Electric meter load monitoring method and system based on electric energy quality analyzer

Abstract

The invention relates to the technical field of electrical detection, in particular to an electricity meter load monitoring method and system based on an electric energy quality analyzer, and the method comprises the following steps: extracting three-phase voltage phase deviation characteristics in a disturbance period, combining with the comparison of a frequency band change trend and fundamental frequency power, recognizing an interference coupling section, and analyzing a jump frequency point; and outputting an ammeter load monitoring result. According to the method, the phase deviation angle of the three-phase voltage is extracted in the voltage disturbance period, and the cross-phase synchronization index is constructed, so that the synchronous response characteristics of different loads to disturbance can be accurately revealed, and the accuracy of load response attribution judgment is enhanced. On the basis of cross-phase synchronous analysis, a corresponding relation between frequency band characteristics and response attribution is constructed, so that load response subdivision has a phase difference identification capability and a frequency behavior calibration capability, and the discrimination robustness under a multi-source interference background is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical detection, and particularly to a method and system for monitoring the load of an electric meter based on a power quality analyzer. Background Art

[0002] The technical field of electrical detection includes the monitoring of the operating states of various electrical equipment and systems, performance evaluation, and safety diagnosis. The core content of this technical field is to identify abnormal equipment operation, system load changes, and potential power fault hazards through the measurement and analysis of key electrical parameters such as voltage, current, frequency, power factor, and power quality.

[0003] Among them, the method for monitoring the load of an electric meter based on a power quality analyzer refers to using a power quality analyzer to collect various electrical parameter data such as voltage, current, harmonics, and voltage deviation of the corresponding circuit of the measured electric meter, and combining the time-periodic characteristics and power characteristics of the electrical load to classify and identify the actual load conditions connected to the electric meter and monitor the changes.

[0004] The load monitoring methods based on traditional power quality analyzers mostly rely on the direct sampling of static electrical parameters such as voltage, current, and harmonics, lacking in-depth characterization of the dynamic response behavior within the disturbance period, and it is difficult to accurately judge the phase response differences of different loads under the action of disturbances. The lack of a quantitative analysis mechanism for the three-phase phase shift path leads to the problem of response attribution confusion in the case of mixed access of multiple loads or superposition of nonlinear interference, making it difficult to clearly analyze different load behaviors. During the load identification process, a single-band or full-band unified analysis strategy is mostly adopted, without differential processing in combination with the frequency-band behavior characteristics, which is easily masked by frequency disturbances and reduces the identification accuracy. The lack of an interference identification mechanism based on the linkage characteristics of periodic power and frequency-band amplitude causes problems such as inaccurate frequency point positioning and high response misjudgment rate in a strong interference background. For example, although the amplitude of some high-frequency harmonic disturbances is small, they are highly synchronized with the fundamental frequency power change. Without matching analysis, they are easily overlooked, resulting in blind spots in the monitoring of abnormal load behaviors. In addition, in the prior art, no clustering analysis is carried out on the jump frequency points, making it difficult to identify representative frequency characteristic points, which cannot be effectively used for subsequent frequency point control and load model optimization. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose a method and system for monitoring the load of an electric meter based on a power quality analyzer.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A method for monitoring the load of an electric meter based on a power quality analyzer includes the following steps: S1: Obtain the three-phase voltage of the electric meter through a power quality analyzer, extract the phase shift angles of the three-phase voltage during the disturbance period to judge the synchronization characteristics between phases, and form a cross-phase synchronization evaluation result; S2: Based on the cross-phase synchronization evaluation result, identify the load response attribution types of the phase shift angles of the three-phase voltage, establish a load response dissection sample set, obtain the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, and establish a frequency band correspondence relationship; S3: Refer to the frequency band correspondence relationship of the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, judge the change trend of the harmonic amplitude within the frequency band, and obtain the frequency band amplitude change data; S4: Obtain the change trend of the fundamental frequency total power of the electric meter load monitored by the power quality analyzer during the disturbance period, compare the change trend of the harmonic amplitude at the corresponding moment in the frequency band amplitude change data with the change trend of the fundamental frequency total power, and identify the interference coupling fluctuation section; S5: Extract the frequency points with prominent jump amplitudes and jump frequencies in the interference coupling fluctuation section, perform active frequency point analysis on the frequency points, and output the electric meter load monitoring result.

[0007] As a further solution of the present invention, the cross-phase synchronization evaluation result includes a phase shift direction feature, an inter-phase response consistency level, and a synchronous stability change condition. The load response dissection sample set includes a load combination structure label, a response mode classification result, and a record of phase synchronization performance. The frequency band amplitude change data is specifically a jump amplitude trajectory within the frequency band, a periodic stability characteristic index, and a frequency fluctuation distribution state. The interference coupling fluctuation section specifically refers to a harmonic disturbance superposition interval, a power change synchronization region, and a frequency band behavior resonance node. The electric meter load monitoring result includes a load type identification result, an interference frequency identification label, and load structure response information.

[0008] As a further solution of the present invention, the specific steps for obtaining the cross-phase synchronization evaluation result are as follows: S111: Obtain the A, B, and C three-phase voltage waveform data recorded by the power quality analyzer within a specified time period, and extract the voltage phase shift angle change sequences during three cycles before, during, and after the disturbance; S112: Based on the voltage phase shift angle change sequences, use the cosine similarity algorithm to calculate the cosine values of the angles between the A-B, B-C, and A-C three pairs of angle change vectors in the sequences respectively, and judge the inter-phase synchronization matching degree at each sampling point within the corresponding cycle according to the cosine values of the angles, and obtain the phase shift path similarity analysis result; S113: Judge the average similarity change trend of the A-B, B-C, and A-C three pairs of angle change vector continuous three-cycle sections in the phase shift path similarity analysis result, and form a cross-phase synchronization evaluation result.

[0009] As a further solution of the present invention, the steps for obtaining the corresponding relationship of frequency bands are specifically as follows: S211: According to the phase - to - phase response patterns of the load to disturbances judged by the three pairs of angle change vectors A - B, B - C, and A - C in the cross - phase synchronization evaluation result, if it shows a similar coincidence state for three consecutive cycles, it is classified as a concentrated - type load response sample; if there are phased or amplitude - based response differences between different phases, it is classified as a parallel - type load response sample; S212: Integrate the concentrated - type load response samples and the parallel - type load response samples, and perform structural and time index labeling on the samples to form a load response dissection sample set; S213: By invoking the phase calibration interface and frequency component measurement function of the power quality analyzer, obtain the frequency band characteristics corresponding to the three - phase voltage in the load response dissection sample set, and establish the corresponding relationship between the three - phase voltage and the frequency band characteristics.

[0010] As a further solution of the present invention, the steps for obtaining the frequency band amplitude change data are specifically as follows: S311: Obtain the harmonic voltage amplitudes within the 2nd, 3rd, and 6th frequency bands recorded in the power quality analyzer, and extract the harmonic amplitude sequences whose frequency band information is consistent with that marked in the load response dissection sample set and the corresponding relationship of frequency bands; S312: Extract the amplitude change paths of each cycle before, during, and after the disturbance in the harmonic amplitude sequences with consistent frequency band information, analyze the number of jumps and amplitude fluctuation characteristics, and statistically analyze the change trends under consecutive cycles, and output the frequency band amplitude change data.

[0011] As a further solution of the present invention, the steps for obtaining the interference coupling fluctuation section are specifically as follows: S411: Obtain the change trend of the fundamental frequency total power of the meter load monitored by the power quality analyzer during the disturbance period, establish power sequence data by cycle to obtain the fundamental frequency total power change sequence; S412: Extract the change trend of the harmonic amplitude at the corresponding moment according to the frequency band amplitude change data, and perform cycle - by - cycle pairing with the fundamental frequency total power change sequence to generate an amplitude - power ratio sequence; S413: Identify the cycle segments in the amplitude - power ratio sequence where the jump trends of the two types of trends are consistent during the disturbance period through the cumulative sum test algorithm, and judge the interference coupling fluctuation section of synchronous fluctuation.

[0012] As a further solution of the present invention, the steps for obtaining the monitoring result of the meter load are specifically as follows: S511: Extract the frequency points with jumps in each period of the interference coupling fluctuation section, and count the corresponding jump amplitudes and jump frequencies to obtain a jump frequency point index set; S512: According to the jump frequency point index set, screen the frequency point set with prominent jump amplitudes and frequencies, and construct a candidate active frequency point sequence; S513: Use the density clustering algorithm to perform clustering analysis on the candidate active frequency point sequence, identify the activation frequency points exceeding the preset frequency threshold, and output the electricity meter load monitoring result for amplitude limiting processing of interference frequency points and load identification optimization.

[0013] An electricity meter load monitoring system based on a power quality analyzer, which is used to execute the above-mentioned electricity meter load monitoring method based on a power quality analyzer. The system includes: The cross-phase synchronization recognition module obtains the three-phase voltages of the electricity meter through the power quality analyzer, extracts the phase offset angles of the three-phase voltages in the disturbance period to judge the synchronization characteristics between phases, and forms a cross-phase synchronization evaluation result; The load response attribution module, based on the cross-phase synchronization evaluation result, identifies the load response attribution types of the phase offset angles of the three-phase voltages, establishes a load response dissection sample set, and obtains the frequency band corresponding relationship of the three-phase voltages in the load response dissection sample set; The frequency band amplitude analysis module refers to the frequency band corresponding relationship of the frequency band characteristics corresponding to the three-phase voltages in the load response dissection sample set, judges the change trend of the harmonic amplitudes in the frequency band, and obtains the frequency band amplitude change data; The power trend comparison module obtains the change trend of the fundamental frequency total power of the electricity meter load monitored by the power quality analyzer in the disturbance period, compares the change trend of the harmonic amplitudes at the corresponding moments in the frequency band amplitude change data with the change trend of the fundamental frequency total power, and identifies the interference coupling fluctuation section; The active frequency point recognition module extracts the frequency points with prominent jump amplitudes and frequencies in the interference coupling fluctuation section, performs active frequency point analysis on the frequency points, and outputs the electricity meter load monitoring result.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, by extracting the phase shift angles of the three-phase voltages within the voltage disturbance period and constructing a cross-phase synchrony index, the synchronous response characteristics of different loads to disturbances can be accurately revealed, enhancing the accuracy of load response attribution judgment. Based on the cross-phase synchronous analysis, the corresponding relationship between the frequency band characteristics and response attribution is constructed, enabling the load response dissection to have the ability to identify phase differences and calibrate frequency behaviors, thereby improving the discrimination robustness under the background of multi-source interference. By combining microscopic disturbance characteristics such as hopping frequency and amplitude with the periodic power change trend for multi-dimensional data comparison, the coupling relationship between the frequency band amplitude fluctuation and load change can be effectively mined, enhancing the ability to identify key frequency points in complex disturbance events. Further, by using density clustering to mine the active feature points in the hopping frequencies, it helps to accurately distinguish the noise frequency points from the effective load frequency points, providing a more targeted frequency basis for subsequent load identification and interference mitigation. Throughout the process, by means of the dynamic matching analysis of harmonic amplitude, phase shift, and fundamental frequency power, not only can the monitoring sensitivity to abnormal load events be improved, but also the load identification dimension can be extended to frequency band behaviors and synchrony performances, realizing the deep perception of the load structure state and comprehensively enhancing the stability, accuracy, and interpretability of the monitoring results. Description of the Drawings

[0015] Figure 1 is a schematic diagram of the working process of the present invention; Figure 2 is a flowchart of step S1 of the present invention; Figure 3 is a flowchart of step S2 of the present invention; Figure 4 is a flowchart of step S3 of the present invention; Figure 5 is a flowchart of step S4 of the present invention; Figure 6 is a flowchart of step S5 of the present invention. Detailed Embodiments

[0016] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0017] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by terms such as "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more, unless otherwise specifically defined.

[0018] Please refer to Figure 1 , the present invention provides a technical solution: a method for monitoring the load of an electric meter based on a power quality analyzer, including the following steps: S1: Obtain the three-phase voltage of the electric meter through the power quality analyzer, extract the phase shift angle of the three-phase voltage during the disturbance period to judge the synchronous characteristics between phases, and form a cross-phase synchronization evaluation result; S2: Based on the cross-phase synchronization evaluation result, identify the load response attribution type of the phase shift angle of the three-phase voltage, establish a load response dissection sample set, and obtain the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set to establish a frequency band correspondence relationship; S3: Refer to the frequency band correspondence relationship of the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, judge the change trend of the harmonic amplitude within the frequency band, and obtain the frequency band amplitude change data; S4: Obtain the change trend of the fundamental frequency total power of the electric meter load monitored by the power quality analyzer during the disturbance period, compare the change trend of the harmonic amplitude at the corresponding moment in the frequency band amplitude change data with the change trend of the fundamental frequency total power, and identify the interference coupling fluctuation section; S5: Extract the frequency points with prominent jump amplitudes and jump frequencies in the interference coupling fluctuation section, perform active frequency point analysis on the frequency points, and output the monitoring result of the electric meter load; The cross-phase synchronization evaluation result includes the phase shift direction feature, the inter-phase response consistency level, and the change situation of the synchronization stability. The load response dissection sample set includes the load combination structure label, the response mode classification result, and the record of the phase synchronization performance. The frequency band amplitude change data is specifically the jump amplitude trajectory within the frequency band, the cycle stability characteristic index, and the frequency fluctuation distribution state. The interference coupling fluctuation section specifically refers to the harmonic disturbance superposition interval, the power change synchronization region, and the frequency band behavior resonance node. The monitoring result of the electric meter load includes the load type identification result, the interference frequency identification label, and the load structure response information.

[0019] Please refer to Figure 2 , the specific steps for obtaining the cross-phase synchronization evaluation result are as follows: S111: Obtain the A, B, and C three-phase voltage waveform data recorded by the power quality analyzer within the specified time period, and extract the voltage phase offset angle change sequence within three cycles before, during, and after the disturbance; Obtain the A, B, and C three-phase voltage waveform data recorded by the power quality analyzer. First, set the data sampling rate to 10 kHz, perform continuous data acquisition during the period from 09:32:00 to 09:32:02 on July 5, 2024. Independently measure the A, B, and C three-phase voltage channels, record the sampled data and save it as a time series array. After obtaining the voltage data of each phase, extract the steady-state waveform data within 5 cycles before the disturbance through the full-cycle truncation method, identify the first abnormal offset point as the disturbance starting point, and extract a waveform sequence with a period length before and after this point as the data before and after the disturbance. When extracting the data during the disturbance, locate it at the center of the disturbance occurrence time point, symmetrically extract a voltage waveform with a period length, perform a fast Fourier transform (FFT) on the data sequence within each cycle, convert the time-domain waveform to the frequency-domain complex form, and then extract the main frequency component phase angle of each phase voltage at the fundamental frequency in complex form. Using the phase angle offset calculation method, assume the initial phase angle of phase A is , during the disturbance is , after the disturbance is , then the three-phase angle offset sequences are , phases B and C are processed in the same way to form three groups of time-angle sequence data. After obtaining the angle change sequences of each phase, organize them into the following array representation. Phase A: , phase B: , phase C: .

[0020] S112: Based on the voltage phase offset angle change sequence, use the cosine similarity algorithm to calculate the cosine values of the included angles between the three pairs of angle change vectors A-B, B-C, and A-C in the sequence respectively, and judge the inter-phase synchronous matching degree at each sampling point within the corresponding cycle according to the cosine values of the included angles to obtain the phase offset path similarity analysis result; Based on the extracted phase offset angle change sequence, respectively construct the phase angle change vector differences between each pair of phases, defined as: A-B vector difference: , B-C vector difference: , A-C vector difference: , for the above three groups of angle difference vectors, calculate the cosine similarity pairwise, and the steps are as follows: Regard each pair of difference vectors as two three-dimensional vectors , , then the calculation formula for the cosine value of the included angle is: ; Among them, : represents the cosine value of the angle between vectors u and v, reflecting the similarity degree of two angle change vectors; : represents the angle between two three-dimensional vectors u and v, with the unit of radian (but in practical applications, the cosine value is directly analyzed); : represents the first group of angle change difference vectors, and the elements , , are the angle difference components of A - B, B - C, or A - C in three sampling periods respectively; : represents another group of angle change difference vectors corresponding to u, and the elements , , are another group of angle difference components in the same position period; : is the dot product of vectors u and v, which is essentially the sum of item-by-item products and represents the coincidence degree of the direction projections of two vectors; : represents the Euclidean norm of vector u, reflecting the overall "strength" of the vector; : Similarly, it is the norm of vector v; The overall fractional structure: is the standard expression for calculating the cosine of the angle between two vectors, and the result range is , where 1 means the directions are completely the same, 0 means perpendicular, and -1 means the directions are completely opposite.

[0021] For example, to calculate the cosine value of the angle between A - B and A - C: Let , v = [0, 2, 1] Substitute into the calculation: Dot product: , Norm: , , Cosine value: . Similarly, to calculate the cosine value of the angle between B - C and A - B: Let ; Dot product: ; Norm: ; Cosine value: .

[0022] The judgment principle is set as follows: If the cosine value , it is judged as "highly synchronous"; if the cosine value is between 0.7 and 0.95, it is judged as "moderately synchronous"; if the cosine value is lower than 0.7, it is judged as "asynchronous". This interval is set according to the tolerance of the symmetrical structure voltage deviation, and the relevant reference standard is that the critical angle of the phase angle synchronization deviation specified in IEC61000 - 4 - 30 is approximately equal to 18°, and the cosine value critical point is The advantage of the formula is that by calculating the cosine angle between the angle difference vectors before and after the three-phase voltage disturbance, the quantitative judgment of the instantaneous phase synchronization degree between the three phases is realized.

[0023] S113: Judge the average similarity change trend of the three pairs of angle change vectors in the three consecutive cycle sections of A-B, B-C, and A-C in the phase shift path similarity analysis result to form the cross-phase synchronization evaluation result; Based on the cosine values of the included angles of the three pairs of angle change vectors, with three consecutive cycles as an analysis window, extract the cosine values within each cycle in sequence, and statistically calculate the average similarity between each pair. Suppose the cosine values of three cycles of A-B are 0.988, 0.993, 0.979, B-C are -0.951, -0.980, -0.925, and A-C are 0.961, 0.974, 0.953. The following steps are executed: First, calculate the average value of three cycles for each pair: The average value of A-B: , The average value of B-C: (-0.951 - 0.980 - 0.925) / , The average value of A-C: . Perform trend judgment on the above results. Adopt the sliding window method to perform differential processing on the average value of each group within three cycles to judge whether it is in an upward, downward or stable state. The judgment criteria are as follows: If the difference between the latter value and the former value , it is judged as "stable"; if the latter value increases compared with the former value , it is judged as "upward"; if the latter value decreases compared with the former value , it is judged as "downward". Record the differential results in sequence, and set that if there are two "downward" occurrences within three cycles in the judgment window, it is judged as "synchronization deterioration", two "upward" as "synchronization recovery", and three "stable" as "synchronization maintenance". In this embodiment, A-B is "synchronization maintenance", B-C is "synchronization deterioration", and A-C is "synchronization maintenance". This result constitutes the cross-phase synchronization evaluation result.

[0024] Please refer to Figure 3 , The steps for obtaining the frequency band correspondence relationship are specifically as follows: S211: Judge the inter-phase response mode of the load to the disturbance according to the three pairs of angle change vectors of A-B, B-C, and A-C in the cross-phase synchronization evaluation result. If it shows an approximately coincident state in three consecutive cycles, it is classified as a concentrated load response sample. If there are phased or amplitude response differences between different phases, it is classified as a parallel load response sample; Based on the A-B, B-C, and A-C pairs of angle change vectors in the cross-phase synchronization evaluation results to judge the inter-phase response mode of the load to disturbances, it is first necessary to extract the cosine value sequences of the angles between each pair and perform classification processing. The values of the three groups of cosine values of A-B, B-C, and A-C in three consecutive periods are set as , and where A-B, B-C, and A-C respectively represent the phase angle change difference vector combinations between phase A and phase B, phase B and phase C, and phase A and phase C. Each group of cosine values is the inter-phase similarity index obtained by calculating the previous vector included angle, denoted as , , , where , , represents the included angle between each group of vectors, in radians. The range of the cosine value of the included angle is . The difference of each group of three-period cosine values is judged. It is set that the difference less than 0.01 is the approximate coincidence standard. According to the standard of IEC61000-4-30, the phase angle offset of voltage fluctuation under the condition of three-phase symmetry should not exceed 18°. The corresponding cosine critical value is 0.951. Therefore, it is set that is the approximate state. It is judged whether each pair of phases all meet this standard. If the three groups of angle differences are all kept within this fluctuation range, the three-phase voltage fluctuation trends in this period segment are consistent, denoted as the "approximate coincidence state", and the corresponding samples are classified as concentrated load response samples. In the above example values, the periodic changes of the three groups of data are all lower than 0.004, meeting the approximate coincidence standard. If it is replaced by another group of samples, where the three-period cosine value of A-B is , and its fluctuation amplitude reaches 0.088, far exceeding the coincidence threshold, indicating that there are amplitude response differences between different phases of the load. This type of sample is denoted as a parallel load response sample. The difference threshold is based on the multi-phase response identification specification shown in IEEE Std1159. It is necessary that at least one pair of cosine fluctuations exceeds 0.03 to be defined as a parallel type. The judgment step needs to calculate the difference between the maximum and minimum cosine values of any group of the three groups of angle differences in three periods one by one. If it is greater than the threshold, it enters the parallel response sample set, marks the sample and saves the corresponding response category.

[0025] S212: Integrate the concentrated load response samples and the parallel load response samples, and perform structure and time index annotation on the samples to form a load response dissection sample set; After integrating the centralized load response samples and the parallel load response samples, all the samples are combined to generate a unified sample index structure. Each set of sample data contains three information fields: phase offset timestamp, sample type identifier (0 represents centralized type, 1 represents parallel type), and sample number (uniquely represented by YYYYMMDDhhmmss + number sequence). During the execution process, the sample number format is fixed as a 17 - digit decimal number. The first 14 digits are the acquisition timestamp, and the last 3 digits are the in - day sequence number. For example, the sample number "202407050932001" means that the sample was collected at 09:32:00 on July 5, 2024, and it is the first sampling segment of the day. Subsequently, the sample information is saved in a structured manner, and at the same time, a "time index field" and a "response category field" are added to each sample node. The file structure is defined in JSON format, and the field contents are as follows: "timestamp" corresponds to the start time of the sample, "response_type" is 0 or 1, and "sequence_id" is a numbered string. When loading the data, the response category is quickly filtered through the index field. The sample structure also needs to record three sets of angle vector values corresponding to the sample. The data format is represented in an array - nested manner as {"vectors":{"A - B":[0.985,0.989,0.987],"B - C":[0.981,0.982,0.983],"A - C":[0.986,0.987,0.985]}}, where "A - B", "B - C", and "A - C" are still the three phase - to - phase combination relationships, and the three values in each vector are the cosine values between the corresponding phases in each cycle, corresponding to , , , where i = 1, 2, 3 represents three consecutive cycles. This number is used to implement a reproducible multi - cycle tracking logic in the sample structure. The sample structure is finally written into the sampling record document in dictionary format, and "response_type" is set as the main retrieval field at the database end.

[0026] S213: By invoking the phase calibration interface and frequency component measurement function of the power quality analyzer, obtain the frequency band characteristics corresponding to the three - phase voltages in the load response dissection sample set, and establish the frequency band correspondence relationship between the three - phase voltages and the frequency band characteristics; When invoking the phase calibration interface and frequency component calculation function of the power quality analyzer, it is necessary to first set the start and end time points of the sampled data, perform FFT processing on the three-phase voltage signals to extract the frequency components. During the processing, the analysis bandwidth is set to 0–2 kHz, the sampling frequency is set to 10 kHz, each group of voltage signals is processed with a window function in units of 1000 points, the corresponding window width is 0.1 second, a Hanning window is applied to each phase voltage to improve the frequency domain resolution. After performing the spectrum calculation, 20 frequency band windows are divided within the 0–2 kHz frequency band for each phase voltage, each segment width is 100 Hz. The square of the amplitude within each frequency band is integrated to obtain its energy value and normalized to represent the relative frequency band energy density. Let the frequency band number be j, then the relative energy density of the j-th frequency band is defined as , , , where , , are the energy integral values of the A, B, and C phase voltages in the j-th frequency band respectively, with the unit of volt squared second (V²·s), and the vector forms are respectively , , , where each element , , represents the normalized frequency band energy density of this phase in the K-th frequency band, which is used to characterize the frequency domain response characteristics. If for any frequency band there is or , then this frequency band is recorded as the "phase difference frequency band", and with the frequency band number K as the primary key, a frequency band difference record table is established in the sample index structure to associate the three-phase frequency domain mapping characteristics.

[0027] Please refer to Figure 4 . The specific steps for obtaining the frequency band amplitude change data are as follows: S311: Obtain the harmonic voltage amplitudes within the 2nd, 3rd, and 6th frequency bands recorded in the power quality analyzer, and extract the harmonic amplitude sequence whose frequency band information is consistent with the marked frequency band in the load response dissection sample set and frequency band correspondence; To obtain the harmonic voltage amplitudes within the 2nd, 3rd, and 6th frequency bands recorded in the power quality analyzer, first call the spectrum analysis module of the analyzer to extract the harmonic component amplitude data it records. Assuming the fundamental frequency of the system is 50 Hz, then the 2nd, 3rd, and 6th harmonics correspond to frequencies of 100 Hz, 150 Hz, and 300 Hz respectively. During the data reading process, assume the sampling period is 0.02 second and the sampling rate is 10 kHz. Each frequency band is processed using the fast Fourier transform (FFT) method to extract the voltage amplitude at the specified frequency points, which are respectively denoted as , , ,in Indicates the 2nd harmonic voltage amplitude (unit V), corresponding to the 100Hz point modulus value in the frequency domain. Indicates the amplitude of the third harmonic (unit: V), corresponding to the modulus value at 150Hz. Indicates the 6th harmonic amplitude (unit V), corresponding to the 300Hz point modulus. Then, according to the frequency band correspondence of the load response split sample set record, filter out the sample records with frequency band numbers 2, 3, and 6 from the field "frequency_band", and establish a mapping relationship with the voltage harmonic amplitude data. For example, if sample A is marked with "frequency_band:3" in the record file, then extract all the measurement points under this sample. Amplitude sequence, forming the corresponding harmonic amplitude sequence , where the subscript Indicates the harmonic order (i.e. 2, 3, 6). The superscripts (1), (2), and (3) respectively represent the harmonic amplitude data corresponding to the three complete cycles before, during, and after the disturbance. The unit is volt (V). For example, if the amplitudes of the third harmonics in a certain sampling record are 1.8V, 3.2V, and 2.6V, respectively, then the third harmonic amplitude sequence of the sample is After all samples in the sampling record are matched one by one with the corresponding frequency band number and harmonic number, all the extracted , , The data is stored as a structured vector sequence for the next step of change analysis.

[0028] S312: extract the amplitude change path of each period before, during and after the disturbance in the harmonic amplitude sequence with consistent frequency band information, analyze the jump times and amplitude fluctuation characteristics, count the change trends under continuous periods, and output the frequency band amplitude change data; Extract the amplitude change path of each period before, during and after the disturbance in the harmonic amplitude sequence with consistent frequency band information. Assume that the current processing frequency band is the third harmonic, and the amplitude sequence is ,in represents the first period before the disturbance Subharmonic amplitude, in volts (V), represents the amplitude of the corresponding period during the disturbance process, Represents the amplitude of the corresponding period after the disturbance. Taking sample A as an example, it is recorded as , respectively calculate the amplitude variation of two adjacent cycles, assuming The amplitude change difference of the subharmonic frequency band during the next cycle is , , in this example , The jump recognition criterion sets the jump determination threshold to This value is obtained by referring to the harmonic stability requirements in the GB / T24337-2009 standard. For any period Then it is recognized as one jump event. Therefore, there are two jump events in the 3rd harmonic in this sample, denoted as the number of jumps Subsequently, the maximum change amplitude of the amplitude in this frequency band during three periods is statistically calculated and denoted as the maximum amplitude fluctuation which is used to reflect the disturbance intensity. Finally, a trend label is generated according to the amplitude change direction in three periods. If then mark "↑", if then mark "↓", and if they are equal, mark "=". Finally, it is combined into a trend string In this example, the trend is "↑↓". Therefore, the statistical result of the change in this sample frequency band is output as: harmonic number amplitude sequence , number of jumps , maximum fluctuation amplitude , trend direction label , where: represents the harmonic serial number; represents the array composed of the amplitudes of the th harmonic in three consecutive periods; represents the number of jump events identified in this array; represents the difference between the maximum value and the minimum value among the three values; =\left [ {0.6,0.62,0.58} \right ] In this case, the following are calculated respectively: (only the first group change exceeds 0.5), , , =0, 0.04V, the trend is "↑↓". Finally, the analysis results of the changes in each frequency band are complete in structure, and the values of each data item are clear, meeting the requirements of sample statistical output.

[0029] Please refer to Figure 5 , the specific steps for obtaining the interference coupling fluctuation section are as follows: S411: Obtain the change trend of the fundamental frequency total power of the meter load monitored by the power quality analyzer within the disturbance period, establish power sequence data according to the period, and obtain the fundamental frequency total power change sequence; To obtain the change trend of the fundamental frequency total power of the meter load monitored by the power quality analyzer, first set the voltage and current signals corresponding to the monitoring time period. After extracting the fundamental wave frequency (50 Hz) from the three-phase waveform data, calculate the instantaneous active power of each period respectively and perform a period average operation to obtain the period average power value, denoted as , where the subscript j represents the period number, with the unit of watt (W). Each represents the average total active power value within the j-th complete power frequency period. Suppose the disturbance duration covers 3 complete periods, then construct the power sequence , where P is the fundamental frequency power change sequence vector, and the element , , are the active powers of the 1st, 2nd, and 3rd periods respectively, with the unit of watt (W). Taking the sample as an example, if the effective values of the three-phase voltages are , = 219 V, = 221 V, and the effective value of the current is 10 A, , = 10.1 A, where , , represent the effective values of the three-phase voltages (unit: V), , , represent the effective values of the three-phase currents (unit: A), and the power factor is . Substitute and calculate , where and are the average values of the three-phase voltage and current respectively, with the units of V and A. If the current in the 2nd period increases to , = 11.8 A, , similarly, , and the current in the 3rd period drops to , , , then 6930 W. Finally, obtain the fundamental frequency active power change sequence , as the record of the total power change within the disturbance period.

[0030] S412: Extract the change trend of the harmonic amplitude at the corresponding moment according to the change data of the frequency band amplitude, and perform period correspondence pairing with the fundamental frequency total power change sequence to generate the amplitude-power ratio pair sequence; Extract the harmonic amplitude change trend at the corresponding moment according to the frequency band amplitude change data. First, call the amplitude sequence corresponding to the th harmonic obtained previously , where represents the frequency band of the 3rd harmonic, represents the amplitude (unit: V) of the th harmonic in the Kth cycle, , with the unit of volt (V), construct an amplitude-power pairing structure, and pair the harmonic amplitude of each cycle with the fundamental frequency power of the corresponding cycle to form paired elements and form a paired sequence , where M is a vector group of "amplitude-power" ratio pairs, and the elements are binary tuples. The former is the harmonic amplitude of this cycle, and the latter is the fundamental frequency active power of this cycle. If in sample A and , then there is , and organize the pairing results into a key-value structure {"t1":[1.8,6270],"t2":[3.2,7524],"t3":[2.6,6930]} according to the time tags, where "t1", "t2", and "t3" respectively correspond to the time sequence tags of the pre-disturbance, in-disturbance, and post-disturbance cycles, and complete the amplitude and power trend alignment pairing structure.

[0031] S413: Identify the cycle segments with consistent jump trends of the two types of trends in the amplitude-power ratio sequence through the cumulative sum test algorithm, and judge the interference coupling fluctuation segments with synchronous fluctuations; Identify the cycle segments with consistent trends in the amplitude-power ratio sequence M through the cumulative sum test algorithm. First, define the trend change method. Let the amplitude change amount of the th harmonic between two cycles, and the fundamental frequency power change amount is , where represents the cycle index, judge whether the signs of the two are the same, and set the sign consistency marking value as , where is the sign function. If the product is greater than 0, then indicates consistent trends, and if the product is less than 0, then indicates inconsistent trends. For example, in sample A , the change in the first segment is , , the product +1.4×+1245= , so =+1; The change in the second segment is , , the product , so , add up all to obtain the total number of periods with consistent trends by cumulative summation , where represents the periods with consistent trends, with the unit of number of segments (dimensionless), and set the threshold for determining consistent trends as . According to the sampling segment number n = 3, the threshold is taken as n - 1, that is, 2. If then it is determined that there is an interference coupling trend section. This value is derived from the principle that the time of synchronous jitter of power and amplitude within the interference section should cover at least two periods. Finally, it is determined that there is an interference coupling fluctuation section with trend synchronization within the disturbance period in sample A, and the coupling characteristic is established.

[0032] Please refer to Figure 6 , and the steps for obtaining the electricity meter load monitoring results are specifically as follows: S511: Extract the frequency points with jumps in each period within the interference coupling fluctuation section, count the corresponding jump amplitudes and jump frequencies, and obtain the jump frequency point index set; Extract the frequency points with jumps in each period within the interference coupling fluctuation section. First, divide the frequency data collected by the electricity meter into multiple time periods according to the period (with 50Hz as the reference), each period is 20ms, and the amount of data in each period is 20 data points according to the 1kHz sampling rate. Read the frequency change sequence in each period in turn. By calculating the frequency difference between adjacent data points, if the frequency difference is greater than the set jump threshold Δf (set as 0.3Hz), then record the current frequency point as a jump point. For example, if the frequency of the 9th data point is 49.8Hz and the 10th is 50.2Hz, then the difference is 0.4Hz, which meets the jump condition, and it is recorded as the jump frequency point 50.2Hz. Record its jump amplitude as 0.4Hz, and the jump frequency is accumulated once within one period. By analogy, traverse all periods, count the number of times each jump frequency point accumulates in the observation period and the corresponding average jump amplitude, and obtain the jump frequency point index set, where each frequency point is recorded as follows: frequency value, number of jumps, average jump amplitude. If 20 periods are sampled, at most 200 data points can be recorded, and the selected frequency points such as 50.2Hz, 49.6Hz, etc. each have independent statistical entries and are stored in the index set.

[0033] S512: According to the jump frequency point index set, screen out the set of frequency points with prominent jump amplitudes and frequencies, and construct a candidate active frequency point sequence; According to the jump frequency point index set, traverse the jump frequencies and average jump amplitudes of each frequency point. First, calculate the mean and standard deviation of all frequency point jump frequencies, and the mean , on this basis, if the hopping frequency of a certain frequency point is greater than +1.5 and its average hopping amplitude is greater than +1.5 , it is determined as a frequency point with prominent hopping characteristics and classified into the candidate active frequency point set. For example, for 50 frequency points, if the average frequency is 4 times and the standard deviation is 1.2 times, and the average amplitude is 0.25 Hz and the standard deviation is 0.05 Hz, then the frequency point needs to satisfy the frequency > 5.8 times and the amplitude > 0.325 Hz. For example, the hopping frequency of 49.6 Hz is 6 times and the amplitude is 0.38 Hz, which meets the conditions and is selected into the candidate set. This set is used as the input sequence for subsequent clustering analysis, including the frequency point value and its statistical characteristics. In this process, all screened frequency points need to meet the dual conditions of frequency and amplitude to avoid introducing errors due to accidental jumps.

[0034] S513: Use the density clustering algorithm to perform clustering analysis on the candidate active frequency point sequence, identify the activation frequency points exceeding the preset frequency threshold, and output the electricity meter load monitoring results for amplitude limiting processing of interference frequency points and load identification optimization; Use the density clustering algorithm to perform clustering analysis on the candidate active frequency point sequence. First, represent each candidate frequency point as a two-dimensional vector , where represents the actual frequency value of the frequency point, with the unit of Hz, represents the number of times of cumulative jumps of this frequency point in multiple cycles, with the unit of times. This vector sequence forms a point set , where n is the number of candidate frequency points. To improve the density perception accuracy, a density weight evaluation parameter is introduced to measure the aggregation degree of a certain frequency point in its local area. Its expression is: , where, is the density value of frequency point i, represents all frequency points that satisfy , indicating the neighborhood range of this point, is the smoothing coefficient, with the unit of Hz and the set value of 0.2 Hz, The neighborhood radius is set to 0.5 Hz, which is used to determine whether points with close frequency values form a clustering relationship. MinPts is the clustering core point threshold, set to 3, indicating that at least 2 neighbor frequency points are required. The judgment logic is as follows: Traverse all candidate frequency points, calculate the number of qualified points in their neighborhoods one by one. If the number of neighborhood points is greater than or equal to MinPts, the current frequency point is a core point, and record all frequency points in its neighborhood as members of the same cluster. Then continue to expand other core points and merge the clusters with overlapping neighborhoods to form several density frequency cluster sets, and calculate the average frequency value within each frequency cluster With the average frequency hopping frequency , set the activation frequency determination threshold and the density threshold . Filter out the clusters of frequency points whose average frequency value is greater than this threshold and whose density is greater than this value, and finally use them as active activation frequency points for subsequent processing. Suppose a set of candidate frequency points is as follows: , = 7 times, = 49.6 Hz, = 6 times, = 49.7 Hz, = 5 times. For the point Hz, its neighborhood is Hz, and the neighborhood contains the point . Calculate the density value as follows: , . Since , and the number of neighborhood points is 2 ≥ MinPts - 1, it is determined as a core point, and its clustering cluster is {49.5, 49.6, 49.7} Hz. The average frequency within the cluster is: . The average frequency hopping frequency is: times. Since , this frequency cluster does not meet the activation condition, so it is not output. However, if another frequency cluster contains the points: times), times) . Calculate its: , times, . It meets and . It is determined as an activation frequency cluster. Finally, output the frequency points 50.1 Hz, 50.2 Hz, and 50.3 Hz within this cluster, which form the activation frequency point sequence for monitoring output and participate in the subsequent interference frequency point amplitude limiting and meter load identification steps. This step constructs a frequency point recognition mechanism with high precision through the combination of frequency hopping frequency and local density.

[0035] An electricity meter load monitoring system based on a power quality analyzer. The electricity meter load monitoring system based on a power quality analyzer is used to execute the above-mentioned electricity meter load monitoring method based on a power quality analyzer. The system includes: The cross-phase synchronization recognition module obtains the three-phase voltages of the electricity meter through the power quality analyzer, extracts the phase shift angles of the three-phase voltages within the disturbance period to judge the synchronization characteristics between phases, and forms a cross-phase synchronization evaluation result; Based on the cross-phase synchronization evaluation results, the load response attribution module identifies the load response attribution type of the three-phase voltage phase offset angle, establishes a load response dissection sample set, obtains the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, and establishes a frequency band correspondence relationship; The frequency band amplitude analysis module refers to the frequency band correspondence relationship of the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, judges the change trend of the harmonic amplitude within the frequency band, and obtains the frequency band amplitude change data; The power trend comparison module obtains the change trend of the fundamental frequency total power of the meter load monitored by the power quality analyzer during the disturbance period, compares the change trend of the harmonic amplitude at the corresponding moment in the frequency band amplitude change data with the change trend of the fundamental frequency total power, and identifies the interference coupling fluctuation section; The active frequency point identification module extracts the frequency points with prominent jump amplitudes and jump frequencies in the interference coupling fluctuation section, performs active frequency point analysis on the frequency points, and outputs the meter load monitoring results.

[0036] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still belong to the protection scope of the technical solution of the present invention.

Claims

1. A method for monitoring the load of an electric meter based on a power quality analyzer, characterized in that, It includes the following steps: S1: Obtain the three-phase voltage of the electric meter through a power quality analyzer, extract the phase shift angles of the three-phase voltage during the disturbance period to judge the synchronization characteristics between phases, and form a cross-phase synchronization evaluation result; S2: Based on the cross-phase synchronization evaluation result, identify the load response attribution types of the phase shift angles of the three-phase voltage, establish a load response dissection sample set, and obtain the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set to establish a frequency band correspondence relationship; S3: Refer to the frequency band correspondence relationship of the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, judge the change trend of the harmonic amplitude within the frequency band, and obtain the frequency band amplitude change data; S4: Obtain the change trend of the fundamental frequency total power of the electric meter load monitored by the power quality analyzer during the disturbance period, compare the change trend of the harmonic amplitude at the corresponding moment in the frequency band amplitude change data with the change trend of the fundamental frequency total power, and identify the interference coupling fluctuation section; S5: Extract the frequency points with prominent jump amplitudes and jump frequencies in the interference coupling fluctuation section, perform active frequency point analysis on the frequency points, and output the electric meter load monitoring result.

2. The method for monitoring the load of an electric meter based on a power quality analyzer according to claim 1, wherein The cross-phase synchronization evaluation result includes the phase shift direction feature, the inter-phase response consistency level, and the synchronous stability change condition. The load response dissection sample set includes the load combination structure label, the response mode classification result, and the phase synchronization performance record. The frequency band amplitude change data is specifically the jump amplitude trajectory within the frequency band, the cycle stability characteristic index, and the frequency fluctuation distribution state. The interference coupling fluctuation section specifically refers to the harmonic disturbance superposition interval, the power change synchronization region, and the frequency band behavior resonance node. The electric meter load monitoring result includes the load type identification result, the interference frequency identification label, and the load structure response information.

3. The method for monitoring the load of an electric meter based on a power quality analyzer according to claim 1, wherein The specific steps for obtaining the cross-phase synchronization evaluation result are as follows: S111: Obtain the A, B, and C three-phase voltage waveform data recorded by the power quality analyzer within a specified time period, and extract the voltage phase shift angle change sequences in three cycles before, during, and after the disturbance; S112: Based on the voltage phase shift angle change sequence, use the cosine similarity algorithm to calculate the cosine values of the included angles between the three pairs of angle change vectors of A-B, B-C, and A-C in the sequence respectively, and judge the inter-phase synchronization matching degree at each sampling point within the corresponding cycle according to the cosine value of the included angle to obtain the phase shift path similarity analysis result; S113: Judge the average similarity change trend of the three pairs of angle change vectors of A-B, B-C, and A-C in the three consecutive cycle sections in the phase shift path similarity analysis result to form a cross-phase synchronization evaluation result.

4. The method for monitoring the meter load based on the power quality analyzer according to claim 3, characterized in that, The specific steps for obtaining the frequency band correspondence relationship are as follows: S211: Judge the inter-phase response mode of the load to the disturbance according to the three pairs of angle change vectors of A-B, B-C, and A-C in the cross-phase synchronization evaluation result. If it shows a similar coincidence state in three consecutive cycles, it is classified as a concentrated load response sample. If there are phase-by-phase or amplitude response differences between different phases, it is classified as a parallel load response sample; S212: Integrate the centralized load response samples and the shunt load response samples, and perform structural and time index annotation on the samples to form a load response dissection sample set; S213: By invoking the phase calibration interface and frequency component measurement function of the power quality analyzer, obtain the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, and establish the frequency band correspondence between the three-phase voltage and the frequency band characteristics.

5. The method for monitoring the load of an electric meter based on a power quality analyzer according to claim 4, wherein The specific steps for obtaining the frequency band amplitude change data are as follows: S311: Obtain the harmonic voltage amplitudes in the 2nd, 3rd, and 6th frequency bands recorded in the power quality analyzer, and extract the harmonic amplitude sequence whose frequency band information is consistent with that marked in the load response dissection sample set and the frequency band correspondence; S312: Extract the amplitude change paths of each cycle before, during, and after the disturbance in the harmonic amplitude sequence with consistent frequency band information, analyze the number of jumps and the amplitude fluctuation characteristics, and statistically analyze the change trend under continuous cycles to output the frequency band amplitude change data.

6. The method for monitoring the load of an electric meter based on a power quality analyzer according to claim 5, characterized in that, The specific steps for obtaining the interference coupling fluctuation section are as follows: S411: Obtain the change trend of the fundamental frequency total power of the meter load monitored by the power quality analyzer during the disturbance period, establish power sequence data according to the cycle to obtain the fundamental frequency total power change sequence; S412: Extract the change trend of the harmonic amplitude at the corresponding moment according to the frequency band amplitude change data, and perform cycle correspondence pairing with the fundamental frequency total power change sequence to generate an amplitude-power ratio sequence; S413: Identify the cycle segments in the amplitude-power ratio sequence where the jump trends of the two types of trends are consistent during the disturbance period through the cumulative sum test algorithm, and judge the interference coupling fluctuation section of synchronous fluctuation.

7. The method for monitoring the load of an electric meter based on a power quality analyzer according to claim 6, wherein The specific steps for obtaining the monitoring result of the meter load are as follows: S511: Extract the frequency points with jumps in each cycle in the interference coupling fluctuation section, and count the corresponding jump amplitudes and jump frequencies to obtain a jump frequency point index set; S512: According to the jump frequency point index set, screen the set of frequency points with prominent jump amplitudes and frequencies, and construct a candidate active frequency point sequence; S513: Use the density clustering algorithm to perform clustering analysis on the candidate active frequency point sequence, identify the activation frequency points exceeding the preset frequency threshold, and output the monitoring result of the meter load for the amplitude limiting processing of interference frequency points and the optimization of load identification.

8. An electric meter load monitoring system based on a power quality analyzer, characterized in that, According to the method for monitoring a meter load based on a power quality analyzer according to any one of claims 1-7, the system includes: The cross-phase synchronization recognition module obtains the three-phase voltage of the meter through the power quality analyzer, extracts the phase offset angles of the three-phase voltage during the disturbance period to judge the synchronization characteristics between phases, and forms a cross-phase synchronization evaluation result; The load response attribution module, based on the cross-phase synchronization evaluation result, identifies the load response attribution type of the phase offset angles of the three-phase voltage, establishes a load response dissection sample set, and obtains the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set to establish a frequency band correspondence; The frequency band amplitude analysis module, with reference to the frequency band correspondence of the frequency band characteristics corresponding to the three-phase voltage in the load response dissection sample set, judges the change trend of the harmonic amplitude within the frequency band to obtain the frequency band amplitude change data; The power trend comparison module obtains the change trend of the fundamental frequency total power of the meter load monitored by the power quality analyzer within the disturbance period, compares the change trend of the harmonic amplitude at the corresponding moment in the frequency band amplitude change data with the change trend of the fundamental frequency total power, and identifies the interference coupling fluctuation section; The active frequency point identification module extracts the frequency points with prominent jump amplitudes and jump frequencies in the interference coupling fluctuation section, performs active frequency point analysis on the frequency points, and outputs the meter load monitoring results.

Images