Power distribution network electrical variable measurement and state analysis method

Abstract

The invention relates to the field of electrical variable measurement, in particular to a power distribution network electrical variable measurement and state analysis method, which comprises the following steps of: obtaining an original current value, an original voltage value and an original power value at each moment; calculating a data matrix of each sub-segment; calculating a decoupling characteristic matrix according to the coupling component; constructing a state analysis model, and training the state analysis model to obtain a trained state analysis model; new current original values at all moments, new voltage original values at all moments and new power original values at all moments are obtained through collected new power distribution network electrical variables, the new current original values, the new voltage original values at all moments and the new power original values at all moments are input into the trained state analysis model after being processed, all state probability vectors are obtained, and the state category corresponding to the maximum probability value is taken as the state analysis result of the sub-segment. The existing power distribution network electrical variable measurement and state analysis method has the problem of low accuracy of state analysis of the power distribution network. The method provided by the invention has high accuracy of state analysis of the power distribution network.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of measuring electrical variables, in particular to a method for measuring and analyzing the state of electrical variables in a power distribution network. BACKGROUND

[0002] A power distribution network refers to a power network that receives electrical energy from a transmission and distribution network or a regional power plant and distributes it to various users through power distribution facilities. The power distribution network is affected by various factors such as load fluctuations, line switching, fault impacts, and electromagnetic transient disturbances during operation, resulting in significant non-stationary characteristics of electrical variables such as current, voltage, and power.

[0003] To ensure the safe and stable operation of the power distribution network and improve efficient power supply for various users, it is necessary to measure the electrical variables (current, voltage) of the power distribution network, analyze the operating state of the power distribution network based on the measured electrical variables, and the operating state includes normal state, short circuit fault, overvoltage, undervoltage, harmonic interference, and transient oscillation.

[0004] The existing method for measuring and analyzing the state of electrical variables in a power distribution network usually uses a fixed sampling rate when measuring electrical variables, so it cannot effectively adapt to changes in transient events in the power distribution network, resulting in insufficient timeliness of data collection at critical moments and difficulty in fully capturing instantaneous changes. When analyzing electrical variables, first, the method lacks pertinence when processing collected data, often relying only on simple filtering techniques for noise suppression, which cannot effectively remove complex interference in power distribution network signals, affecting the accuracy and reliability of electrical variables. Second, the conventional signal segmentation method uses a fixed length sliding window, which cannot flexibly adapt to complex events in the power distribution network, so it often cannot retain important electromagnetic transient information, leading to distortion of feature extraction. Finally, current multivariate feature extraction often directly concatenates multiple electrical variables, ignoring the nonlinear coupling relationship between them, which can easily lead to redundant features, reducing the performance and recognition ability of the state analysis model. These factors can affect the performance of the final state analysis model, thereby affecting the accuracy of the state analysis of the power distribution network.

[0005] Therefore, the existing method for measuring and analyzing the state of electrical variables in a power distribution network has the problem of low accuracy in state analysis of the power distribution network. SUMMARY

[0006] The technical problem to be solved by the present application is to overcome the shortcomings of the prior art and provide a method for measuring and analyzing the state of electrical variables in a power distribution network with high accuracy.

[0007] To solve the above technical problems, the present application provides a method for measuring and analyzing the state of electrical variables in a power distribution network, comprising: S1. Measure power distribution network electric variable data based on a sampling time sequence, calculate instantaneous power, calculate current original value at each time, voltage original value at each time and power original value at each time; S2. Calculate current joint mutation index at each time, calculate a segmented boundary set based on the current joint mutation index at each time, divide the current original value at each time, the voltage original value at each time and the power original value at each time into a plurality of subsegments based on the boundary points of the segmented boundary set, calculate a data matrix of each subsegment, the data matrix of each subsegment including a current sequence of each subsegment, a voltage sequence of each subsegment, and a power sequence of each subsegment; S3. Calculate coupling components based on the current sequence of each subsegment, the voltage sequence of each subsegment, and the power sequence of each subsegment, and calculate a decoupling feature matrix according to the coupling components; S4. Construct a state analysis model, and train the state analysis model to obtain a trained state analysis model; S401. Calculate a dual-channel fusion feature matrix; S402. Calculate a state probability vector of each subsegment and a total loss function; S403. Based on the total loss function, repeat S401-S402 to train and iteratively optimize the state analysis model; S5. Obtain new current original value at each time, new voltage original value at each time and new power original value at each time from new power distribution network electric variables collected by S1, input the new current original value at each time, the new voltage original value at each time and the new power original value at each time into the trained state analysis model after executing S2-S3, obtain each state probability vector, and take the state category corresponding to the maximum probability value as the state analysis result of the subsegment.

[0008] As a further improvement of the present application: the steps of S1 include: S101. For the non-stationary characteristics of the power distribution network signal, a variable step dynamic adjustment sampling interval is used for sampling, and a sampling time sequence is calculated; S102. Respectively apply Kalman filtering to the sampled current signal at each time and the voltage signal at each time to reduce noise, and obtain filtered current value at each time and filtered voltage value at each time; S103. Convert the filtered current value at each time and the filtered voltage value at each time into digital signals through an electric current quantization function, and obtain current original value at each time and voltage original value at each time; S104. Calculate power original value at each time based on current original value at each time and voltage original value at each time.

[0009] Preferably, the sampling time sequence is composed of n sampling times; the calculation formula of the mth sampling time is: , wherein,​ This is a time index, which defaults to the current time point; For the first Each sampling time; For sampling point index, , This represents the total number of sampling points; For the first Each sampling time; For the first Each sampling interval is determined through adaptive dynamic adjustment. No. sampling interval The calculation formula is: , In the formula, The reference sampling interval; It is the L2 norm, also known as the Euclidean norm, and is used in... Represents the input object; specifically, the input object here refers to... ; for Simulated current signals are constantly measured using a current transformer. for The simulated voltage signal is constantly measured by a voltage transformer. For the first One sampling time; for Simulated current signals are constantly measured using a current transformer. for The simulated voltage signal is constantly measured by a voltage transformer. For the first Each sampling time.

[0010] As a further improvement of the present invention: step S2 includes: S201. Calculate the joint current abrupt change index at each time step; S202. Calculate the set of segmented boundaries based on the joint current abrupt change index at each time point; S203. Based on the boundary points of the segment boundary set, divide the original values ​​of current, voltage, and power at each time moment into sub-segments to obtain the data matrix of each sub-segment composed of the current sequence, voltage sequence, and power sequence.

[0011] Preferably, the joint current abrupt change index at each time point, i.e., the joint current abrupt change index at time t. The calculation formula is: , In the formula, The standard deviation of the signal within the window; For local window width; From Time's up The definite integral at time step; The variable is the integral variable, representing the time index within the local time window; for The original value of the current at that moment; The average value of the signal within the window; The standard Gaussian kernel function; Frequency index; Minimum frequency; Maximum frequency; The short-time Fourier transform of the signal within the window at time t is at the t-th The power spectrum percentage at a given frequency point; It is a logarithmic function, with the default base being the natural constant.

[0012] As a further improvement of the present invention: step S3 includes: S301. Based on the current sequence, voltage sequence, and power sequence of each sub-segment, calculate the current sequence tensor, voltage sequence tensor, and power-current physical relationship constraint tensor of each sub-segment to obtain the coupling components. S302. Apply sparse constraints to the coupled tensor using Tucker decomposition to obtain the core tensor and the factor matrix representing the coupling relationship of variables, the time evolution mode and the physical constraint mode. Define the decoupling feature matrix based on the factor matrix.

[0013] As a further improvement of the present invention: step S401 includes: S4011. Calculate the spectral convolution output features of each column of the decoupled feature matrix to obtain the spectral convolution output matrix; S4012. Calculate the temporal convolution output features of each column of the decoupled feature matrix to obtain the temporal convolution output matrix; S4013. Calculate a dual-channel fusion feature matrix with spatiotemporal joint representation capability based on the spectral convolution output matrix and the temporal convolution output matrix.

[0014] Preferably, the dual-channel fused feature matrix The calculation formula is: , In the formula, For the fusion weight matrix; The output matrix is ​​the spectral convolution matrix; For splicing operations; This is the output matrix of the temporal convolution; For Hadamah accumulation; For the Softmax function; This is a pattern 3 factor matrix; For dimension A vector of all 1s.

[0015] As a further improvement of the present invention: step S402 includes: S4021. Calculate the enhanced feature matrix of each sub-segment based on the dual-channel fusion feature matrix; S4022. Calculate the physical constraint mode difference scalar between each sub-segment and its predecessor based on the enhanced feature matrix of each sub-segment; S4023. Calculate the state probability vector of each sub-segment; S4024. Calculate the total loss function based on the physical constraint modal difference scalar between each sub-segment and its predecessor, and the state probability vector of each sub-segment.

[0016] Preferably, the total loss function The calculation formula is: , In the formula, k is the sub-segment index. ; This represents the total number of sub-segments; For the first The true state label of each sub-segment; The power sequence of the k-th sub-segment; For the first The state probability vector of each sub-segment; This is the scaling factor; For the first The physical constraint modal difference scalar between each sub-segment and the previous sub-segment.

[0017] The beneficial effects of the present invention are as follows: The method for measuring and analyzing the electrical variables of a distribution network provided by the present invention has high accuracy in analyzing the state of the distribution network.

[0018] First, S1's multi-rate adaptive sampling method can dynamically adjust the sampling rate to ensure that more data is collected when the distribution network signal changes drastically, thereby improving the sampling accuracy and time synchronization of the signal and overcoming the limitations of traditional fixed sampling rate in capturing transient events. By combining Kalman filtering for noise suppression, the signal-to-noise ratio of current and voltage measurement signals is significantly improved, effectively eliminating the influence of noise interference in traditional measurement methods and ensuring the accuracy of the data. Furthermore, S2 employs a joint mutation index and an adaptive signal segmentation method. By combining local kurtosis with spectral entropy, it can accurately detect structural mutation points in the signal, avoiding feature loss or distortion that may occur with traditional equal-length segmentation methods. Finally, S3 constructs a third-order coupling tensor embedded with power-current constraints in the decoupling of multivariate coupling features, and extracts decoupled features through Tucker decomposition, eliminating the redundant coupling among current, voltage and power, and improving the recognition ability of power distribution network state analysis; In summary, by optimizing the data set used to train the state analysis model, optimizing the data set preprocessing method, and optimizing the state analysis model itself, the performance of the state analysis model is improved, thereby making the state analysis of the power distribution network more accurate. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 is a schematic diagram of the principle of the present application; Figure 2 is a current data graph collected by a conventional fixed sampling method at fixed time intervals; Figure 3 is a current data graph collected by the adaptive sampling method of the present technology; Figure 4 is a display graph of the original current signal; Figure 5 is a display graph of joint mutation index detection; Figure 6 is a comparison graph of classification accuracy of different methods under various power grid states. DETAILED DESCRIPTION

[0020] The specific embodiments of the present application will be further described in detail below with reference to the accompanying drawings.

[0021] As shown in Figure 1 , the method for measuring and analyzing the state of the power distribution network provided by the present application comprises: S1. Measure the power distribution network electric variable data based on the sampling time sequence, calculate the instantaneous power, and calculate the current original value at each time, the voltage original value at each time, and the power original value at each time; The power distribution network electric variable data includes the analog current signal at each time measured by the current transformer and the analog voltage signal at each time measured by the voltage transformer S101. In view of the non-stationary characteristics of the power distribution network signal, a variable step dynamic sampling interval adjustment method is used for sampling, the sampling rate is automatically increased when the signal changes sharply, and the sampling rate is reduced when it is stable, so as to balance the data quantity and resolution, and the sampling time sequence is calculated; The sampling time sequence is composed of n sampling times; the calculation formula of the i-th sampling time is as follows: , In the formula, is the time index, which corresponds to the current time point by default; ​the first sampling time point in the sequence of sampling time points; the first sampling time point in the sequence of sampling time points; the total number of sampling points; , the total number of sampling points; the first sampling time point in the sequence of sampling time points; the first sampling time point in the sequence of sampling time points; the first sampling interval, determined by adaptive dynamic adjustment; the first sampling interval the first sampling interval the first sampling interval the first sampling interval , wherein, the reference sampling interval, which can be set to 0.001 seconds; the L2 norm, i.e., the Euclidean norm, represented by the input object, which specifically refers to ; the analog current signal at the first sampling time point; the analog voltage signal at the first sampling time point; the analog current signal at the first sampling time point; the analog voltage signal at the first sampling time point; the first sampling time point in the sequence of sampling time points; the analog current signal at the first sampling time point; the analog voltage signal at the first sampling time point; the analog current signal at the first sampling time point; the analog voltage signal at the first sampling time point; the first sampling time point in the sequence of sampling time points; S102. Respectively, the analog current signal and the analog voltage signal at each sampling time point are subjected to Kalman filtering for noise reduction to suppress measurement noise and interference, improve signal-to-noise ratio, and obtain filtered current values and filtered voltage values at each time point; the filtered current value at the first sampling time point the filtered current value at the first sampling time point , wherein, the Kalman gain of the current at the first sampling time point, calculated by the Kalman filtering algorithm; the analog current signal at the first sampling time point; the filtered current value at the first sampling time point; the filtered current value at the first sampling time point; the filtered voltage value at the first sampling time point; the filtered voltage value at the first sampling time point; the filtered voltage value at the first sampling time point the filtered voltage value at the first sampling time point ​​, In the formula, is Kalman gain of the voltage at the moment, which is calculated by a Kalman filtering algorithm; is an analog voltage signal at the moment; is a filtered voltage value at the moment.

[0022] S103. The filtered current value at each moment and the filtered voltage value at each moment are converted into digital signals through a current quantization function to obtain the current original value at each moment and the voltage original value at each moment, so as to ensure that the current and voltage values are aligned at the same time point; The current original value at each moment represents a measured current original value, The current original value at each moment The calculation formula is: , In the formula, is a current quantization function for mapping an analog value to a digital value; is a filtered current value at the moment; The input of the current quantization function is defined as , then The calculation formula is: , In the formula, is a rounding function; is the minimum value of the current, that is, the lower limit of the quantization range, which is specifically set according to the measurement range of the current transformer or the minimum value of historical data, such as-1000A; is the maximum value of the current, that is, the upper limit of the quantization range, which is specifically set according to the measurement range of the current transformer or the maximum value of historical data, such as 1000A; is the number of quantization bits, which is set to 16 by default, and is used for the number of bits of the digital value in analog-to-digital conversion, which determines the resolution of quantization, that is, the number of discrete levels of the digital value; The voltage original value at each moment represents a measured voltage original value, The voltage original value at each moment The calculation formula is: , In the formula, is a voltage quantization function for mapping an analog value to a digital value; The input of the voltage quantization function is defined as , The calculation formula is: , , is the minimum value of the voltage, i.e. the lower limit of the quantization range, which is set according to the minimum value of the measurement range or historical data of the voltage transformer, for example, -35 kV; is the maximum value of the voltage, i.e. the upper limit of the quantization range, which is set according to the maximum value of the measurement range or historical data of the voltage transformer, for example, 35 kV; S104. Calculate the power raw value at each time based on the current raw value at each time and the voltage raw value at each time using the instantaneous power theory; The power raw value at each time represents the measured power raw value, The power raw value at each time The calculation formula of the power raw value at each time is: , , is a cosine function; is is the phase difference between the current and voltage at each time, which is calculated by Hilbert transform; The phase difference between the current and voltage at each time The calculation formula of the phase difference between the current and voltage at each time is: , , is the phase angle; is Hilbert transform; is a conjugate complex number; The power raw value at each time is calculated based on the current raw value at each time and the voltage raw value at each time using the instantaneous power theory. Figure 2 、 Figure 3It can be observed that in regions where the signal waveform changes rapidly, the sampling point distribution of conventional fixed sampling methods is relatively sparse, which may not be able to fully capture the detailed features of transient events. However, the adaptive sampling method of this technology significantly increases the sampling point density in these key regions, forming a denser sampling distribution. The dynamic adjustment capability enables the adaptive sampling method of this technology to record abrupt changes and transient processes in the signal more accurately while ensuring data acquisition efficiency.

[0023] S2. Calculate the joint current mutation index at each time moment, calculate the segment boundary set based on the joint current mutation index at each time moment, and divide the original current value, the original voltage value, and the original power value at each time moment into multiple sub-segments based on the boundary points of the segment boundary set. Calculate the data matrix of each sub-segment. The data matrix of each sub-segment includes the current sequence, the voltage sequence, and the power sequence of each sub-segment. S201. Calculate the joint current abrupt change index at each time step; The kurtosis quantification time-domain abrupt change intensity of the signal within a local window and the spectral entropy quantification frequency-domain complexity based on the power spectrum ratio are weighted and fused through a sliding Gaussian window to form a current joint abrupt change index, thereby achieving comprehensive quantification of nonstationarity intensity. The joint current abrupt change index at each time step, i.e., the joint current abrupt change index at time t, characterizes the nonstationarity of the signal in both the time and frequency domains. A larger value indicates a more pronounced abrupt change in the signal structure. The calculation formula is: , In the formula, The standard deviation of the signal within the window represents local fluctuations and is obtained by calculating the standard deviation of the signal within the local window. The local window width is preferably set to 256 points, which is used to define the local calculation interval. From Time's up The definite integral at time step; for The original value of the current at that moment; The variable is the integral variable, representing the time index of the change within the local time window; The mean of the signal within the window represents the local benchmark and is obtained by calculating the mean of the signal within the local window. For standard Gaussian kernel function, Characterized by a standard Gaussian kernel function that weights the signals within a local window, centered at time t, with the weights varying accordingly. The decay occurs as time moves further away from time t; Frequency index; The minimum frequency can be set to 0.1Hz to cover low-frequency components; The maximum frequency can be set to 2kHz to cover high-frequency components; The short-time Fourier transform of the signal within the window at time t is at the t-th The power spectral density of a frequency point is calculated using a short-time Fourier transform and is used to quantize the frequency complexity of the signal. It is a logarithmic function, with the default base being the natural constant; In the above formula, The term characterizes local kurtosis, quantifies the intensity of abrupt changes in the signal in the time domain, and is used to detect impulse or disturbance events. The term characterizes the local spectral entropy and quantifies the complexity of the signal in the frequency domain. It is used to detect spectral changes. The two complement each other to more comprehensively detect abrupt changes in non-stationary signals, avoid false detections by a single indicator, preserve the complete electromagnetic transient process, and improve segmentation accuracy.

[0024] The short-time Fourier transform of the signal within the window at time t is at the t Power spectral density at frequency points The calculation formula is: , In the formula, for The short-time Fourier transform of the signal within the time window is at the 1st The power spectral density value at a frequency point is obtained by performing a short-time Fourier transform on the signal within a local window; S202. Calculate the set of segmented boundaries based on the joint current mutation index at each time point to achieve adaptive boundary determination; Multi-scale wavelet mode maxima detection is performed on the joint current abrupt change index at various time points. Extreme points with curvature exceeding a dynamic threshold are screened to determine the locations of signal structure abrupt changes, resulting in an adaptive set of segmented boundaries. This set contains all detected abrupt change points, corresponding to time points where signal characteristics undergo significant changes, and exhibits both adaptability and multi-scale characteristics. The calculation formula is: , In the formula, The set of all time points that satisfy the conditions; The joint current mutation index at time t The second derivative with respect to time characterizes the curvature change of the joint mutation index. The curvature maxima correspond to mutation points, that is, the index changes drastically at the mutation points. This is the curvature threshold coefficient, which can be set to 0.15 to determine significant abrupt changes; For joint mutation index sequences The absolute maximum value, The joint mutation indicator sequence, i.e. A sequence that changes over time. The t-th element is .

[0025] S203. Based on the boundary points of the segment boundary set, divide the original values ​​of current, voltage, and power at each time moment into multiple sub-segments, calculate the data matrix of each sub-segment, and the data matrix of each sub-segment includes the current sequence, voltage sequence, and power sequence of each sub-segment. The dimension of the data matrix of the kth sub-segment is... , For the first The length of each sub-segment, where each row corresponds to a time point, each column corresponds to an electrical variable, and each sub-segment corresponds to a continuous interval between signal structure abrupt changes; the data matrix of the k-th sub-segment. The calculation formula is: , In the formula, This is the current sequence of the k-th sub-segment; This is the voltage sequence of the k-th sub-segment; Let k be the power sequence of the k-th sub-segment; k is the sub-segment index. ; This represents the total number of sub-segments; Distribution network current, voltage, and power time-series data exhibit non-stationary characteristics, with transient disturbances and steady-state fluctuations intertwined. Conventional equal-length segmentation methods for signal segmentation analysis, by simply dividing the data at equal intervals based on time length, easily fragment a complete transient process into different sub-segments, compromising event integrity. Furthermore, fixed-length sliding windows cannot adapt to the duration of dynamic changes in events such as fault impacts and harmonic oscillations, leading to distorted feature extraction. S2 constructs a joint abrupt change index by fusing kurtosis and spectral entropy indices to dynamically detect abrupt changes in signal structure, achieving adaptive non-stationary signal segmentation while preserving the complete electromagnetic transient process. Figure 4 , Figure 5 As shown, the joint mutation index curve used in S2 reflects the comprehensive change characteristics of the signal in the time and frequency domains. Regions with higher index values ​​correspond to time points when the signal characteristics change significantly. By detecting these mutation points, the segment boundaries are adaptively determined. The segment boundaries based on the mutation index can more accurately correspond to the actual location of the signal characteristic changes, ensuring the relative consistency of the signal characteristics within each sub-segment.

[0026] S3. Calculate the coupling components based on the current sequence, voltage sequence, and power sequence of each sub-segment, and calculate the decoupling characteristic matrix based on the coupling components; S301. Based on the current sequence, voltage sequence, and power sequence of each sub-segment, calculate the current sequence tensor, voltage sequence tensor, and power-current physical relationship constraint tensor of each sub-segment to obtain the coupling components. The coupling tensor of the k-th sub-segment represents the first... The physical coupling relationship of current, voltage and power in each segment is embedded with power-current constraints to eliminate redundancy and characterize the multivariable nonlinear physical relationship. No. The tensor dimension of the current sequence of each sub-segment is Store current timing data, the first Current sequence tensor of each sub-segment The calculation formula is: ; No. The voltage sequence tensor dimension of each sub-segment is Store voltage timing data, the first Voltage sequence tensor of each segment The calculation formula is: ; No. The power-current physical relationship constraint tensor dimension of each segment is... Characterizing the physical constraints of the power-current relationship, the first Power-current physical relationship constraint tensor of each segment The calculation formula is: , In the formula, For Hadamah accumulation; The current sequence of the kth sub-segment The element-wise reciprocal; The coupling tensor of the k-th sub-segment Dimensions , For the first The length of each sub-segment, i.e., the number of time points; It corresponds to 3 modalities and 3 feature channels; among them, Corresponding Coupled Tensor The third dimension is the feature channel with an index of 1. Corresponding Coupled Tensor The third dimension has a feature channel index of 2. Corresponding Coupled Tensor The third dimension has a feature channel index of 3; S302. Apply sparse constraints to the coupled tensor using Tucker decomposition to obtain the core tensor and the factor matrix representing the coupling relationship, time evolution mode, and physical constraint mode of the variables. Define the decoupling feature matrix based on the factor matrix. The optimization function that minimizes the non-negativity constraint Tucker decomposition through an iterative process is expressed as: , In the formula, This is the mode 1 factor matrix, with dimension 1. The solution is obtained through optimization algorithms, which characterize the coupling relationship between variables. It is calculated independently for each sub-segment, with the current sub-segment being the k-th sub-segment by default. This is a pattern 2 factor matrix with dimension 1. The time evolution pattern is obtained by optimizing the algorithm. It is calculated independently for each sub-segment, with the current sub-segment being the k-th sub-segment by default. This is a pattern 3 factor matrix with dimension 1. The solution is obtained through optimization algorithms, which characterize the physical constraint modes. Each sub-segment is calculated independently, with the current sub-segment being the k-th sub-segment by default. For about , and The minimization optimization problem is solved by calculating and optimizing the k-th sub-segment. It is the Frobenius norm; Let be the coupling tensor of the k-th sub-segment, with dimension . , characterizing the The physical coupling relationship of current, voltage and power in each segment is embedded with power-current constraints to eliminate redundancy; The core tensor, representing the latent dimensional relationships after decomposition, is obtained through optimization using the non-negativity-constrained Tucker decomposition algorithm. Initial values ​​are randomly generated, and iterative updates are performed to minimize the objective function, with dimensions of [dimensionality missing]. Calculations are performed independently for each sub-segment, with the current sub-segment being the k-th sub-segment by default. The rank of mode 1, with a default value of 8, represents the potential dimension of the variable coupling relationship; The rank of mode 2, with a default value of 8, represents the potential dimension of the time evolution mode; The rank of mode 3, with a default value of 3, represents the latent dimension of the physical constraint mode; The product of mode 1, i.e., the core tensor and Multiply along pattern 1; The product of mode 2, i.e., the core tensor and Multiply along pattern 2; For mode 3 product, i.e. and Multiply along pattern 3; This is the sparsity constraint coefficient, with a value of 0.01, used to control the sparsity of the factor matrix; For pattern index, Corresponding variable coupling, Corresponding to the evolution over time, Corresponding physical constraints; It is an L1 norm; For the first The factor matrix of each pattern is obtained by solving an optimization algorithm; definition To decouple the characteristic matrix, due to the factor matrix Characterizing the time evolution pattern, therefore taking , dimension Each sub-segment is calculated independently, with the current sub-segment being the k-th sub-segment by default.

[0027] In practical implementation, by solving this optimization problem, we can obtain... and , , Factor matrix Feature representations that include a time dimension can effectively preserve information about the dynamic changes in signals and are suitable as input features for time-frequency feature extraction.

[0028] Strong nonlinear coupling exists between electrical variables. Directly concatenating these variables as inputs leads to feature redundancy, while conventional independent channel processing methods ignore physical constraints, thus reducing state identifiability. S3 constructs a third-order coupling tensor with embedded power and current physical constraints and applies a non-negative constraint Tucker decomposition to extract a decoupling feature matrix characterizing the coupling relationship and time evolution mode of variables, thereby eliminating redundant correlations between electrical variables.

[0029] S4. Construct a state analysis model and train the state analysis model to obtain the trained state analysis model; S401. Calculate the dual-channel fusion feature matrix to construct the spectrum-time domain dual-channel feature extraction module; S4011. Calculate the spectral convolution output features of each column of the decoupled feature matrix to obtain the spectral convolution output matrix; The spectral convolution output feature of the j-th column feature, with dimension . The calculation is performed on the k-th sub-segment to characterize the adaptive frequency domain features, which can highlight the frequency domain patterns in signal abrupt change regions; the spectral convolution output features of the j-th column features. The calculation formula is: , In the formula, To modify the activation function of the linear unit; This is the inverse Fourier transform, which converts the frequency domain signal back to the time domain; Fourier transform converts the time-domain signal to the frequency domain. For decoupling characteristic matrix The j-th column feature vector has dimension . The calculation is performed on the k-th sub-segment; This is an element-wise multiplication operation in the frequency domain; For complex spectral kernels, the dimension is The elements are complex numbers and are trainable parameters. Frequency domain features are adaptively learned through frequency domain convolution, and specific harmonic or transient frequency patterns of the distribution network are learned during training. Use the Sigmoid activation function; For the first The local spectral entropy vector of the column features, with dimension . It is used to modulate the convolution output, enhancing the sensitivity to complex frequency domain changes; Time of the first Local spectral entropy vector of column features The calculation formula is: , In the formula, Frequency index; For decoupling characteristic matrix The Column feature vector exist The short-time Fourier transform of the signal within the time window is at the 1st The power spectrum ratio at a frequency point is used to quantify the frequency complexity of the feature vector; Decoupling feature matrix The Column feature vector exist The short-time Fourier transform of the signal within the time window is at the 1st Power spectral density at frequency points The calculation formula is: , In the formula, For decoupling characteristic matrix The Column feature vector exist The short-time Fourier transform of the signal within the time window is at the 1st The power spectral density value at a frequency point is obtained by performing a short-time Fourier transform on the feature vector within a local window; S4012. Calculate the time-domain convolution output features of each column feature of the decoupling feature matrix to obtain a time-domain convolution output matrix; The time-domain convolution output feature of the jth column feature represents an adaptive time-domain feature that can sensitively respond to the local change rate of the current and voltage; the calculation formula of the time-domain convolution output feature of the jth column feature is: , In the formula, is a convolution kernel offset; is a convolution kernel time radius, and the default value is 5, defining a convolution window range of ; is a weight of the time-domain convolution kernel at the offset of , which is a trainable parameter, dynamically adjusts the response to the time series through learning, and enhances the capture of mutation features; is the value of the jth column feature vector of the decoupling feature matrix at the time of ; is a sensitivity coefficient, which can be set to 5, to control the sensitivity to feature changes; is a feature difference, representing a local change rate; The calculation formula of the feature difference is: , In the formula, is the value of the jth column feature vector of the decoupling feature matrix at the time of ; is the value of the jth column feature vector of the decoupling feature matrix at the time of .

[0030] S4013. Calculate a dual-channel fusion feature matrix with spatio-temporal joint representation capability based on the spectral convolution output matrix and the time-domain convolution output matrix; The dimension of the dual-channel fusion feature matrix is , which has spatio-temporal joint representation capability and fuses frequency domain and time domain features; the calculation formula of the dual-channel fusion feature matrix is: , In the formula, is a fusion weight matrix with a dimension of , which is a trainable parameter that linearly transforms the spliced features to a low-dimensional space to reduce redundancy; is a spectral convolution output matrix with a dimension of , and the ith column vector is the spectral convolution output feature of the jth column feature.​ For splicing operations; The output matrix of the temporal convolution has dimensions of . The i-th column vector is the temporal convolution output feature of the j-th column feature. ; For the Softmax function; For dimension A vector of all 1s; Decoupling features still lack frequency domain information. Conventional Fast Fourier Transform ignores time-frequency locality, while wavelet transform has fixed basis functions, making it difficult to adapt to diverse fault spectra. S401 overcomes this problem by performing frequency domain convolution with learnable spectral kernels and time domain convolution with feature differential modulation on the decoupling feature matrix in parallel, fusing them to generate dual-channel features with spatiotemporal joint representation capabilities.

[0031] S402. Calculate the state probability vector and total loss function for each sub-segment; S4021. Calculate the enhanced feature matrix of each sub-segment based on the dual-channel fusion feature matrix; The vector is obtained by multiplying the core tensor with the physical constraint mode, then linearly mapping the context weight matrix, modulating it with the Sigmoid function, and expanding the modulated vector to the same dimension as the two-channel feature matrix through the outer product operation. Then, the two-channel feature matrix is ​​enhanced by the Hadamard product to generate an enhanced feature matrix that can more effectively capture the key modes of state change. No. The enhanced feature matrix of each sub-segment, with dimension [missing information]. Characterizing the spatiotemporal features modulated by physical constraints, it can more effectively capture key patterns of state changes; Enhanced feature matrix of each sub-segment The calculation formula is: , In the formula, For dimension A vector of all 1s; This is the context weight matrix, with dimension 1. , are trainable parameters used to map the product vector of the core tensor and the physical constraint mode to the feature space; This is the outer product operation, used to copy a vector into a matrix; For vectorization operations; S4022. Calculate the physical constraint mode difference scalar between each sub-segment and its predecessor based on the enhanced feature matrix of each sub-segment; By calculating the Frobenius norm difference between the current sub-segment physical constraint mode and the previous sub-segment physical constraint mode, a physical constraint mode difference scalar is obtained, which is used to quantify the intensity of state changes between sub-segments. No. The larger the value of the physical constraint mode difference scalar between the first sub-segment and the previous sub-segment, the more obvious the state abrupt change. scalar of physical constraint modal difference between each sub-segment and the previous sub-segment The calculation formula is: , In the formula, This is the mode 3 factor matrix corresponding to the previous sub-segment, representing the physical constraint mode of the previous sub-segment. The current sub-segment is assumed to be the [number missing]th [segment missing]. Each segment; S4023. Calculate the state probability vector of each sub-segment; By using the current joint mutation index as attention weight, the enhanced feature matrix is ​​weighted and averaged over time points. Then, through the linear transformation of the classification weight matrix and the classification bias vector, the Softmax function is applied to obtain the state probability vector, thereby highlighting the importance of mutation points for state classification. No. The state probability vector of each sub-segment has a dimension of . , characterizing the The state classification results of each sub-segment; the first The state probability vector of each sub-segment The calculation formula is: , In the formula, This is the classification weight matrix, with dimension 1. , are trainable parameters used to linearly map the weighted averaged enhanced features to the state class space. This represents the number of status categories. By default, it includes six categories: normal state, short circuit fault, overvoltage, undervoltage, harmonic interference, and transient oscillation. In actual implementation, it must correspond to the number of categories and category names that are pre-marked by the user. For the first In each sub-segment The attention weight at time step is such that the larger the value, the more significant the change at that time step, representing the importance of the time step to the state classification. For the first Enhanced feature matrix of each sub-segment exist The row vector at time t, i.e. All feature values ​​at time , dimension ; This is the classification bias vector, with dimension . , are trainable parameters; No. In each sub-segment Attention weight at any moment The calculation formula of the current joint mutation index at time s is: , In the formula, The current joint mutation index at time s is: is distinguished from the time index t; S4024. Calculate the total loss function based on the physical constraint modal difference scalar between each sub-section and its previous sub-section, and the state probability vector of each sub-section; The total loss function is calculated by combining the classification cross-entropy loss and the state transition constraint loss, wherein the state transition constraint loss is obtained based on the L2 norm of the adjacent sub-section state probability vector and the exponential weighting of the state transition feature, and the total loss function is used to promote the state analysis model to be accurate in classification while meeting the physical law of state transition; the calculation formula of the total loss function is: , In the formula, is the true state label of the m-th sub-section, which is a one-hot vector with a dimension of is a manually pre-labeled data label; is the state transition constraint loss weight, which is a hyperparameter and can be set to 0.1, serving as the weight hyperparameter of the state transition constraint loss to control the strength of the smoothness constraint; is the state probability vector of the m-th sub-section, with a dimension of ; is a scaling factor, which is a hyperparameter and can be set to 1.0, used to adjust the influence of the physical constraint modal difference on the smoothness loss; The state analysis of the power distribution network needs to accurately identify the operating state, but there is a time sequence correlation between states, and direct independent classification ignores the physical law of state transition, which can easily lead to inconsistent classification results. The conventional method only uses classification loss without considering the smoothness of the state sequence, which reduces the distinguishability of the state. S402 enhances the double-channel feature by using the physical constraint modal, quantifies the state change strength between sub-sections, and combines the mutation index weighting classification, and introduces the state transition constraint through the joint loss function to ensure that the state classification result meets the accuracy and time sequence consistency.

[0032] S403. Based on the total loss function, repeat S401-S402 to perform state analysis model training and iterative optimization; The training continues until the stopping condition is met, and the stopping condition is: when the validation set loss does not decrease for 20 consecutive rounds, or the total number of iterations reaches the preset upper limit of 1000 rounds, the training is terminated, and the state analysis model parameters with the optimal performance of the validation set are saved.​​​​​

[0033] S5. The new power grid electric variable data collected by S1 are used to calculate new current original values at each time, new voltage original values at each time, and new power original values at each time, which are input into the trained state analysis model after S2-S3 are performed, to obtain state probability vectors, and the state category corresponding to the maximum probability is taken as the state analysis result of the sub-section.

[0034] The classification accuracy of the technical method under each power grid state is compared with the classification accuracy of other methods under each power grid state, and Table 1 is obtained. Figure 6 As shown in Table 1, Figure 6 the classification accuracy of the technical method under each power grid state is the highest, and the average accuracy is also relatively leading. In the power grid state where other methods perform poorly, the accuracy can also reach more than 85%, which proves the excellent performance of the technical method.

[0035] Table 1 Comparison of classification accuracy of different methods under each power grid state

[0036] To verify the performance improvement of the technical method relative to the optimal benchmark method (herein referred to as the machine learning method), the accuracy of the technical method and the optimal benchmark method under different scenarios is compared, and Table 2 is obtained. The overall average accuracy of the technical method is relatively high because the whole process is optimized in coordination, from measurement, segmentation, decoupling to modeling, and the state identification capability is systematically improved. The steady-state scenario (normal state) accuracy is relatively high because the multi-rate sampling and Kalman filtering provide steady-state data with higher signal-to-noise ratio, laying a more accurate foundation for classification. The transient and fault scenario accuracy is significantly improved because the joint mutation index segmentation ensures the integrity of the transient event, and the dual-channel feature extraction effectively captures the time-frequency local features of the fault. The accuracy of the complex disturbance scenario (harmonic, oscillation) is most significantly improved because the coupled tensor decoupling eliminates the redundancy between variables, and the state transition constraint loss makes the model learn the continuous disturbance pattern more in line with the physical law.

[0037] Table 2 Performance improvement analysis table of the technical method relative to the optimal benchmark method

[0038] The misjudgment rates of the optimal benchmark method (here, 100% accuracy) and the technical method under various power grid states are obtained as shown in Table 3. As shown in Table 3, the misjudgment rate of the technical method under normal state is low because high-precision measurement reduces false triggering, and characteristic decoupling avoids misjudging normal fluctuations as abnormal; the misjudgment rate of the technical method under short-circuit state is low because adaptive sampling completely captures fault peaks, and dual-channel characteristics enhance the recognition of mutation patterns; the misjudgment rate of the technical method under harmonic interference is low because the spectral entropy index and the learnable spectral kernel convolution are specially optimized for the separation ability of complex frequency domain patterns; and the misjudgment rate of the technical method under transient oscillation is low because the state transition constraint effectively smooths the state jump in the oscillation process, reducing isolated misjudgment points.

[0039] Table 3 Misjudgment rate reduction effect of the technical method under various power grid states

Claims

1. A method for measuring and analyzing the state of electrical variables in a power distribution network, characterized by, The method comprises the following steps: S1. Measure the power distribution network electric variable data based on the sampling time sequence, calculate the instantaneous power, calculate the current original value at each time, the voltage original value at each time, and the power original value at each time; S2. Calculate the current joint mutation index at each time, calculate the segmentation boundary set based on the current joint mutation index at each time, divide the current original value at each time, the voltage original value at each time, and the power original value at each time into multiple sub-sections based on the boundary points of the segmentation boundary set, calculate the data matrix of each sub-section, and the data matrix of each sub-section includes the current sequence of each sub-section, the voltage sequence of each sub-section, and the power sequence of each sub-section; S3. Calculate the coupling component based on the current sequence of each sub-section, the voltage sequence of each sub-section, and the power sequence of each sub-section, and calculate the decoupling feature matrix according to the coupling component; S4. Construct a state analysis model, and train the state analysis model to obtain a trained state analysis model; S401. Calculate a double-channel fusion feature matrix; S402. Calculate the state probability vector of each sub-section and the total loss function; S403. Based on the total loss function, repeat S401-S402 to perform state analysis model training and iterative optimization; S5. Obtain new current original value at each time, new voltage original value at each time, and new power original value at each time through new power distribution network electric variables collected by S1, input the new current original value at each time, the new voltage original value at each time, and the new power original value at each time into the trained state analysis model after executing S2-S3, obtain each state probability vector, and take the state category corresponding to the maximum probability value as the state analysis result of the sub-section.

2. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 1, characterized in that, The step S1 comprises: S101. For the non-stationary characteristics of the power distribution network signal, a variable step dynamic sampling interval adjustment method is used for sampling to calculate the sampling time sequence; S102. The sampling obtained current signal at each time and the sampling obtained voltage signal at each time are respectively subjected to Kalman filtering for noise reduction to obtain the filtered current value at each time and the filtered voltage value at each time; S103. The filtered current value at each time and the filtered voltage value at each time are converted into digital signals through a current quantization function to obtain the current original value at each time and the voltage original value at each time; S104. The power original value at each time is calculated based on the current original value at each time and the voltage original value at each time.

3. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 2, characterized in that, The sampling time sequence consists of n sampling times; the first sampling time The calculation formula of the first sampling time is as follows: ， In the formula, This is a time index, which defaults to the current time point; For the first Each sampling time; For sampling point index, , This represents the total number of sampling points; For the first Each sampling time; For the first Each sampling interval is determined through adaptive dynamic adjustment. The first sampling interval The calculation formula is: The calculation formula is: ， In the formula, is a reference sampling interval; is an L2 norm, i.e., a Euclidean norm, which is represented by represents an input object, and the input object specifically refers to ; is an analog current signal at the time t, which is measured by a current transformer; is an analog voltage signal at the time t, which is measured by a voltage transformer; is the first sampling time; is an analog current signal at the time t, which is measured by a current transformer; is an analog voltage signal at the time t, which is measured by a voltage transformer; is the first sampling time.

4. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 1, characterized in that, The step S2 comprises: S201. Calculate the current joint mutation index at each time; S202. Calculate the segmentation boundary set based on the current joint mutation index at each time; S203. Divide the current original value at each time, the voltage original value at each time, and the power original value at each time into sub-sections according to the boundary points of the segmentation boundary set to obtain the data matrix of each sub-section composed of the current sequence, the voltage sequence, and the power sequence.

5. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 4, characterized in that, The current joint mutation index at each time point, i.e., the current joint mutation index at time t The calculation formula is: ， In the formula, The standard deviation of the signal within the window; For local window width; From Time's up The definite integral at time step; The variable is the integral variable, representing the time index within the local time window; for The original value of the current at that moment; The average value of the signal within the window; The standard Gaussian kernel function; Frequency index; Minimum frequency; Maximum frequency; The short-time Fourier transform of the signal within the window at time t is at the t-th The power spectrum percentage at a given frequency point; It is a logarithmic function, with the default base being the natural constant.

6. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 1, characterized in that, The step S3 comprises: S301. Calculate the current sequence tensor of each sub-section, the voltage sequence tensor of each sub-section, and the power-current physical relationship constraint tensor of each sub-section based on the current sequence of each sub-section, the voltage sequence of each sub-section, and the power sequence of each sub-section to obtain the coupling component; S302. Apply sparse constraint Tucker decomposition to the coupling tensor to obtain the core tensor and the factor matrix representing the variable coupling relationship, the time evolution mode, and the physical constraint mode, and define the decoupling feature matrix based on the factor matrix.

7. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 1, characterized in that, The steps of S401 include: S4011. Calculate the spectral convolution output features of each column feature of the decoupling feature matrix to obtain a spectral convolution output matrix; S4012. Calculate the time domain convolution output features of each column feature of the decoupling feature matrix to obtain a time domain convolution output matrix; S4013. Calculate a dual-channel fusion feature matrix with spatiotemporal joint representation capability based on the spectral convolution output matrix and the time domain convolution output matrix.

8. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 7, characterized in that, The dual-channel fusion feature matrix The calculation formula is: ， where, is the fusion weight matrix; is the spectral convolution output matrix; is the concatenation operation; is the time domain convolution output matrix; is the Hadamard product; is the Softmax function; is the mode 3 factor matrix; is an all-ones vector of dimension .

9. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 1, characterized in that, The steps of S402 include: S4021. Calculate an enhanced feature matrix of each sub-section based on the dual-channel fusion feature matrix; S4022. Calculate a physical constraint modal difference scalar between each sub-section and its previous sub-section based on the enhanced feature matrix of each sub-section; S4023. Calculate a state probability vector of each sub-section; S4024. Calculate a total loss function based on the physical constraint modal difference scalar between each sub-section and its previous sub-section and the state probability vector of each sub-section.

10. The method for measuring and analyzing the state of electric variables of a power distribution network according to claim 9, characterized in that, The total loss function The calculation formula is: ， In the formula, k is the sub-segment index. ; This represents the total number of sub-segments; For the first The true state label of each sub-segment; The power sequence of the k-th sub-segment; For the first The state probability vector of each sub-segment; This is the scaling factor; For the first The physical constraint modal difference scalar between each sub-segment and the previous sub-segment.

Images