Energy consumption recognition system based on power transformation aliasing scenario

Abstract

The present application relates to the technical field of electric energy identification, in particular to an energy consumption identification system based on a power transformation aliasing scene, which comprises a disturbance boundary extraction module, a boundary stability determination module, an aliasing state identification module, a periodic power calculation module and an energy consumption attribution determination module.In the present application, the start and end points of the periodic voltage fluctuation index are continuously tracked to form a voltage sag record with phase offset characteristics, a stable fluctuation determination mechanism is constructed in combination with offset statistics, and the accurate identification of the aliasing section is realized with the help of disturbance amplitude and frequency extraction, the power sequence is constructed on the basis of voltage and current synchronous sampling values, the periodic integral is completed, the high-precision calculation result of energy consumption is formed, the fine division of the cross-interference periodic energy consumption is realized under the condition of no physical isolation, the dependence on external metering equipment is avoided, the independent quantification of the overlapping load energy consumption is realized, and a feasible means is provided for non-intrusive load identification and energy efficiency evaluation in a high-coupling scene.

Description

Technical Field

[0001] This invention relates to the field of power consumption identification technology, and in particular to a power consumption identification system based on substation aliasing scenarios. Background Technology

[0002] The field of power energy identification technology involves the accurate identification and quantification of the operating status, power consumption characteristics, and changing patterns of various energy-consuming devices in power systems. This includes the acquisition and analysis of basic electrical parameters such as current and voltage, modeling and discrimination of power usage behavior, and non-intrusive power consumption identification methods in multi-load interference or complex environments. These methods are widely used in systemic scenarios such as smart grids, energy-saving monitoring, load management, and energy efficiency assessment. Traditional energy consumption identification systems based on substation aliasing scenarios refer to systems that independently identify the power consumption of each device in the context of multiple substation devices operating simultaneously and experiencing cross-interference. These systems typically acquire electrical parameters from each device through the installation of multiple independent power metering instruments in a physically isolated manner, or rely on existing classified power data to estimate the energy consumption data of each circuit through a circuit-by-circuit matching method.

[0003] Traditional energy consumption identification in substation aliasing scenarios mainly relies on multiple independent energy metering instruments to collect parameters of electrical equipment in a physically isolated manner, or on path matching and calculation based on existing classified energy data. In the context of multiple loads connected in parallel and harmonic disturbances, it lacks the ability to identify real-time disturbance characteristics and makes it difficult to accurately delineate overlapping energy consumption areas. For example, sag fluctuations generated when multiple devices start up may be uniformly classified into a single channel, leading to misjudgments and energy consumption attribution errors. In practical applications where independent metering devices cannot be installed or data classification is unclear, the system's identification performance and data resolution are both limited. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies and propose an energy consumption identification system based on substation aliasing scenarios.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an energy consumption identification system based on substation aliasing scenarios includes:

[0006] The disturbance boundary extraction module acquires periodic voltage signal data under the parallel operation state of substation loads, collects the start and end point index positions of voltage fluctuation areas in each cycle, calculates phase offset, statistically analyzes and records the absolute value of offset, and generates voltage sag event records.

[0007] The boundary stability determination module calculates the difference range between the maximum and minimum offset values ​​in any five consecutive cycles based on the voltage sag event records. If the difference range is less than the voltage fluctuation tolerance, it is determined to be a stable boundary and the cycle number is recorded to generate a list of disturbance stable cycles.

[0008] The aliasing state identification module extracts the peak amplitude of the disturbance and the number of disturbances within the corresponding period based on the list of disturbance stable periods, calculates the period disturbance intensity, and if it is greater than the total harmonic distortion rate threshold, it is marked as an aliasing period, and the start and end index intervals of all overlapping segments are located to generate aliasing period identification results.

[0009] The periodic power calculation module extracts the voltage and current sampling values ​​point by point within the corresponding segment based on the aliasing period identification result, performs the product of the corresponding time points to form a power sequence, and accumulates and integrates the sequence to obtain the total energy consumption value within a unit period, and outputs the periodic energy consumption calculation data.

[0010] As a further embodiment of the present invention, the voltage sag event record includes an absolute value of offset statistics, a phase offset index, and an event start and end marker; the disturbance stability cycle list includes a set of cycle numbers, tolerance comparison results, and difference range evaluation indicators; the aliasing cycle identification results include an aliasing cycle number, an overlapping segment index, and a total harmonic exceedance marker; and the cycle energy consumption calculation data includes a power time series, a cycle cumulative energy value, and voltage-current sampling pairs.

[0011] As a further aspect of the present invention, the disturbance boundary extraction module includes:

[0012] The periodic voltage acquisition submodule acquires voltage signal data for each cycle under the parallel operation of substation loads, plots the voltage fluctuation curve within each cycle, and extracts the voltage extreme points, voltage fluctuation edge points and corresponding timestamp information in each cycle to generate periodic voltage signal fluctuation values.

[0013] The start and end index positioning submodule detects the voltage fluctuation region boundary within each cycle based on the periodic voltage signal fluctuation value, identifies the index position of the start and end of the voltage fluctuation by setting a voltage fluctuation judgment threshold, and judges whether the fluctuation region has stability characteristics based on the boundary changes of adjacent cycles, thereby obtaining the voltage fluctuation boundary index sequence.

[0014] The phase offset statistics submodule performs difference statistics and absolute value calculation on the start and end index positions within a continuous period according to the voltage fluctuation boundary index sequence and the voltage fluctuation measurement requirements. It summarizes the degree of offset of the boundary index in the periodic sequence and statistically obtains the voltage sag event records.

[0015] As a further aspect of the present invention, the boundary stability determination module includes:

[0016] The offset difference extraction submodule extracts the offset values ​​corresponding to all cycles involved in each event based on the voltage sag event record. It sorts the offset values ​​in each set of five consecutive cycles, identifies the maximum and minimum values ​​in the sequence, and uses the difference as the difference range of the current sequence to obtain the offset difference range set.

[0017] The voltage tolerance comparison submodule compares the difference range of each sequence with the voltage fluctuation tolerance threshold according to the offset difference range set. If the difference range is less than the voltage fluctuation tolerance threshold, it is marked as a boundary stable cycle; otherwise, it is marked as an unstable cycle, thus obtaining the boundary stability label sequence.

[0018] The stable period generation submodule extracts the numbers of all periods marked as stable states based on the boundary stability label sequence, establishes an index mapping according to the original order in the full period sequence, and establishes a unique identifier structure for the period sequence to obtain a list of perturbation stable periods.

[0019] As a further aspect of the present invention, the aliasing state identification module includes:

[0020] The disturbance feature extraction submodule extracts voltage disturbance data for the corresponding period based on the disturbance stability period list, counts the peak amplitude of the disturbance waveform and the total number of disturbance events in each period, and performs the above operations sequentially for all periods to obtain disturbance intensity calculation input data;

[0021] The intensity threshold judgment submodule calculates the input data based on the disturbance intensity, calculates the disturbance intensity of each cycle, and judges it item by item with the total harmonic distortion rate threshold. If it is greater than the total harmonic distortion rate threshold, the cycle is marked as an aliasing cycle, and a corresponding structure of cycle number and judgment result is established to generate an aliasing cycle label sequence.

[0022] The aliasing segment location submodule extracts the cycle number of all aliasing cycles according to the aliasing cycle label sequence, performs segment location in the original voltage disturbance signal data, extracts the start point index and end point index of the disturbance curve in each aliasing cycle, records the start and end index values ​​of each disturbance interval and forms an aliasing segment index set, and obtains the aliasing cycle identification result.

[0023] As a further aspect of the present invention, the periodic power calculation module includes:

[0024] The power sample extraction submodule locates the complete time interval of the period in the original signal sequence based on the sampling start and end point index of the corresponding period in the aliasing period identification result, extracts the voltage and current values ​​at each time point, and synchronizes them according to the time index. Each time point has a set of corresponding voltage-current data pairs, and the electrical parameter synchronization sequence is obtained by sorting them out.

[0025] The instantaneous power accumulation submodule performs a product operation based on the voltage and current values ​​at each time point in the electrical parameter synchronization sequence to form instantaneous power data at each point, calculates and obtains the total energy consumption per unit cycle, and generates cycle energy consumption integral data.

[0026] The cycle energy consumption generation submodule reconstructs the energy consumption value sequence within all cycles according to the cycle energy consumption integral data and the cycle number sequence. It uses the cycle number as the index and the cycle energy consumption value as the content to integrate them into a two-dimensional cycle energy consumption array structure and outputs the cycle energy consumption calculation data.

[0027] As a further aspect of the present invention, the system further includes:

[0028] The energy consumption attribution determination module extracts three feature indicators—boundary difference, voltage slope, and current amplitude—of all disturbance sources in the same period based on the periodic energy consumption calculation data, constructs a disturbance feature vector, compares it with the typical feature vector in the stable period, filters the load units to which the high similarity features belong and assigns corresponding power, and generates energy consumption identification results for substation aliasing scenarios.

[0029] The energy consumption identification results of the substation aliasing scenario include a set of disturbance feature vectors, a similarity comparison matrix, and load unit energy consumption allocation information.

[0030] As a further aspect of the present invention, the energy consumption attribution determination module includes:

[0031] Based on the periodic energy consumption calculation data, the feature acquisition submodule extracts the sampling sequence within the corresponding period, extracts the voltage values ​​at the start and end of the disturbance in the corresponding sampling point sequence, calculates the difference as the boundary difference, constructs a continuous difference sequence for all voltage sampling points within the disturbance segment, calculates the mean of the sequence as the voltage slope, and counts the maximum amplitude of all current sampling points within the disturbance segment as the current amplitude, and integrates to obtain the disturbance feature vector set.

[0032] The similarity vector comparison submodule compares each vector with the known typical load feature vector in the stable period according to the disturbance feature vector set, calculates the three-dimensional distance as the similarity index, sorts them according to the distance value from smallest to largest, selects the target load number with the smallest distance as the matching load, and obtains the disturbance-load similarity mapping set.

[0033] The energy consumption allocation generation submodule divides the energy consumption of each disturbance source into equal parts based on the matching relationship between each disturbance source and the corresponding load number in the disturbance-load similarity mapping set, combined with the energy consumption value corresponding to the periodic energy consumption calculation data. It then summarizes the allocated energy consumption of each disturbance source under the corresponding load number to obtain the energy consumption identification result of the substation aliasing scenario.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0035] In this invention, by continuously tracking the start and end point indexes of periodic voltage fluctuations, a voltage sag record with phase offset characteristics is formed. A stable fluctuation judgment mechanism is constructed by combining offset statistics, and the aliasing section is accurately identified by extracting the disturbance amplitude and frequency. A power sequence is constructed based on the synchronous sampling values ​​of voltage and current and the periodic integration is completed to form a high-precision calculation result of energy consumption. This enables the fine division of cross-interference periodic energy consumption under conditions without physical isolation, avoids dependence on external metering equipment, improves the accuracy of energy consumption attribution judgment in complex interference scenarios, and realizes independent quantification of overlapping load energy consumption. This provides a feasible means for non-intrusive load identification and energy efficiency assessment in highly coupled scenarios. Attached Figure Description

[0036] Figure 1 This is a system flowchart of the present invention;

[0037] Figure 2 This is a flowchart of the disturbance boundary extraction module of the present invention;

[0038] Figure 3 This is a flowchart of the boundary stability determination module of the present invention;

[0039] Figure 4 This is a flowchart of the aliasing state recognition module of the present invention;

[0040] Figure 5 This is a flowchart of the periodic power calculation module of the present invention;

[0041] Figure 6 This is a flowchart of the energy consumption attribution determination module of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 merely illustrative and not intended to limit the invention.

[0043] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0044] Please see Figure 1 The energy consumption identification system based on substation aliasing scenarios includes:

[0045] The disturbance boundary extraction module acquires periodic voltage signal data under the parallel operation state of the substation load, collects the start and end point index positions of the voltage fluctuation area in each cycle, calculates the phase offset of the start and end index in the continuous cycle (in accordance with the voltage fluctuation measurement method in IEC61000-4-30 standard), counts the absolute value of the offset and records it, and generates a voltage sag event record (refer to the power quality monitoring standard).

[0046] The boundary stability determination module calculates the difference range between the maximum and minimum offset values ​​in any five consecutive cycles based on the voltage sag event records. It compares the difference range with the voltage fluctuation tolerance (implementing the ±10% standard of "Power Quality Supply Voltage Deviation"). If it is less than the voltage fluctuation tolerance, it is determined to be a stable boundary and the cycle number is recorded to generate a list of disturbance stability cycles.

[0047] The aliasing state identification module extracts the peak amplitude of the disturbance and the number of disturbances within the corresponding period based on the list of stable disturbance periods, calculates the periodic disturbance intensity, and compares it with the total harmonic distortion rate threshold (≤8% according to the standard). If it is greater than the total harmonic distortion rate threshold, it is marked as an aliasing period, and the start and end index intervals of all overlapping segments are located to generate aliasing period identification results.

[0048] The periodic power calculation module extracts the voltage and current sampling values ​​of the corresponding segment based on the aliasing period identification result, performs the product of the corresponding time points to form a power sequence, and accumulates and integrates the sequence to obtain the total energy consumption value within a unit period, and outputs the periodic energy consumption calculation data.

[0049] The energy consumption attribution determination module extracts three feature indicators—boundary difference, voltage slope, and current amplitude—of all disturbance sources in the same cycle based on the cycle energy consumption calculation data, constructs a disturbance feature vector, compares it with the typical feature vector in the stable cycle, filters the load units to which the high similarity features belong and assigns corresponding power, and generates energy consumption identification results for substation aliasing scenarios.

[0050] The voltage sag event record includes the absolute value of the offset statistics, the phase offset index, and the event start and end markers. The disturbance stability cycle list includes the cycle number set, tolerance comparison results, and difference range evaluation indicators. The aliasing cycle identification results include the aliasing cycle number, the overlapping section index, and the total harmonic exceedance marker. The cycle energy consumption calculation data includes the power time series, the cycle cumulative energy value, and the voltage and current sampling pairs. The substation aliasing scenario energy consumption identification results include the disturbance feature vector set, the similarity comparison matrix, and the load unit energy consumption allocation information.

[0051] Please see Figure 2 The disturbance boundary extraction module includes:

[0052] The periodic voltage acquisition submodule acquires voltage signal data for each cycle under the parallel operation of substation loads, plots the voltage fluctuation curve within each cycle, and extracts the voltage extreme points, voltage fluctuation edge points and corresponding timestamp information in each cycle to generate periodic voltage signal fluctuation values.

[0053] Voltage signal data for each cycle under parallel operation of substation loads is collected by sampling four typical load outgoing circuits in an actual urban power distribution system. The total load current is set between 20 and 50 A. Sampling points are deployed at the load output terminals, and voltage sensors are configured to continuously record the voltage signal curve at a frequency of 5000 Hz. The sampling time is 10 seconds, with a single cycle set to 20 ms, corresponding to 100 sampling points. Maximum and minimum values ​​are extracted from the waveform data within each cycle to identify voltage extreme points. A moving average correction method is used to filter out high-frequency interference. Extreme points are determined by analyzing the relationship between each point and its two preceding and following points. Index position; In cycle 1, the maximum value is identified as 235.6V (index 12) and the minimum value as 231.4V (index 48), so the fluctuation amplitude is 4.2V. Then, the slope of the waveform in each cycle is calculated, and the rate of change of each point is derived using the three-point center difference method. When the rate of change is continuously positive and greater than 1.5V / ms, the starting point of the rising edge of the fluctuation is determined, and when it is continuously negative and less than −1.5V / ms, the ending point of the falling edge is determined. In cycle 2, the sliding detection start index is 11 and the end index is 46, the fluctuation amplitude is 4.8V, and the cycle duration is 20ms. Finally, the fluctuation boundary characteristic data under different cycles are obtained, as detailed in the table below.

[0054] Table 1. Measurement data of periodic voltage fluctuations:

[0055] ;

[0056] As shown in Table 1, the voltage fluctuation amplitude from cycle 1 to cycle 4 is between 4.0V and 5.1V, and the corresponding start and end index ranges are mostly concentrated between 12 and 50, thus obtaining the periodic voltage signal fluctuation value.

[0057] The start and end index positioning submodule detects the voltage fluctuation region boundary within each cycle based on the periodic voltage signal fluctuation value. It identifies the index position of the start and end of the voltage fluctuation by setting a voltage fluctuation judgment threshold, and judges whether the fluctuation region has stability characteristics based on the boundary changes of adjacent cycles, thereby obtaining the voltage fluctuation boundary index sequence.

[0058] Based on the periodic voltage signal fluctuation values, the process of detecting the boundary of the voltage fluctuation region within each cycle is performed by identifying the start and end indices of the fluctuation and comparing their range with a set threshold. The fluctuation boundary identification threshold is set to 30 sampling points, corresponding to a time span of 6ms. Specifically, the start and end indices of each cycle are extracted from Table 4, the fluctuation region span is calculated, and compared with the threshold. In cycle 1, the index span is 48−12=36, and in cycle 2 it is 46−11=35, both exceeding the threshold. Therefore, they are determined to be valid fluctuation segments. The amplitude information in this segment is then extracted for subsequent boundary stability analysis. Subsequently, the start and end points of the fluctuations in different cycles are... The point indices are compared horizontally to determine whether the continuity meets the stability criteria. The stability boundary judgment condition is set as follows: if the fluctuation of the boundary start and end indices does not exceed ±5 sampling points in three consecutive cycles, the boundary is considered stable. In cycles 1 to 3, the starting indices are 12, 11, and 13, with a variation range of ±1 to 2, and the ending indices are 48, 46, and 50, with a variation range of ±2 to 4. The stability criteria are met, and the index points are recorded to form the boundary index sequence {12, 11, 13} and {48, 46, 50}. Each segment is numbered for subsequent offset calculations, and the voltage fluctuation boundary index sequence is finally obtained.

[0059] The phase offset statistics submodule performs difference statistics and absolute value calculation on the start and end index positions within a continuous period according to the voltage fluctuation boundary index sequence and the voltage fluctuation measurement requirements. It summarizes the degree of offset of the boundary index in the period sequence and statistically obtains the voltage sag event records.

[0060] Based on the voltage fluctuation boundary index sequence, period boundary offset statistical processing is performed. The starting offset between period 2 and period 1 is extracted as 11-12=-1, and the ending offset is 46-48=-2. The starting offset between period 3 and period 2 is 13-11=+2, and the ending offset is 50-46=+4. The absolute values ​​of each offset are 1, 2, 2, and 4, respectively. The total offset of these four items is recorded as 9 sampling points. Then, it is compared with the event recording benchmark value, which is set to 5 sampling points. If the cumulative value of continuous offset exceeds 5, it is recorded as a sag event. Periods 1 to 3 have reached 9, which meets the condition. Further judgment is made on whether the offset is continuous. For example, the starting offset of period 4 is 14-13=+1, the ending offset is 49-50=-1, the absolute offset value is 2, and the cumulative total offset is 9+2=11. If it continues to exceed the limit, it is confirmed as a continuous sag segment. The start and end times and amplitudes are recorded in the sag event table, and finally the voltage sag event record is obtained.

[0061] Please see Figure 3 The boundary stability determination module includes:

[0062] The offset difference extraction submodule extracts the offset values ​​corresponding to all cycles involved in each event based on the voltage sag event record. It sorts the offset values ​​in each set of five consecutive cycles, identifies the maximum and minimum values ​​in the sequence, and uses the difference as the difference range of the current sequence to obtain the offset difference range set.

[0063] Based on voltage sag event records, all periodic data involved in each event are extracted and arranged chronologically to construct an offset value sequence. The offset value represents the variation of the boundary index of each period. A sliding window structure is formed by defining a sequence unit of five consecutive periods. Periods 1 to 5 constitute the first window, periods 2 to 6 constitute the second window, and so on. Within each five-period window, five corresponding offset values ​​are extracted, sorted by numerical value, and the maximum and minimum values ​​are recorded. The difference between these two values ​​yields the offset difference range corresponding to that window. For example, if the offset value sequence in a certain event is {4, 7, 5, 3, 6, 9, 8}, then the offset values ​​for window 1 are {4, 7, 5, 3, 6}, with a maximum of 7, a minimum of 3, and a difference of 4. Window 2 is {7, 5, 3, 6, 9}, with a maximum of 9, a minimum of 3, and a difference of 6. This process is repeated for all windows to calculate the offset difference range, forming a data sequence of difference ranges corresponding to the time window. A numbered record table is then established, assigning an ID to each sequence and binding it to the starting period to form a traceable structure, as shown in the table below.

[0064] Table 2: Offset Difference Measurement Data

[0065] ;

[0066] As shown in Table 2, the difference range values ​​correspond one-to-one with the window period number, and are mapped to subsequent judgment operations through the number index, ultimately obtaining the offset difference range set.

[0067] The voltage tolerance comparison submodule compares the difference range of each sequence with the voltage fluctuation tolerance threshold based on the offset difference range set. If the difference range is less than the voltage fluctuation tolerance threshold, it is marked as a boundary stable period; otherwise, it is marked as an unstable period, thus obtaining the boundary stability label sequence.

[0068] Based on the set of offset difference ranges, each set of difference values ​​is compared with the set voltage fluctuation tolerance threshold. According to the ±10% requirement in the power industry standard "Power Quality Supply Voltage Deviation", the voltage reference value is set to 230V, corresponding to a tolerance bandwidth of 23V. However, since the unit of offset value is the sampling point index, it needs to be converted to the sampling point change amplitude equal to the tolerance. After conversion, the voltage offset tolerance threshold is set to 3 sampling points. If a difference range is less than or equal to 3, it is marked as stable; otherwise, it is unstable. For example, in Table 5, window 1 has a difference of 4, and window 2 has a difference of 6, both exceeding 3, therefore both are marked as unstable. If a window has a difference of 2, it can be marked as a stable period. Throughout the comparison process, each difference value needs to be binary judged and the corresponding label value 1 or 0 output. Simultaneously, a label sequence table is established and bound to the difference source window number, as shown below:

[0069] Table 3 Boundary Stability Labels:

[0070] ;

[0071] As shown in Table 3, stability labels can be used as the basis for screening stability cycles, and finally the boundary stability label sequence can be obtained.

[0072] The stable period generation submodule extracts the numbers of all periods marked as stable states based on the boundary stability label sequence, establishes an index mapping according to the original order in the whole period sequence, and establishes a unique identifier structure for the period sequence to obtain a list of perturbation stable periods.

[0073] Based on the boundary stability label sequence, extract all window numbers with a label value of 1, and split and filter their corresponding period numbers. For example, if window number 3 in the label sequence is stable, its corresponding period range is period 3 to 7. Then, extract the period numbers within this range, i.e., periods 3, 4, 5, 6, and 7, as a candidate set of stable periods. Then, merge the stable periods in all windows, remove duplicate period numbers, and arrange them according to the original sampling order to form a perturbation stable period sequence. In this process, each period number also needs to be assigned a unique index and recorded in the stable period mapping table for subsequent event aggregation and stable segment location operations. Finally, summarize all labeled period numbers and output them as a stable period result set to form a complete set of periodic stable segments, and output it as a list of perturbation stable periods.

[0074] Please see Figure 4 The aliasing state recognition module includes:

[0075] The disturbance feature extraction submodule extracts voltage disturbance data for the corresponding period based on the list of disturbance stable periods, counts the peak amplitude of the disturbance waveform and the total number of disturbance events in each period, and performs the above operations sequentially for all periods to obtain the disturbance intensity calculation input data;

[0076] Based on the list of stable disturbance cycles, the voltage disturbance curve corresponding to each stable cycle is extracted. The number of disturbances in each cycle is counted, and key disturbance amplitude features, the number of duration points of the disturbance segment, and the disturbance time value are extracted for each disturbance segment. The number of disturbances is obtained by detecting the number of changes in the first-order differential slope of the waveform. The disturbance amplitude is the maximum value of the voltage change. The number of disturbance segments is the number of sampling points between the start and end indices. The disturbance duration is calculated by multiplying the segment length by the single-point time interval (sampling frequency of 5000Hz, corresponding to 0.2ms / point). The disturbance count for cycle number 1 is 3, and the disturbance amplitudes are 6.2V and 5V respectively. The disturbance values ​​are 0.8V and 7.0V, with 12, 10, and 15 disturbance points, and durations of 4.8ms, 4.0ms, and 5.2ms, respectively. Period 2 has 4 disturbances with amplitudes of 7.5V, 6.9V, 5.3V, and 6.1V, corresponding to 10, 8, 9, and 11 disturbance points, and durations of 4.0ms, 3.2ms, 3.6ms, and 4.4ms, respectively. Period 3 contains 2 disturbances with amplitudes of 8.0V and 7.2V, durations of 9 and 10 disturbance points, and durations of 3.6ms and 4.0ms, respectively. The extracted data for each period is summarized in Table 4, providing the input data for disturbance intensity calculation.

[0077] Table 4 Input data table for disturbance intensity calculation:

[0078] ;

[0079] The intensity threshold determination submodule calculates the input data based on the disturbance intensity using the following formula:

[0080] ;

[0081] The perturbation intensity for each cycle is calculated and compared with the total harmonic distortion (THD) threshold. If the intensity exceeds the THD threshold, the cycle is marked as an aliasing cycle with a label value of 1; otherwise, it is marked as 0. A correspondence structure between cycle numbers and the judgment results is established, generating an aliasing cycle label sequence. Indicates the first The disturbance intensity per cycle, in V· , Indicates the first The first in the cycle Peak amplitude of the disturbance (in V). For the first Number of sampling points for each disturbance This is the single-point sampling time interval, in milliseconds (ms) per point, with a set value of 0.2 ms. The duration of the corresponding disturbance (in milliseconds). For the first Number of disturbances within a period, threshold The upper limit of statistical intensity of non-aliasing periodic disturbances in a typical 10kV distribution network sample is derived from 95% of the non-aliasing periodic disturbances.

[0082] Based on the input data for disturbance intensity calculation, the disturbance intensity formula is derived by substituting the number of disturbances within a period, the peak amplitude of each disturbance, the number of disturbance segments, and the duration. ms is the sampling interval. Now, we will calculate the specific disturbance intensity for period 1:

[0083] First disturbance: ;

[0084] Second disturbance: ;

[0085] Third disturbance: ;

[0086] The total intensity of the period 1 disturbance is ;

[0087] Calculation of disturbance intensity for period 2:

[0088] ;

[0089] ;

[0090] The intensity of the period 3 disturbance is:

[0091] ;

[0092] The disturbance intensity value is compared with the total harmonic distortion threshold. If a comparison is made, If it is marked as an aliasing period, the label value is 1; otherwise, it is 0. The results are as follows:

[0093] Table 5. Labeling of Overlapping Periods:

[0094] ;

[0095] Finally, the aliased periodic label sequence is obtained.

[0096] Periodic disturbance intensity is a composite index used to measure the combined impact of all disturbance events within a voltage cycle. Its core significance lies in simultaneously reflecting the amplitude, duration, and frequency of disturbances. It not only considers the maximum change in voltage fluctuations but also introduces the time dimension of fluctuation duration and the frequency of occurrence, thereby constructing a numerical dimension that can describe the "intensity" of disturbances within the cycle. The larger the index, the more severe, concentrated, or numerous the disturbance events are within the cycle. It is a key basis for judging whether the cycle has reached an abnormal power quality level (such as aliasing, voltage distortion, etc.) and can serve as an important reference for subsequent disturbance classification and identification and abnormal area determination.

[0097] The formula first applies to each perturbation event. In the cycle The contributions of each event are calculated separately, and then the effects of all events are summed up to form a complete value of the periodic disturbance intensity. The formula uses the absolute value symbol. This is to unify the handling of the impact of positive and negative disturbances on amplitude, emphasizing that the disturbance intensity is considered only in terms of its absolute magnitude, followed by the number of disturbance duration points. Multiplication reflects the length extension of the disturbance in the time dimension; multiplying by Converted to time units, the normalized representation represents the true duration of the disturbance, with the denominator being the duration of the corresponding disturbance segment. The square root is used to adjust the nonlinear effect of the disturbance response rhythm on the intensity. Physically, the shorter the disturbance time (i.e., the faster the high-frequency disturbance), the stronger its impact on the system. Therefore, the square root is used... The reciprocal structure introduces a time compression correction term, forming an amplitude-duration co-function of the disturbance intensity. Finally, by summing all disturbance segments, the total amount of the cumulative effect of the disturbance energy within the period is obtained, which characterizes the overall intensity of the disturbance in that period.

[0098] The aliasing segment location submodule extracts the cycle number of all aliasing cycles based on the aliasing cycle label sequence, performs segment location in the original voltage disturbance signal data, extracts the start point index and end point index of the disturbance curve in each aliasing cycle, records the start and end index values ​​of each disturbance interval and forms an aliasing segment index set, and obtains the aliasing cycle identification result.

[0099] Based on the aliasing cycle label sequence, cycle 2 is selected as the aliasing cycle with a label value of 1. The start and end point indices of its disturbance segment are located from the original disturbance index data. The four disturbance segments correspond to sampling points 20–30, 45–53, 60–69, and 76–87, respectively. The voltage start and end point differences are calculated for each of these four disturbance segments. The disturbance amplitude filtering threshold is set to 3V. If the voltage change value of a certain segment is greater than this threshold, it is determined to be a valid disturbance segment. After detection, the first three segments meet the condition, while the fourth segment, with a change amplitude of 2.4V, does not meet the filtering requirements and is removed. Finally, there are three valid segments in aliasing cycle 2: 20–30, 45–53, and 60–69. The output is the aliasing cycle identification result.

[0100] Please see Figure 5 The periodic power calculation module includes:

[0101] The power sample extraction submodule locates the complete time interval of the period in the original signal sequence based on the sampling start and end point index of the corresponding period in the aliasing period identification result, extracts the voltage and current values ​​at each time point, and synchronizes them according to the time index. Each time point has a set of corresponding voltage-current data pairs, and the electrical parameter synchronization sequence is obtained by sorting and processing.

[0102] Based on the start and end index positions of period number 1 in the aliasing period identification results, all sampling points within the corresponding time period in the voltage and current sampling data are located. The voltage and current values ​​of each sampling point are extracted, and the sampling interval of each point is confirmed to be 0.2ms based on the sampling frequency of 5000Hz. The sampling data are aligned one by one according to the time point order to form a one-to-one voltage-current sampling pair. Furthermore, adjacent difference operations are performed on the voltage sampling point sequence to obtain the perturbation amplitude of each point. For example, the voltage difference between point 2 and point 1 is V, as the disturbance amplitude value at point 2, is used to obtain the disturbance amplitude at point 3 using the same method. V; then backtrack 10 points to construct the current disturbance frequency count for the current point. For example, there are 2 sudden current disturbance events near point 2 and 1 event near point 3; finally, the cumulative time of each point relative to the start point of the cycle is calculated. For example, point 1 is 0.2ms, point 2 is 0.4ms, and point 3 is 0.6ms. The parameters of the above sampling points are set into a periodic sample table, as shown below:

[0103] Table 6. Periodic Sampling Parameters:

[0104] ;

[0105] As shown in Table 6, a sampling point-parameter alignment structure has been constructed with period number 1 to obtain the electrical parameter synchronization sequence.

[0106] The instantaneous power accumulation submodule performs a product operation based on the voltage and current values ​​at each time point in the electrical parameter synchronization sequence to generate instantaneous power data for each point, using the following formula:

[0107] ;

[0108] The total energy consumption per unit cycle is calculated, and integral data of cycle energy consumption is generated. Indicates the first Total energy consumption per cycle For period Number of internal sampling points and They represent the first Voltage and current values ​​at each sampling point It is the absolute value of the difference between the current voltage and the previous sampling voltage, in V, and is dimensionless after normalization. It represents the current disturbance count within a fixed window (e.g., 10 points) preceding the current sampling point, and is dimensionless after normalization; This represents the cumulative time from the current sampling point to the start of the period, in milliseconds. The sampling time interval is in milliseconds (ms).

[0109] Based on the parameters of each sampling point in the electrical parameter synchronization sequence shown in Table 6, the perturbation weighted power accumulation and integration are performed point by point according to the formula, and specific calculations are carried out for sampling points 2 and 3 respectively:

[0110] Sampling point 2:

[0111] ;

[0112] ;

[0113] ;

[0114] ;

[0115] ;

[0116] Sampling point 3:

[0117] ;

[0118] ;

[0119] ;

[0120] ;

[0121] ;

[0122] Ignoring the first point without disturbance, and only summing the values ​​at points 2 and 3, we obtain the total energy consumption integral for period number 1:

[0123] ;

[0124] The calculation result is the energy consumption numerical result within cycle number 1, generating cycle energy consumption integral data.

[0125] The total energy consumption of a cycle refers to the energy quantification result obtained by accumulating and integrating the instantaneous power values ​​of all sampling points on the time axis within a complete voltage cycle. It reflects the effective energy output generated by the interaction between voltage and current within the cycle. If there are disturbances, current surges, or voltage distortions in the cycle, these factors will be reflected in the total energy consumption through instantaneous power changes. Therefore, this value not only represents the actual power transmission capacity within the cycle, but also introduces dynamic weight adjustments for disturbance amplitude and frequency into the structure, enabling it to reveal the relationship between electrical parameter disturbance characteristics and energy response. The larger the total energy consumption of a cycle, the denser the power fluctuation amplitude and interference frequency within the cycle. It is an important basic quantity for analyzing the energy behavior of aliasing cycles, locating abnormal cycles, and performing power quality classification judgments.

[0126] The formula first passes through Get the The instantaneous power value at each sampling point represents the power intensity generated by the combined effect of voltage and current at that moment, followed by the disturbance amplitude. With disturbance frequency The summation reflects the combined effect of amplitude and frequency disturbances on the sampling point, taking into account the weakening trend of the system response strength due to the propagation of disturbances over time, using the cumulative time since the start of the period at the sampling point. Construct the denominator and use the square root operation to construct the nonlinear time decay factor. This is used to suppress the amplified effect of disturbances in the later part of the cycle on the overall power assessment, and then multiplied by the sampling interval. The results are mapped to time-domain integration units, and the total interference power weights of all sampling points within the period are summed to form the energy disturbance assessment result. The overall structure integrates instantaneous power of electrical parameters, disturbance characteristic intensity, and time-domain correction factor, and has the ability to analyze dynamic response processes.

[0127] The cycle energy consumption generation submodule reconstructs the energy consumption value sequence within all cycles according to the cycle energy consumption integral data and the cycle number order. It uses the cycle number as the index and the cycle energy consumption value as the content to integrate them into a two-dimensional cycle energy consumption array structure and outputs the cycle energy consumption calculation data.

[0128] Based on the cycle energy consumption integral data, the integral value of 2276.25 corresponding to cycle number 1 is recorded. The sample regions of cycle numbers 2 and 3 are then processed, and their energy consumption integral values ​​of 2211.30 and 2137.88 are obtained respectively. The three cycle numbers and energy consumption values ​​are combined to construct a cycle-power integral table, summarized as follows:

[0129] Table 7: Cycle Energy Consumption Calculation Data

[0130] ;

[0131] As shown in Table 7, the energy consumption integral values ​​corresponding to cycles 1 to 3 have been calculated and organized, and the output is the cycle energy consumption calculation data.

[0132] Please see Figure 6 The energy consumption attribution determination module includes:

[0133] The feature acquisition submodule extracts the sampling sequence within the corresponding period based on the periodic energy consumption calculation data. It extracts the voltage values ​​at the start and end points of the disturbance in the corresponding sampling point sequence, calculates the difference as the boundary difference, constructs a continuous difference sequence for all voltage sampling points in the disturbance segment, calculates the mean of the sequence as the voltage slope, and counts the maximum amplitude of all current sampling points in the disturbance segment as the current amplitude. It then integrates these to obtain the disturbance feature vector set.

[0134] Based on the calculated periodic energy consumption data, the start and end positions of each disturbance source within the periodic sampling interval are extracted sequentially. For each disturbance source segment, the voltage values ​​at the start and end sampling points are recorded. For example, the start voltage of disturbance source 1 is 222.4V, the end voltage is 225.1V, and the boundary difference is 2.7V; the start voltage of disturbance source 2 is 219.7V, the end voltage is 221.5V, and the boundary difference is 1.8V. Then, a first-order difference is calculated for all voltage sampling points within each disturbance interval. The calculations were performed by taking the average of the differences between adjacent voltages. The average difference for disturbance source 1 was 0.86 V / ms, and for disturbance source 2 it was 0.60 V / ms, which were used as their voltage slopes. At the same time, the maximum values ​​were extracted from the current sequences of each disturbance segment. The maximum current value for disturbance source 1 was 5.3 A, and for disturbance source 2 it was 4.9 A. The boundary differences, voltage slopes, and current amplitudes of the above three characteristic parameters were combined to form the three-dimensional parameter feature vectors of the corresponding disturbance sources, as shown in Table 8, thus obtaining the disturbance feature vector set.

[0135] Table 8. Characteristic parameters of disturbance sources:

[0136] ;

[0137] As shown in Table 8, a set of voltage, current and dynamic change characteristic parameters corresponding to the disturbance source has been constructed as a set of three-dimensional disturbance feature inputs.

[0138] The similarity vector comparison submodule compares each vector with the known typical load feature vector in the stable period according to the disturbance feature vector set, calculates the three-dimensional distance as the similarity index, sorts them in ascending order of distance value, selects the target load number with the smallest distance as the matching load, and obtains the disturbance-load similarity mapping set.

[0139] Based on the three-dimensional parameter feature vectors of disturbance source 1 and disturbance source 2 in Table 8, their vector differences are compared with the registered load vectors in the typical stable load feature sample library. The feature vector (2.7, 0.86, 5.3) of disturbance source 1 is aligned with the stable sample numbers A (2.6, 0.85, 5.4) and B (1.9, 0.72, 4.8) in sequence. The coordinate difference is calculated according to the three-dimensional features, and the square root of the difference is obtained to get the three-dimensional distance between disturbance source 1 and samples A and B. Similarly, the distance between disturbance source 2 and samples A and B is obtained, which are 0.173 for sample A and 0.153 for sample B. The corresponding distances for disturbance source 2 are 0.852 for sample A and 0.078 for sample B. After sorting, disturbance source 1 is closest to sample B and disturbance source 2 is closest to sample B. Therefore, disturbance source 1 and 2 are matched to load B respectively to obtain the disturbance-load similarity mapping set.

[0140] Table 9: Similarity Matching Results

[0141] ;

[0142] As shown in Table 9, both disturbance source 1 and disturbance source 2 are closest to the typical characteristic vector of load B, and a mapping relationship is established.

[0143] The energy consumption allocation generation submodule divides the data equally into each disturbance source item based on the matching relationship between each disturbance source and the corresponding load number in the disturbance-load similarity mapping set, combined with the energy consumption value corresponding to the periodic energy consumption calculation data. It then summarizes the allocated energy consumption of each disturbance source under the corresponding load number to obtain the energy consumption identification result of the substation aliasing scenario.

[0144] Combining the information of both disturbance source 1 and 2 corresponding to load B in the disturbance-load similarity mapping set, and allocating the periodic energy consumption value within the period, assuming the periodic energy consumption value is 2300J, since there are 2 disturbance sources within the period, the energy consumption is divided equally into 1150J to the two disturbance items, and then the energy consumption values ​​of disturbance source 1 and 2, 1150J, are allocated to their corresponding load B respectively. Load B accumulates an energy consumption of 2300J, as shown in Table 10, and the energy consumption identification results of the substation aliasing scenario are obtained.

[0145] Table 10 Energy Consumption Allocation Results:

[0146] ;

[0147] As shown in Table 10, load B is the closest to the eigenvector of all disturbance sources, and thus the total energy consumption value for the current cycle is obtained.

[0148] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A power consumption recognition system based on a variable aliasing scenario, characterized in that, The system comprises: The disturbance boundary extraction module acquires periodic voltage signal data under the parallel operation state of the substation load, collects the start and end point index positions of the voltage fluctuation region in each period, calculates the phase shift, counts and records the absolute value of the shift, and generates a voltage sag event record; The boundary stability determination module calculates the difference range of the maximum and minimum shift values in any five consecutive periods according to the voltage sag event record, and if the difference range is less than the voltage fluctuation tolerance, it is determined to be a stable boundary and the period number is recorded, and a disturbance stable period list is generated; The aliasing state recognition module extracts the disturbance peak amplitude and the number of disturbances in the corresponding period based on the disturbance stable period list, calculates the period disturbance intensity, and if it is greater than the total harmonic distortion threshold, it is marked as an aliasing period, and the start and end index intervals of all overlapping sections are located, and an aliasing period identification result is generated; The period power calculation module extracts the point-by-point sampling values of voltage and current in the corresponding section according to the aliasing period identification result, performs the product of the corresponding time points to form a power sequence, and integrates the sequence to obtain the total energy value in a unit period, and outputs the period energy consumption calculation data; The energy consumption attribution determination module extracts the boundary difference value, voltage slope and current amplitude of all disturbance sources in the same period according to the period energy consumption calculation data, constructs a disturbance feature vector, and compares it with the typical feature vector in the stable period, selects the load unit with high similarity features and allocates the corresponding power, and generates a substation aliasing scene energy consumption recognition result.

2. The energy consumption identification system based on the variable power aliasing scenario of claim 1, wherein, The voltage sag event record includes the absolute value of the shift, the phase shift index, the event start and end marker, the disturbance stable period list includes the period number set, the tolerance comparison result, the difference range evaluation index, the aliasing period identification result includes the aliasing period number, the overlapping section index, and the total harmonic overrun marker, and the period energy consumption calculation data includes the power time sequence, the period cumulative energy value, and the voltage and current sampling pair. 3.The power consumption identification system based on the variable power aliasing scenario of claim 1, wherein, The disturbance boundary extraction module comprises: The period voltage acquisition submodule acquires the voltage signal data of each period under the parallel operation state of the substation load, draws the voltage fluctuation curve in each period, extracts the voltage extreme point, voltage fluctuation edge point and corresponding timestamp information in each period, and generates the period voltage signal fluctuation value; The start and end index positioning submodule detects the voltage fluctuation region boundary in each period based on the period voltage signal fluctuation value, identifies the index positions of the voltage fluctuation start and end points through the set voltage fluctuation determination threshold, judges whether the fluctuation region has stability characteristics according to the boundary change of adjacent periods, and acquires the voltage fluctuation boundary index sequence; The phase shift statistical submodule calculates the difference value and absolute value of the start and end index positions in the consecutive periods according to the voltage fluctuation boundary index sequence and the voltage fluctuation measurement requirements, summarizes the shift degree of the boundary index in the period sequence, and acquires the voltage sag event record.

4. The energy consumption identification system based on variable power aliasing scenarios of claim 1, wherein, The boundary stability determination module comprises: The offset difference extraction submodule extracts, based on the voltage sag event record, offset values corresponding to all cycles involved in each event, sorts the offset values in each sequence of five consecutive cycles, identifies the maximum and minimum values in the sequence, takes the difference as the difference range of the current sequence, and obtains a set of offset difference ranges; The voltage tolerance comparison submodule compares each sequence difference range with a voltage fluctuation tolerance threshold according to the set of offset difference ranges, and if the difference range is less than the voltage fluctuation tolerance threshold, marks it as a boundary stable cycle, otherwise as an unstable cycle, to obtain a boundary stability label sequence; The stable cycle generation submodule extracts, according to the boundary stability label sequence, all cycles marked as stable, establishes an index mapping according to the original order in the full cycle sequence, and establishes a cycle sequence unique identification structure to obtain a list of disturbance stable cycles.

5. The energy consumption identification system based on variable power aliasing scenarios of claim 1, wherein, The aliasing state recognition module includes: The disturbance feature extraction submodule extracts, based on the list of disturbance stable cycles, voltage disturbance data corresponding to the cycles, counts the peak amplitude of the disturbance waveform and the total number of disturbance events in each cycle, and sequentially performs the above operations on all cycles to obtain disturbance intensity calculation input data; The intensity threshold judgment submodule calculates the disturbance intensity of each cycle according to the disturbance intensity calculation input data, and judges each cycle according to the total harmonic distortion threshold. If the disturbance intensity is greater than the total harmonic distortion threshold, the cycle is marked as an aliasing cycle, and a corresponding structure of cycle number and judgment result is established to generate an aliasing cycle label sequence; The aliasing section positioning submodule extracts the cycle numbers of all aliasing cycles according to the aliasing cycle label sequence, positions the sections in the original voltage disturbance signal data, extracts the start point index and end point index of the disturbance curve in each aliasing cycle, records the start and end index values of each disturbance interval, and forms an aliasing section index set to obtain an aliasing cycle identification result.

6. The energy consumption identification system based on variable power aliasing scenarios of claim 1, wherein, The cycle power calculation module includes: The power sample extraction submodule positions the complete time interval of the cycle in the original signal sequence based on the sampling start and end point index of the corresponding cycle in the aliasing cycle identification result, extracts the voltage and current values at each time point, and synchronously aligns them according to the time index. Each time point has a corresponding voltage-current data pair. The voltage-current synchronous sequence is obtained; The instantaneous power accumulation submodule performs a product operation to form instantaneous power data at each point based on the voltage and current values at each time point in the voltage-current synchronous sequence, calculates the total energy consumption per cycle, and generates cycle energy consumption integral data; The cycle energy generation submodule reconstructs the energy consumption value sequence in all cycles according to the cycle energy consumption integral data, takes the cycle number as the index and the cycle energy consumption value as the content, and integrates them into a two-dimensional cycle energy array structure to output cycle energy consumption calculation data.

7. The energy consumption identification system based on variable power aliasing scenarios of claim 1, wherein, The substation aliasing scenario energy consumption recognition result includes a disturbance feature vector set, a similarity comparison matrix, and load unit energy consumption allocation information.

8. The energy consumption identification system based on variable power aliasing scenarios of claim 1, wherein, The energy consumption attribution determination module includes: The feature acquisition submodule extracts a sampling sequence in a corresponding period based on the period energy consumption calculation data, extracts voltage values of a disturbance starting point and an ending point in the corresponding sampling point sequence, calculates a difference value as a boundary difference value, constructs a continuous difference value sequence for all voltage sampling points in the disturbance section, calculates a sequence mean value as a voltage slope, and calculates a maximum amplitude of all current sampling points in the disturbance section as a current amplitude. The disturbance feature vector set is integrated and acquired; The similar vector comparison submodule compares each vector with a known typical load feature vector in a stable period according to the disturbance feature vector set, calculates a three-dimensional distance as a similarity index, sorts the distance values from small to large, selects a target load number with the smallest distance as a matching load, and obtains a disturbance-load similarity mapping set. The energy consumption allocation generation submodule divides the energy consumption value corresponding to the period energy consumption calculation data into each disturbance source item in equal parts according to the matching relationship between each disturbance source and the corresponding load number in the disturbance-load similarity mapping set, and summarizes the allocated energy consumption of each disturbance source to the corresponding load number to obtain a power transformation aliasing scene energy consumption identification result.

Images