Electrical potential safety hazard identification method and device for novel grid-connected main body

Abstract

The invention discloses an electrical potential safety hazard identification method and device for a novel grid-connected main body. The method comprises the following steps: acquiring waveform data of voltage and current in operation of the novel grid-connected main body; performing differential processing on the waveform data of the voltage and the current to generate differential signals of the voltage and the current, and constructing voltage-current coupling track characteristics based on the differential signals of the voltage and the current; performing multi-scale time-frequency analysis based on the waveform data of the voltage and the current to extract time-frequency energy texture features; performing feature fusion on the voltage-current coupling track feature and the time-frequency energy texture feature to generate a joint criterion; and electrical potential safety hazard identification and state determination of the novel grid-connected main body are realized based on a joint criterion. The invention aims to improve the integrity and identification accuracy of hidden danger characterization and enhance the depiction capability of unsteady-state spectrum behaviors, harmonic waves and inter-harmonic evolution characteristics, thereby improving the stability and robustness of an electrical safety hidden danger identification result under a complex operation condition.

Description

Technical Field

[0001] This invention relates to the field of electrical safety hazard detection technology for novel grid-connected entities, specifically to a method and device for identifying electrical safety hazards in novel grid-connected entities. Background Technology

[0002] With the large-scale integration of new grid-connected entities such as distributed photovoltaic power, electric vehicle charging piles, and energy storage devices into the distribution network, the operation mode of the power system has gradually evolved from the traditional source-follow-load dynamic to a complex operation mode of bidirectional source-load interaction and rapid power fluctuation. These new grid-connected entities typically feature multiple control links, frequent switching of operating states, and fast dynamic response speeds. During their operation, the coupling relationship between voltage and current exhibits significant nonlinearity, time-varying characteristics, and non-steady-state features, leading to more diverse types of electrical safety hazards, with significantly enhanced concealment and suddenness.

[0003] Existing electrical safety hazard monitoring and identification technologies mainly focus on single electrical quantities such as voltage, current, and frequency. They typically employ threshold discrimination, statistical analysis, or pattern recognition methods based on steady-state characteristics, offering some detection capability for overvoltage, undervoltage, overcurrent, and certain harmonic issues. However, during the operation of new grid-connected systems, hazards often do not manifest as anomalies in a single electrical quantity, but rather as subtle changes in voltage-current coupling or non-steady-state evolution of the spectral structure. Traditional methods based on single-domain features are insufficient to comprehensively characterize these hazards. To enhance the ability to perceive complex operating states, some existing technologies have begun to introduce multi-feature analysis or time-frequency analysis methods. These methods extract frequency domain or time-frequency domain features of signals through Fourier transform, short-time Fourier transform, or wavelet transform to characterize harmonic content, spectral energy distribution, and their changing trends. These methods enhance the ability to describe non-steady-state spectral behavior to some extent, but they typically employ fixed time scales or single-band features and fail to effectively correlate spectral evolution information with voltage-current coupling behavior. Under complex operating conditions and multiple disturbances, anomalous features are easily masked by noise or changes in operating conditions. Furthermore, some studies have attempted to jointly characterize electrical signals through phase space reconstruction, trajectory analysis, or statistical modeling to depict the dynamic characteristics of the system's operating state. However, these methods often focus on the relationship between single signals or similar signals, paying insufficient attention to the coupling relationship between voltage and current under transient conditions and during cross-fundamental period changes. They also lack a collaborative analysis mechanism with spectral features, making it difficult to comprehensively reflect the multidimensional characteristics of the operating state of new grid-connected entities. In summary, existing electrical safety hazard identification schemes for new grid-connected entities mainly include: hazard identification methods based on single electrical quantities or static statistical features, non-steady-state analysis methods based on frequency domain or time-frequency domain features, and operating state characterization methods based on trajectory or phase space analysis. The above methods can reflect some of the operating characteristics of new grid-connected entities to a certain extent, but they generally have problems such as insufficient characterization of voltage-current transient coupling relationship, limited adaptability to non-steady-state spectrum evolution behavior, and insufficient ability to perform joint analysis of multiple features. They are difficult to meet the needs of accurate identification and stability determination of electrical safety hazards in complex operating environments. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a method and device for identifying electrical safety hazards in new grid-connected main bodies, addressing the aforementioned problems in the prior art. This invention aims to improve the completeness and accuracy of hazard characterization and enhance the ability to depict non-steady-state spectral behavior, harmonic and interharmonic evolution characteristics, thereby improving the stability and robustness of electrical safety hazard identification results under complex operating conditions.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for identifying electrical safety hazards in new grid-connected entities includes the following steps: acquiring waveform data of voltage and current during the operation of the new grid-connected entity; generating differential signals of voltage and current by differential processing of the waveform data; constructing voltage-current coupling trajectory features based on the differential signals of voltage and current; extracting time-frequency energy texture features by performing multi-scale time-frequency analysis on the waveform data of voltage and current; fusing the voltage-current coupling trajectory features and the time-frequency energy texture features to generate a joint criterion; and identifying and determining the status of electrical safety hazards in the new grid-connected entity based on the joint criterion.

[0006] Optionally, the construction of voltage-current coupling trajectory features based on differential voltage and current signals includes: normalizing the differential voltage and current signals, aligning and combining them on the time axis to construct coupling trajectory points, dividing the two-dimensional plane containing the differential voltage-differential current coupling trajectory points according to time windows, and processing the coupling trajectory points in each time window. We perform statistical analysis and construct a two-dimensional occupancy probability matrix to determine the spatial distribution characteristics of the differential voltage-differential current coupling trajectories in a two-dimensional plane. The trajectory domain evolution is used to measure the magnitude of change in coupled trajectories within adjacent time windows. And the trajectory geometry index used to reflect the degree of energy concentration of the coupled trajectory in the principal direction. Thus, the two-dimensional occupancy probability matrix for each time window is obtained. trajectory domain evolution quantity and trajectory geometric morphology index The characteristics of the voltage-current coupling trajectory formed.

[0007] Optionally, the two-dimensional occupancy probability matrix The expression for the computation function is: ; in, For the k-th time window Within the two-dimensional binning range The two-dimensional occupancy probability matrix within the differential voltage-differential current coupling trajectory is divided into multiple two-dimensional bin intervals. Indicates the k-th time window Falling into the two-dimensional compartment range The number of valid trajectory points within the range; This indicates an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise. and For the aligned voltage and current normalized differential signals, and Represent the first and second axes on the differential voltage axis and differential current axis, respectively. The and the first Each container compartment; The trajectory domain evolution quantity The expression for the computation function is: ; in, For the (k-1)th time window Within the two-dimensional binning range The two-dimensional occupancy probability matrix within, For the first time window The trajectory domain evolution quantity; The trajectory geometric shape index The expression for the computation function is: ; in, and The coupling trajectory of differential voltage and differential current in the k-th time window The inner covariance matrix eigenvalues, A stabilizing term is introduced to prevent the denominator from being zero, and we have: ; Here, cov is the covariance operator.

[0008] Optionally, the multi-scale time-frequency analysis based on voltage and current waveform data to extract time-frequency energy texture features includes: multiplying the discrete signals of voltage and current to obtain a discrete signal of instantaneous power; performing a short-time Fourier transform on the discrete signal of instantaneous power using multiple sets of different window functions or frame shift parameters to obtain the time-frequency energy distribution of multiple sub-bands; extracting energy mutation features, harmonic evolution features, and spectral centroid features from the time-frequency energy distribution of multiple sub-bands, and indexing them with the time frame of the short-time Fourier transform. As a unified time scale, and as an index of discrete signals in voltage and current waveform data. satisfy: ; in, The time step between adjacent frames is the analysis window length for the short-time Fourier transform.

[0009] Optionally, the calculation function expression for the energy mutation characteristic is: ; ; in, For the first Frame energy mutation characteristics and The first Frame and the -1 frame of full-band energy, For the first Subband of a frame The time-frequency energy distribution; the harmonic evolution characteristics are harmonic band energy characteristics. ,in For the first The first frame The energy of the subharmonic band is calculated using the following function expression: ; in, For the first The center frequency of the subharmonic Harmonic bandwidth; The calculation function expression for the spectral centroid feature is as follows: ; in, For the first Spectral centroid features of the frame For children The frequency magnitude.

[0010] Optionally, the step of fusing the voltage-current coupled trajectory features and the time-frequency energy texture features to generate a joint criterion includes: constructing vectors from the voltage-current coupled trajectory features and the time-frequency energy texture features to obtain a trajectory domain comprehensive feature vector. and time-frequency domain eigenvectors ; synthesize the feature vector of the trajectory domain and time-frequency domain eigenvectors Normalization is performed separately, and the fused feature vector shown in the following formula is constructed. : ; in, These are the weighting coefficients; the fused feature vectors The dimensionality-reduced fused feature vector is obtained by performing dimensionality reduction mapping according to the following formula. : ; in, This is a linear dimensionality reduction matrix, where the number of rows represents the target dimensionality reduction dimension and the number of columns represents the fused feature vectors. The dimension is determined; ultimately, a feature vector is obtained by fusing voltage-current coupling trajectory features, time-frequency energy texture features, and dimension-reduced features. The combined criteria constituted.

[0011] Optionally, the identification and status determination of electrical safety hazards of new grid-connected entities based on joint criteria includes: ① performing feature consistency analysis, time-series joint analysis, and frequency band correlation analysis based on joint criteria, wherein the feature consistency analysis includes: comparing the trajectory domain evolution of adjacent time windows. Trajectory Geometric Morphology Index Energy mutation characteristics Spectral centroid characteristics The analysis considers the direction and magnitude of change of the four features. If the direction and magnitude of change of the four features are consistent, the result of the feature consistency analysis is determined to be "normal consistency state"; otherwise, the result of the feature consistency analysis is determined to be "abnormal consistency state". The temporal joint analysis includes: extracting and fusing feature vectors after dimensionality reduction. The amount and rate of change within a specified time window, if the feature vectors are fused after dimensionality reduction. If the amount and rate of change within a specified time window do not exceed a preset threshold, the result of the time-series co-analysis is determined to be "time-series co-analysis normal state"; otherwise, the result of the time-series co-analysis is determined to be "time-series co-analysis abnormal state". The frequency band correlation analysis includes: targeting the harmonic band energy characteristics If the change in energy of one or more harmonic bands exceeds a preset threshold within a time window of a certain frame, and the trajectory domain evolution... and trajectory geometric morphology index If the change in the frequency band correlation analysis also exceeds the corresponding preset threshold within the time window of the frame, then the result of the frequency band correlation analysis is determined to be "spectrum-driven coupling anomaly"; otherwise, the result of the frequency band correlation analysis is determined to be "spectrum-driven coupling normal". ② The electrical safety status of the new grid-connected entity is comprehensively determined by combining the results of feature consistency analysis, time-series joint analysis and frequency band correlation analysis: if the result of feature consistency analysis is "consistency normal state", the result of time-series joint analysis is "time-series joint normal state" and the result of frequency band correlation analysis is "spectrum-driven coupling normal", then the electrical safety status of the new grid-connected entity is determined to be safe; otherwise, the electrical safety status of the new grid-connected entity is determined to have potential safety hazards.

[0012] The present invention also provides an electrical safety hazard identification device for a new type of grid-connected entity, comprising a microprocessor and a memory interconnected thereto, wherein the microprocessor is programmed or configured to execute the electrical safety hazard identification method for the new type of grid-connected entity.

[0013] The present invention also provides a computer-readable storage medium storing a computer program or instructions that are programmed or configured to execute the electrical safety hazard identification method for a new type of grid-connected entity by a processor.

[0014] The present invention also provides a computer program product, including a computer program or instructions, which are programmed or configured to execute the electrical safety hazard identification method for a new type of grid-connected entity via a processor.

[0015] Compared with existing technologies, the present invention mainly achieves the following beneficial effects: 1. The present invention proposes a method for identifying electrical safety hazards in new grid-connected entities by differentially processing voltage and current signals and constructing voltage-current coupling trajectories based on differential voltage and differential current signals. This method differentially processes voltage and current waveforms collected during the operation of the grid-connected entity, highlighting their transient change characteristics, and constructs voltage-current coupling trajectories based on differential voltage and differential current signals. The coupling trajectory features are formed by the trajectory shape and its evolution characteristics, which are used to characterize the electrical safety hazard characteristics under changes in the operating state of the grid-connected entity. 2. To address the problem that the voltage-current coupling trajectory alone cannot fully reflect the non-steady-state spectral evolution characteristics, the present invention introduces multi-scale time-frequency analysis as a supplementary characterization method for the coupling trajectory features. By performing multi-scale time-frequency processing on voltage and / or current signals, time-frequency feature information reflecting harmonics, interharmonics, and spectral changes over time is extracted to compensate for the shortcomings of the coupling trajectory features in characterizing spectral evolution. 3. This invention performs synergistic analysis of voltage-current coupling trajectory characteristics and multi-scale time-frequency characteristics to form a joint discrimination mechanism for identifying electrical safety hazards in new grid-connected entities, and provides a corresponding device for implementation. Through unified processing and joint discrimination of coupling trajectory characteristics and time-frequency characteristics, effective identification of electrical safety hazards and determination of operating status of new grid-connected entities can be achieved. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the basic process of the method in an embodiment of the present invention.

[0017] Figure 2 This is a comparison diagram of the differential signal voltage-current coupling trajectory characteristics under different operating conditions in the embodiments of the present invention.

[0018] Figure 3 This is the fusion feature vector score of the voltage sag signal A in this embodiment of the invention.

[0019] Figure 4 This is the fusion feature vector score of the voltage sag signal B in this embodiment of the invention.

[0020] Figure 5 This is the fused feature vector score of the voltage sag signal C in this embodiment of the invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

[0022] like Figure 1 As shown, the electrical safety hazard identification method for novel grid-connected main bodies in this embodiment includes the following steps: S1, collects waveform data of voltage and current during the operation of the new grid-connected main body; S2, the voltage and current waveform data are processed differentially to generate differential voltage and current signals, and voltage-current coupling trajectory features are constructed based on the differential voltage and current signals. S3, based on voltage and current waveform data, performs multi-scale time-frequency analysis to extract time-frequency energy texture features; S4, feature fusion of voltage-current coupling trajectory features and time-frequency energy texture features to generate joint criteria; S5, based on joint criteria, identifies and determines the electrical safety hazards and status of new grid-connected entities.

[0023] Steps S2 and S3 are parallel steps. This embodiment's method for identifying electrical safety hazards in novel grid-connected systems can characterize the voltage-current transient coupling relationship and its dynamic evolution characteristics during the operation of these systems, thereby improving the completeness of hazard characterization and the accuracy of identification. By introducing a multi-scale time-frequency feature characterization mechanism, this invention enhances the ability to characterize unsteady-state spectral behavior, harmonic and interharmonic evolution characteristics, thus improving the stability and robustness of electrical safety hazard identification results under complex operating conditions.

[0024] In step S1 of this embodiment, when collecting the voltage and current waveform data of the new grid-connected main body during operation, the collected discrete voltage and current waveforms are as follows: ; in, and These are the voltage and current at the nth sampling point, respectively. The sampling point number is [number], and the sampling period is [period]. , The sampling point length of one period, the sampling frequency .

[0025] Perform on signal The order difference is defined mathematically as follows: ; ; in, for The first difference, For waveform, for p-th order difference, for p-1 order difference, for The p-th order difference. (The rest is missing from the original text.) Substitute voltages respectively Current You can get Step differential voltage and Step differential current In this embodiment, a first-order difference is used to highlight the transient changes between adjacent sampling points: ; .

[0026] When the system fundamental frequency is The corresponding fundamental period is In the discrete domain, the number of delayed samples is taken. : , in, This indicates a rounding operation performed on the input variable.

[0027] definition Voltage and current delay difference per cycle and for: ; ; The voltage and current difference across a single cycle and The function expression is: ; ; The differential voltage and differential current obtained by choosing any of the above differential forms can be denoted as follows: , .

[0028] In this embodiment, step S2, constructing voltage-current coupling trajectory features based on differential voltage and current signals, includes: normalizing the differential voltage and current signals, aligning and combining them on the time axis to construct coupling trajectory points, dividing the two-dimensional plane containing the differential voltage-differential current coupling trajectory points according to time windows, and processing the coupling trajectory points in each time window. We perform statistical analysis and construct a two-dimensional occupancy probability matrix to determine the spatial distribution characteristics of the differential voltage-differential current coupling trajectories in a two-dimensional plane. The trajectory domain evolution is used to measure the magnitude of change in coupled trajectories within adjacent time windows. And the trajectory geometry index used to reflect the degree of energy concentration of the coupled trajectory in the principal direction. Thus, the two-dimensional occupancy probability matrix for each time window is obtained. trajectory domain evolution quantity and trajectory geometric morphology index The characteristics of the voltage-current coupling trajectory formed, Figure 2 This is a comparison chart of the differential signal voltage-current coupling trajectory characteristics under different operating conditions in this embodiment. The voltage sag in the chart is the voltage sag signal, generated by Simulink simulation of the IEEE 33 system connected to the new grid-connected main body. This voltage signal is generated by adding 580kW active load and 80kW reactive load. Figure 2 As shown, this is based on the voltage difference component within the time interval of 0.20-0.5s. With current difference component The constructed Lissajous coupling trajectory diagram (voltage-current coupling trajectory feature diagram) shows that both the voltage difference component and the current difference component have been Z-score normalized to eliminate the influence of amplitude scale differences on the trajectory morphology. Blue trajectories correspond to voltage sag conditions; orange trajectories correspond to normal operating conditions. Trajectories under normal operating conditions exhibit a concentrated, stable, and symmetrical distribution; trajectories under voltage sag conditions exhibit a dispersed, stretched, and significantly directional distribution.

[0029] To reduce the impact of signal amplitude scale differences on the coupling trajectory morphology, in this embodiment, the normalized function expression for the differential signals of voltage and current is as follows: ; in, and These are the normalized voltage and current differential signals at the nth sampling point, respectively. and These are the voltage and current differential signals at the nth sampling point, respectively. and These are the average values ​​of the differential voltage and differential current, respectively. and These are the standard deviations of the differential voltage and differential current, respectively. To prevent extremely small positive numbers from being divided by zero, amplitude normalization is performed on both voltage and current signals to eliminate amplitude differences caused by variations in operating conditions. This normalization is a linear scaling transformation, which does not affect the basic shape and phase coupling relationship of the Lissajous curve trajectory formed by the two signals, thereby improving the stability and comparability of the hazard characterization results. When the timestamps corresponding to the differential voltage and differential current are not completely consistent, they can be aligned by interpolation within the overlapping time interval to obtain a sequence under a unified time axis. , A Lissajous curve is a curve trajectory synthesized by two sinusoidal vibrations along mutually perpendicular directions. This concept is commonly used in signal and image processing, and its functional expression is: ; ; in, and There are two signals. and The amplitudes of the two signals are... and Let be the angular frequencies of the two signals. and Let be the phase of the two signals, and t be time. When the parameter... When changing, point The trajectory formed in a plane is called a Lissajous curve. Under the condition of the same frequency... Below, the trajectory shape is determined by the amplitude ratio and phase difference The implicit relation of the decision can be represented in elliptic form: ; Therefore, it can be seen that when the phase relationship or amplitude ratio between two signals changes, the corresponding trajectory shape will change. In this embodiment, differential voltage and differential current are used as two-dimensional coordinate axes, and the discrete time... Constructing coupled trajectory points: ; Within a preset time window, the point set The resulting planar trajectory can be viewed as a Lissajous-like coupling trajectory of differential voltage and differential current. When the operating state of the new grid-connected entity changes, the transient coupling relationship between voltage and current changes accordingly, leading to alterations in trajectory morphology, distribution density, and evolution characteristics. This is a discrete-time index, belonging to the set of integers, used to represent the sampling time of the differential voltage and differential current signals. Because the differential operation introduces a delay term based on the fundamental period, this discrete-time index... The value of must satisfy In the first Within a preset time window The range of values ​​is determined by the time window index set. Given a set of time window indices A continuous subset of integers within the domain of the differential signal is defined as: ; in, For the first The starting index of each time window. The window length is defined. A time-delay differential operation based on the fundamental period is performed on the acquired voltage and current signals to obtain a differential sequence characterizing the cross-cycle changes of the signals. A two-dimensional coupled trajectory is then constructed based on this differential sequence to reflect the dynamic evolution characteristics of the grid-connected entity during operation, thereby achieving accurate characterization of potential electrical safety hazards.

[0030] The two-dimensional plane containing the coupling trajectory of differential voltage and differential current is divided into several two-dimensional bin intervals. Let the first... The time window is Within this window, the coupling trajectory points are statistically analyzed to construct a two-dimensional occupancy probability matrix that characterizes the spatial distribution characteristics of the differential voltage-differential current coupling trajectory on a two-dimensional plane. In this embodiment, the two-dimensional occupancy probability matrix The expression for the computation function is: ; in, For the k-th time window Within the two-dimensional binning range The two-dimensional occupancy probability matrix within the differential voltage-differential current coupling trajectory is divided into multiple two-dimensional bin intervals. Indicates the k-th time window Falling into the two-dimensional compartment range The number of valid trajectory points within the range; This indicates an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise. and For the aligned voltage and current normalized differential signals, and Represent the first and second axes on the differential voltage axis and differential current axis, respectively. The and the first The binning intervals are divided into several sections. The obtained two-dimensional occupancy probability matrix satisfies the normalization constraint: ; The resulting matrix represents The spatial distribution of the differential voltage-differential current coupling trajectory within the window on the two-dimensional plane can be used as the coupling trajectory feature corresponding to the window.

[0031] To characterize the degree of change in coupling relationships over time, a trajectory distribution change measure is defined to quantify the change in the spatial distribution of coupled trajectories within adjacent time windows. For adjacent windows... and The corresponding two-dimensional occupancy probability matrices are as follows: and Thus, the trajectory domain evolution can be obtained. In this embodiment, the trajectory domain evolution quantity The expression for the computation function is: ; in, For the (k-1)th time window Within the two-dimensional binning range The two-dimensional occupancy probability matrix within, For the first time window The trajectory domain evolution quantity. It is used to measure the change in the coupling trajectory within adjacent time windows. The larger the value, the more significant the evolution of the differential voltage-differential current coupling relationship between adjacent windows.

[0032] To depict time windows The geometric morphological characteristics of the internally coupled trajectory are used to construct the covariance matrix of the differential voltage-differential current trajectory points within the window: ; Here, cov is the covariance operator.

[0033] Let the covariance matrix be... eigenvalues Then, the principal axis energy percentage is defined as the trajectory geometry index. Therefore, in this embodiment, the trajectory geometry index The expression for the computation function is: ; in, and The coupling trajectory of differential voltage and differential current in the k-th time window The inner covariance matrix eigenvalues, A stabilizing term introduced to prevent the denominator from being zero. Trajectory geometry index. Used to reflect the degree of energy concentration of the coupled trajectory in the principal direction. When When the value is large, it indicates that the trajectory exhibits significant stretching characteristics in the principal axis direction; when When the value is small, it indicates that the orbital distribution tends to be uniform and the anisotropic characteristics are not obvious.

[0034] By jointly modeling the features of the two-dimensional occupancy probability matrix, the trajectory distribution change, and the trajectory geometric morphology index, the differential voltage-differential current coupling trajectory can be characterized from multiple dimensions such as spatial distribution, temporal evolution, and geometric structure. This provides a stable and comparable feature basis for subsequent identification of the operating status of new grid-connected entities and identification of electrical safety hazards.

[0035] To characterize the non-stationary spectral evolution of electrical signals during the operation of the new grid-connected main body, in step S3 of this embodiment, multi-scale time-frequency analysis is performed on the voltage waveform, current waveform, or instantaneous power waveform calculated from the two to obtain the energy distribution of the signal in the time and frequency dimensions, and time-frequency energy texture features are constructed accordingly.

[0036] Based on the collected discrete voltage and current waveforms, the corresponding instantaneous power waveform is: ; in, This is the instantaneous power waveform at the nth sampling point. It can be directly applied to... , and For time-frequency analysis, the input signal is denoted as... To reduce the impact of the DC component on the spectrum analysis, the following can be done: Perform mean processing: ; in, for The mean processing result, for or .

[0037] In this embodiment, multi-scale time-frequency analysis based on voltage and current waveform data is used to extract time-frequency energy texture features. This includes multiplying the discrete signals of voltage and current to obtain a discrete signal of instantaneous power; performing short-time Fourier transform on the discrete signal of instantaneous power using multiple sets of different window functions or frame shift parameters to obtain the time-frequency energy distribution of multiple sub-bands; extracting energy mutation features, harmonic evolution features, and spectral centroid features from the time-frequency energy distribution of multiple sub-bands, and using the time frame index of the short-time Fourier transform. As a unified time scale, and as an index of discrete signals in voltage and current waveform data. satisfy: ; in, The time step between adjacent frames is the analysis window length for the short-time Fourier transform.

[0038] Discrete signals after mean processing The Short-Time Fourier Transform (STFT) can be expressed as: ; in, For the first Near frame, angular frequency Time-frequency energy at that location For window functions, For time frame indexing, For time frame step, ω is the angular frequency.

[0039] The time-frequency energy distribution obtained from the short-time Fourier transform is defined as the spectral energy map: ; in, Indicates the first Near frame, angular frequency The energy level at that point. The frequency axis is... When expressing this, it can be written as In this embodiment, a set of preset window lengths and frame shift parameters are used to perform time-frequency analysis. In further embodiments, multiple sets of different window functions or frame shift parameters can be used. Repeatedly performing the short-time Fourier transform yields S groups of time-frequency energy distributions: ; This enables multi-scale time-frequency characterization at different time-frequency resolutions. Based on time-frequency energy distribution... It can extract energy mutation features, harmonic evolution features, and spectral centroid features.

[0040] In this embodiment, the calculation function expression for the energy mutation characteristic is: ; ; in, For the first Frame energy mutation characteristics and The first Frame and the -1 frame of full-band energy, For the first Subband of a frame The time-frequency energy distribution.

[0041] Given the fundamental frequency In the case of the first The center frequency of the second harmonic is: ; Set harmonic bandwidth In this embodiment, the harmonic evolution characteristics are the harmonic band energy characteristics. ,in For the first The first frame The energy of the subharmonic band is calculated using the following function expression: ; in, For the first The center frequency of the subharmonic For harmonic bandwidth; by the first The first frame The time series of subharmonic band energy, i.e.: harmonic band energy characteristics It can be used to characterize the evolution of harmonic energy over time.

[0042] To characterize the energy migration and potential interharmonic components in the frequency domain, a spectral energy centroid can be defined. In this embodiment, the calculation function expression for the spectral centroid characteristics is: ; in, For the first Spectral centroid features of the frame This represents the frequency of the subband.

[0043] Through the above time-frequency analysis and feature extraction, the energy mutation characteristics are obtained. Harmonic band energy characteristics and spectral centroid sequence The above features together constitute the time-frequency energy texture features, which will be jointly analyzed and fused with the coupling trajectory features in the future to form a multi-perspective feature description for identifying electrical safety hazards in new grid-connected entities.

[0044] To form a multi-perspective feature description that complements the trajectory domain and the time-frequency domain, step S4 in this embodiment is used to perform feature alignment, normalization and fusion processing on the coupled trajectory features obtained in step S2 and the time-frequency energy texture features obtained in step S3, and output a fused feature vector for subsequent electrical safety hazard identification.

[0045] In step S3, for the first A sliding time window To obtain the features of the two-dimensional occupancy probability matrix , and its derived scalar characteristics and Expand the two-dimensional matrix into a vector: ; And constitute the trajectory domain comprehensive feature vector: ; Among them, superscript Represents trajectory domain features; This indicates transpose.

[0046] In step S4, regarding the window The corresponding time-frequency analysis results yield the time-frequency energy distribution. And extract energy mutation features Harmonic band energy characteristics and spectral centroid characteristics This constitutes the time-frequency domain feature vector: ; in, Let be the spectral centroid eigenvalue of the i-th window. For the sliding window index, when When sliding with the window, Constructing the spectral centroid feature sequence; superscript Representing time-frequency domain characteristics, the superscript indicates... This indicates transposition. In this embodiment, the time is the midpoint of the window. As an index, for trajectory domain features Time-frequency domain characteristics A one-to-one correspondence is established; when the temporal resolutions of the two types of features are inconsistent, alignment can be achieved using interpolation or resampling. To avoid the adverse effects of differences in numerical scale, units, and dynamic range between different types of features on the feature fusion results, the trajectory domain features and time-frequency domain features can be normalized separately, as defined below: ; in, and These represent the normalization operators set for trajectory domain features and time-frequency domain features, respectively. Because the two types of features differ significantly in physical meaning, statistical distribution, and numerical dynamic range, the normalization operators... and It is not required that the same normalization form be used; a suitable normalization method can be selected based on the numerical characteristics of the corresponding features.

[0047] In step S4 of this embodiment, fusing the voltage-current coupling trajectory features and time-frequency energy texture features to generate a joint criterion includes: constructing vectors from the voltage-current coupling trajectory features and time-frequency energy texture features to obtain trajectory domain comprehensive feature vectors. and time-frequency domain eigenvectors ; synthesize the feature vector of the trajectory domain and time-frequency domain eigenvectors Normalization is performed separately, and the fused feature vector shown in the following formula is constructed. : ; in, These are weighting coefficients used to adjust the contribution of trajectory domain information and time-frequency domain information to the fused features; they are also used to reduce the weighting of the high-dimensional fused feature vector. To eliminate redundant information, improve noise robustness, and enhance the generalization ability of subsequent discrimination or classification models, feature vectors will be fused. The dimensionality-reduced fused feature vector is obtained by performing dimensionality reduction mapping according to the following formula. : ; in, This is a linear dimensionality reduction matrix, where the number of rows represents the target dimensionality reduction dimension and the number of columns represents the fused feature vectors. The dimension is determined; ultimately, a feature vector is obtained by fusing voltage-current coupling trajectory features, time-frequency energy texture features, and dimension-reduced features. The combined criteria are as follows. In this embodiment, voltage sag signals A, B, and C are generated by Simulink simulation of the IEEE 33 system connected to the new grid-connected entity. Voltage sag signal A is generated by adding 580kW active load and 80kW reactive load; voltage sag signal B is generated by adding 1160kW active load and 80kW reactive load; and voltage sag signal C is generated by adding 1160kW active load and 160kW reactive load. The scores obtained by projecting these signals along the first principal component direction after PCA are as follows. Figure 3 , Figure 4 and Figure 5 As shown, the Y-axis The fused feature vector of the voltage sag signal Projected values ​​along the direction of the first principal component after joint PCA. From Figure 3 The results show that the first principal component of the combined PCA... It exhibits a significant response to voltage sag events while remaining stable under normal operating conditions. Figure 4 Under a larger active power load disturbance, the first principal component of the joint PCA Score change trend and Figure 3 Maintaining a high degree of consistency, the normal signal did not show similar abrupt changes, but only exhibited slow, continuous, small fluctuations, indicating that... The score is robust to changes in disturbance intensity, and its anomalous patterns are independent of specific load parameters. Figure 5 As reactive load further increases, the first principal component of the voltage sag signal... The score still forms a significant negative peak in the transient range, and the normal signal maintains the stable characteristics consistent with the first two operating conditions, the first principal component. The score can not only distinguish the presence of voltage sags, but also stably characterize their occurrence time and duration. Within the voltage sag occurrence interval, the first principal component of the voltage sag signal... All showed significant and consistent negative abrupt changes, while the score of the normal signal remained fluctuating slightly around zero. This result indicates that the first principal component of the voltage sag signal... It can effectively distinguish between voltage sags and normal operating conditions, and has a stable and consistent characterization ability for voltage sags under different disturbance intensities.

[0048] After obtaining the fused feature vector or its dimensionality reduction representation Based on this, step S5 is used to perform joint analysis on the coupled trajectory features and time-frequency energy texture features to construct a joint criterion for identifying electrical safety hazards. In this embodiment, the identification and status determination of electrical safety hazards of new grid-connected entities based on the joint criterion includes: ① performing feature consistency analysis, time-series joint analysis, and frequency band correlation analysis based on the joint criterion. The feature consistency analysis includes: comparing the trajectory domain evolution of adjacent time windows. Trajectory Geometric Morphology Index Energy mutation characteristics Spectral centroid characteristics The analysis considers the direction and magnitude of change of the four features. If the direction and magnitude of change of the four features are consistent, the result of the feature consistency analysis is determined to be "normal consistency state"; otherwise, the result of the feature consistency analysis is determined to be "abnormal consistency state". The temporal joint analysis includes: extracting and fusing feature vectors after dimensionality reduction. The amount and rate of change within a specified time window, if the feature vectors are fused after dimensionality reduction. If the amount and rate of change within a specified time window do not exceed a preset threshold, the result of the time-series joint analysis is determined to be "time-series joint normal state"; otherwise, the result of the time-series joint analysis is determined to be "time-series joint abnormal state". The frequency band correlation analysis includes: targeting each specified harmonic evolution feature. Subharmonic energy If the change in energy of one or more harmonic bands exceeds a preset threshold within a time window of a certain frame, and the trajectory domain evolution... and trajectory geometric morphology index If the change in the frequency band correlation analysis also exceeds the corresponding preset threshold within the time window of the frame, then the result of the frequency band correlation analysis is determined to be "spectrum-driven coupling anomaly"; otherwise, the result of the frequency band correlation analysis is determined to be "spectrum-driven coupling normal". ② The electrical safety status of the new grid-connected entity is comprehensively determined by combining the results of feature consistency analysis, time-series joint analysis and frequency band correlation analysis: if the result of feature consistency analysis is "consistency normal state", the result of time-series joint analysis is "time-series joint normal state" and the result of frequency band correlation analysis is "spectrum-driven coupling normal", then the electrical safety status of the new grid-connected entity is determined to be safe; otherwise, the electrical safety status of the new grid-connected entity is determined to have potential safety hazards.

[0049] When performing feature consistency analysis, temporal joint analysis, and frequency band correlation analysis based on joint criteria, feature consistency analysis is used to analyze the consistency of the changing trends between the evolution of trajectory domain features, geometric morphology indices, and energy mutation features and spectral centroid features in the time-frequency domain. This is used to determine whether the synergy between voltage-current coupling and spectral energy changes is abnormal. Temporal joint analysis is used to analyze the temporal changes of fused feature vectors within a continuous sliding time window, extracting joint change rates or mutation amplitude indicators to characterize the dynamic evolution characteristics of the operating state. Frequency band correlation analysis is used to analyze the correlation between harmonic energy changes and coupling trajectory morphology changes in a specific frequency band, identifying coupling anomalies caused by harmonics or unsteady-state spectra. Through these joint analyses, joint anomaly criteria reflecting the operating state of new grid-connected entities are constructed, providing a decision-making basis for subsequent identification of electrical safety hazards. Ultimately, the electrical safety status of the new grid-connected entity can be comprehensively determined by integrating the results of feature consistency analysis, timing joint analysis, and frequency band correlation analysis: if the feature consistency analysis result is "consistency normal state," the timing joint analysis result is "timing joint normal state," and the frequency band correlation analysis result is "spectrum-driven coupling normal," then the electrical safety status of the new grid-connected entity is determined to be safe; otherwise, the electrical safety status of the new grid-connected entity is determined to have potential safety hazards. Safety hazards include voltage-current coupling anomalies or spectrum anomalies caused by factors such as changes in operating conditions, control mode switching, grid connection / disconnection operations, and harmonic or non-steady-state spectrum injection.

[0050] Furthermore, when the fused feature vector or its reduced-dimensional vector representation meets preset abnormal conditions, it can be determined that the operating status of the new grid-connected entity within the corresponding time window has potential electrical safety hazards. These abnormal conditions include, but are not limited to: the fused features deviating from the distribution range of normal operating features; the disruption of the synergistic relationship between the coupling trajectory features and the time-frequency energy features; and abrupt changes or abnormal enhancements in the fused features along the time or frequency dimensions.

[0051] In summary, this embodiment uses differential processing on the voltage and current waveforms collected during the operation of the new grid-connected main body to highlight the transient changes of the signal across sampling points and across the fundamental frequency period. Based on this, a voltage-current coupling trajectory is constructed using the differential voltage and current signals, and the spatial distribution, temporal evolution, and geometric morphology of the coupling trajectory are modeled to form coupling trajectory features that can characterize changes in operating state. Simultaneously, multi-scale time-frequency analysis is introduced to extract time-frequency energy texture features to characterize the time-varying properties of harmonics, interharmonics, and non-steady-state spectral components. The coupling trajectory features and time-frequency features are aligned, normalized, and fused to construct a joint discrimination mechanism, enabling the identification and status determination of electrical safety hazards in the new grid-connected main body. Therefore, the advantages of this embodiment are as follows: By using differential processing and coupling trajectory feature construction, the transient coupling relationship between voltage and current can be effectively characterized, making up for the shortcomings of existing methods that are based only on a single electrical quantity or static feature; by fusing coupling trajectory features with multi-scale time-frequency features, the coupling relationship information and spectrum evolution information are taken into account, improving the completeness, accuracy and stability of electrical safety hazard identification under complex disturbances and changing operating conditions, thus making it more suitable for engineering application scenarios of new grid-connected entities.

[0052] Those skilled in the art will understand that the technical solutions provided by this invention can take the form of a method, a system, or a computer program product. Therefore, this invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. For example, this invention can provide an electrical safety hazard identification device for a novel grid-connected entity, including a microprocessor and a memory interconnected, the microprocessor being programmed or configured to execute the electrical safety hazard identification method for a novel grid-connected entity. This invention can provide a computer-readable storage medium storing a computer program or instructions programmed or configured to execute the electrical safety hazard identification method for a novel grid-connected entity via a processor. This invention can provide a computer program product including a computer program or instructions programmed or configured to execute the electrical safety hazard identification method for a novel grid-connected entity via a processor. Furthermore, this invention can take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0053] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for identifying electrical safety hazards in new grid-connected main bodies, characterized in that, The process includes the following steps: collecting waveform data of voltage and current during the operation of the new grid-connected main body; generating differential signals of voltage and current through differential processing of the waveform data of voltage and current; and constructing voltage-current coupling trajectory features based on the differential signals of voltage and current. Multi-scale time-frequency analysis based on voltage and current waveform data is performed to extract time-frequency energy texture features; voltage-current coupling trajectory features and time-frequency energy texture features are fused to generate a joint criterion; based on the joint criterion, the electrical safety hazards of new grid-connected entities are identified and their status is determined.

2. The method for identifying electrical safety hazards in a new type of grid-connected main body according to claim 1, characterized in that, The construction of voltage-current coupling trajectory features based on differential voltage and current signals includes: normalizing the differential voltage and current signals, aligning and combining them on the time axis to construct coupling trajectory points, dividing the two-dimensional plane containing the differential voltage-differential current coupling trajectory points according to time windows, and processing the coupling trajectory points in each time window. We perform statistical analysis and construct a two-dimensional occupancy probability matrix to determine the spatial distribution characteristics of the differential voltage-differential current coupling trajectories in a two-dimensional plane. The trajectory domain evolution is used to measure the magnitude of change in coupled trajectories within adjacent time windows. And the trajectory geometry index used to reflect the degree of energy concentration of the coupled trajectory in the principal direction. Thus, the two-dimensional occupancy probability matrix for each time window is obtained. trajectory domain evolution quantity and trajectory geometric morphology index The characteristics of the voltage-current coupling trajectory formed.

3. The method for identifying electrical safety hazards in a new type of grid-connected main body according to claim 2, characterized in that, The two-dimensional occupancy probability matrix The expression for the computation function is: ； in, For the k-th time window Within the two-dimensional binning range The two-dimensional occupancy probability matrix within the differential voltage-differential current coupling trajectory is divided into multiple two-dimensional bin intervals. Indicates the k-th time window Falling into the two-dimensional compartment range The number of valid trajectory points within the range; This indicates an indicator function that takes the value 1 when the condition inside the parentheses is true, and 0 otherwise. and For the aligned voltage and current normalized differential signals, and Represent the first and second axes on the differential voltage axis and differential current axis, respectively. The and the first Each container compartment; The trajectory domain evolution quantity The expression for the computation function is: ； in, For the (k-1)th time window Within the two-dimensional binning range The two-dimensional occupancy probability matrix within, For the first time window The trajectory domain evolution quantity; The trajectory geometric shape index The expression for the computation function is: ； in, and The coupling trajectory of differential voltage and differential current in the k-th time window The inner covariance matrix eigenvalues, A stabilizing term is introduced to prevent the denominator from being zero, and we have: ； Here, cov is the covariance operator.

4. The method for identifying electrical safety hazards in a new type of grid-connected main body according to claim 1, characterized in that, The multi-scale time-frequency analysis based on voltage and current waveform data to extract time-frequency energy texture features includes: multiplying the discrete signals of voltage and current to obtain a discrete signal of instantaneous power; performing a short-time Fourier transform on the discrete signal of instantaneous power using multiple sets of different window functions or frame shift parameters to obtain the time-frequency energy distribution of multiple sub-bands; extracting energy mutation features, harmonic evolution features, and spectral centroid features from the time-frequency energy distribution of multiple sub-bands, and using the time frame index of the short-time Fourier transform. As a unified time scale, and as an index of discrete signals in voltage and current waveform data. satisfy: ； in, The time step between adjacent frames is the analysis window length for the short-time Fourier transform.

5. The method for identifying electrical safety hazards in a new type of grid-connected main body according to claim 4, characterized in that, The calculation function expression for the energy mutation characteristic is: ； ； in, For the first Frame energy mutation characteristics and The first Frame and the -1 frame of full-band energy, For the first Subband of a frame The time-frequency energy distribution; the harmonic evolution characteristics are harmonic band energy characteristics. ,in For the first The first frame The energy of the subharmonic band is calculated using the following function expression: ； in, For the first The center frequency of the subharmonic Harmonic bandwidth; The calculation function expression for the spectral centroid feature is as follows: ； in, For the first Spectral centroid features of the frame For children The frequency magnitude.

6. The method for identifying electrical safety hazards in a new type of grid-connected main body according to claim 1, characterized in that, The feature fusion of voltage-current coupled trajectory features and time-frequency energy texture features to generate a joint criterion includes: constructing vectors from the voltage-current coupled trajectory features and the time-frequency energy texture features to obtain a trajectory domain comprehensive feature vector. and time-frequency domain eigenvectors ; synthesize the feature vector of the trajectory domain and time-frequency domain eigenvectors Normalization is performed separately, and the fused feature vector shown in the following formula is constructed. : ； in, These are the weighting coefficients; the fused feature vectors The dimensionality-reduced fused feature vector is obtained by performing dimensionality reduction mapping according to the following formula. : ； in, This is a linear dimensionality reduction matrix, where the number of rows represents the target dimensionality reduction dimension and the number of columns represents the fused feature vectors. The dimension is determined; ultimately, a feature vector is obtained by fusing voltage-current coupling trajectory features, time-frequency energy texture features, and dimension-reduced features. The combined criteria constituted.

7. The method for identifying electrical safety hazards in a new type of grid-connected main body according to claim 6, characterized in that, The method for identifying and determining the electrical safety hazards and status of new grid-connected entities based on joint criteria includes: ① performing feature consistency analysis, time-series joint analysis, and frequency band correlation analysis based on joint criteria. The feature consistency analysis includes comparing the trajectory domain evolution of adjacent time windows. Trajectory Geometric Morphology Index Energy mutation characteristics Spectral centroid characteristics The analysis considers the direction and magnitude of change of the four features. If the direction and magnitude of change of the four features are consistent, the result of the feature consistency analysis is determined to be "normal consistency state"; otherwise, the result of the feature consistency analysis is determined to be "abnormal consistency state". The temporal joint analysis includes: extracting and fusing feature vectors after dimensionality reduction. The amount and rate of change within a specified time window, if the feature vectors are fused after dimensionality reduction. If the amount and rate of change within a specified time window do not exceed a preset threshold, the result of the time-series co-analysis is determined to be "time-series co-analysis normal state"; otherwise, the result of the time-series co-analysis is determined to be "time-series co-analysis abnormal state". The frequency band correlation analysis includes: targeting the harmonic band energy characteristics If the change in energy of one or more harmonic bands exceeds a preset threshold within a time window of a certain frame, and the trajectory domain evolution... and trajectory geometric morphology index If the change in the frequency band correlation analysis also exceeds the corresponding preset threshold within the time window of the frame, then the result of the frequency band correlation analysis is determined to be "spectrum-driven coupling anomaly"; otherwise, the result of the frequency band correlation analysis is determined to be "spectrum-driven coupling normal". ② The electrical safety status of the new grid-connected entity is comprehensively determined by combining the results of feature consistency analysis, time-series joint analysis and frequency band correlation analysis: if the result of feature consistency analysis is "consistency normal state", the result of time-series joint analysis is "time-series joint normal state" and the result of frequency band correlation analysis is "spectrum-driven coupling normal", then the electrical safety status of the new grid-connected entity is determined to be safe; otherwise, the electrical safety status of the new grid-connected entity is determined to have potential safety hazards.

8. An electrical safety hazard identification device for a new type of grid-connected main body, comprising a microprocessor and a memory interconnected, characterized in that, The microprocessor is programmed or configured to execute the electrical safety hazard identification method for a novel grid-connected entity as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program or instructions, characterized in that, The computer program or instructions are programmed or configured to execute, via a processor, the electrical safety hazard identification method for a novel grid-connected entity as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program or instructions, characterized in that, The computer program or instructions are programmed or configured to execute, via a processor, the electrical safety hazard identification method for a novel grid-connected entity as described in any one of claims 1 to 7.

Images