculated using adjacent window data to obtain a coupling characteristic sequence, the average value of the waveform correlation coefficient sequence is calculated to obtain a coupling strength index, a polynomial fitting is performed on the coupling strength index and the capacitance attenuation degree data to obtain a fitting curve, the fitting curve is extracted to obtain the correlation between the coupling strength index and the capacitance attenuation degree.

6. The non-intrusive monitoring method of cascade modules according to claim 1, characterized in that: The error prediction sequence and confidence level of the computing monitoring process include: The prediction neural network is used to extract the characteristics of the wave coupling information and obtain the monitoring characteristic vector; The monitoring feature vector is input into the constructed correlation model of coupling relationship and monitoring accuracy for processing, so as to obtain the error prediction sequence and confidence level of the monitoring process.

7. The non-intrusive monitoring method of cascade modules according to claim 6, characterized in that: The method of extracting features from the fluctuation coupling information using a prediction neural network to obtain a monitoring feature vector includes: Normalizing the wave coupling information to obtain a standardized feature matrix; A processor is used to perform convolution operation on the standardized feature matrix to obtain a coupling feature map; A second layer of convolution operation is performed on the coupling feature map to obtain a propagation feature map, a pooling operation is used to reduce the dimension of the propagation feature map to obtain a reduced-dimensional feature, and a fully connected layer is used to map and transform the reduced-dimensional feature to obtain a monitoring feature vector.

8. The non-intrusive monitoring method of cascade modules according to claim 6, characterized in that: The prediction neural network is trained by using sample fluctuation coupling information, including: Obtain sample fluctuation coupling information, input the sample fluctuation coupling information into the initial prediction neural network for prediction, and obtain the initial monitoring result; Calculating the mean square error of the initial monitoring result to obtain a monitoring error sequence, performing probability distribution fitting on the monitoring error sequence to obtain standard deviation and mean parameters; Calculate the monitoring confidence interval according to the standard deviation and the mean parameter to obtain a monitoring accuracy evaluation result; The parameters of the initial prediction neural network are optimized according to the monitoring accuracy evaluation result and the monitoring error sequence, and the feature extraction parameters and the prediction parameters are updated to obtain the prediction neural network.

9. The non-intrusive monitoring method of cascade modules according to claim 1, characterized in that: The step of analyzing the voltage fluctuation data of the DC bus according to the error prediction sequence and the confidence level to obtain a monitoring result includes: Calculating filter center frequency and bandwidth parameters according to the error prediction sequence and the confidence level; The bus voltage fluctuation data is filtered using the filter center frequency parameter and the bandwidth parameter to obtain filtered data, and the filtered data is extracted to obtain a multi-band fluctuation feature sequence; Performing wavelet multi-scale decomposition on the multi-band fluctuation characteristic sequence to obtain a harmonic characteristic sequence, and using a support vector machine to perform classification operation on the harmonic characteristic sequence to obtain an attenuation characteristic vector; Calculating a capacitance attenuation quantification index according to the attenuation characteristic vector, normalizing the capacitance attenuation index to obtain a correction result sequence, comparing the correction result sequence with the original capacitance attenuation degree to obtain a correction deviation, updating the filter parameters according to the correction deviation, and obtaining an optimized non-intrusive monitoring model; The optimized non-intrusive monitoring model is used to analyze the voltage fluctuation data of the DC bus to obtain a monitoring result.

10. A cascade module non-intrusive monitoring system, characterized in that: include: An acquisition module is used to acquire a voltage signal of a DC bus in a preset time window length, obtain a voltage fluctuation time domain signal, perform signal conversion and extraction on the voltage fluctuation time domain signal using a fast Fourier transform, obtain a plurality of main frequency components, and obtain a harmonic characteristic sequence according to each of the main frequency components; A first calculation module is used to obtain a voltage waveform of a submodule according to a first preset sampling frequency, perform a time-frequency domain conversion on the voltage waveform using Fourier transform to obtain a frequency domain spectrum, perform characteristic frequency extraction according to the frequency domain spectrum to obtain a plurality of characteristic frequency points, extract a capacitance attenuation characteristic frequency point based on the amplitude corresponding to each of the characteristic frequency points and a capacitance attenuation characteristic library, and obtain an amplitude change trend sequence according to the amplitude corresponding to the capacitance attenuation characteristic frequency point; The second calculation module is used to obtain a fitting curve of a coupling strength index and a capacitance attenuation degree by using the capacitance attenuation characteristic frequency point, the amplitude change trend sequence and the voltage fluctuation data of adjacent sub-modules of the DC bus, obtain the correlation between the coupling strength index and the capacitance attenuation degree according to the fitting curve, judge whether the coupling strength index is correlated with the capacitance attenuation degree based on the correlation, and if so, calculate the error prediction sequence and the confidence level of the monitoring process, analyze the voltage fluctuation data of the DC bus according to the error prediction sequence and the confidence level, and obtain a monitoring result, wherein the monitoring result is the capacitance attenuation degree.

Images