An emulation-based power distribution area fault active early warning and positioning method, system, device and storage medium

Abstract

The application discloses a kind of based on simulation district fault active early warning and positioning method, system, equipment and storage medium, it is related to the field of smart grid fault detection and positioning, method includes: the electrical data of power distribution district multi-node is collected and fused, disturbance driving factor is constructed;While based on topology and equipment parameter, construct digital twin simulation model, obtain macroscopic deviation score by comparing real-time data and simulation trajectory;Further combined with the physical relationship between nodes, quantize disturbance propagation path and calculate path impact integral quantity;Fusion deviation score and path integral quantity are early warned, and according to the energy of each node propagation after early warning trigger, risk score is carried out, and the accurate positioning of fault source is realized;The application realizes the early perception of distribution network implicit fault and slowly varying fault, significantly improves the accuracy of early warning and positioning precision, effectively reduces the misjudgment rate, and quantitatively evaluates the propagation influence of disturbance in complex network.

Description

Technical Field

[0001] This invention relates to the field of smart grid fault detection and location, and in particular to a simulation-based method, system, device, and storage medium for proactive early warning and location of transformer area faults. Background Technology

[0002] With the development of smart grids and distribution automation technologies, distribution transformer substations, as a crucial link in the power system closest to the user side, are becoming increasingly important for ensuring power supply reliability and user safety through real-time monitoring and anomaly response capabilities. However, fault detection and location in distribution transformer substations still face numerous technical bottlenecks.

[0003] Traditional fault identification methods in power distribution systems generally rely on simple limit judgments or univariate analysis, triggering alarms only when monitored values ​​exceed set thresholds. They lack the ability to model the dynamic evolution of the system, making it difficult to detect latent faults, slowly changing faults, or anomalies under complex disturbances in a timely manner. Existing solutions often rely on point measurements or local data, failing to fully consider the topological connections and coupling characteristics between multiple nodes in the power distribution network structure. Especially in multi-branch interconnection or multi-source power supply structures, fault propagation paths are complex and have spatial delays, making traditional location methods based on current distribution or tripping information prone to misjudgment in such situations. Currently, there is a lack of a dynamic verification mechanism that integrates power system simulation models with real-time operating conditions, making simulation results difficult to use for actual operational decisions and hindering closed-loop control from prediction and monitoring to response. The propagation of abnormal information between multiple nodes is affected by multiple factors such as physical paths, electrical parameters, and load conditions. Disturbance signals often exhibit complex characteristics of nonlinearity, multi-scale, and multi-channel operation, and existing technologies generally lack a unified and interpretable indicator system for feature extraction and propagation modeling. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a simulation-based active early warning and location method, system, device and storage medium for transformer area faults.

[0005] Therefore, the technical problem solved by this invention is: how to achieve early warning of distribution radio station faults and accurate source tracing in complex networks.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a simulation-based method for proactive early warning and location of transformer area faults, comprising: Multidimensional electrical data is acquired and interpolated and aligned to obtain aligned time-series data. Based on the aligned time-series data, a structural interlocking disturbance driving factor construction method is introduced to construct the disturbance driving factor for each node. Based on the topology parameters and electrical equipment parameters of the transformer substation, a digital twin transformer substation simulation model is constructed to obtain simulation time series data. This data is then compared with the aligned time series data to obtain residual values. Based on these residual values, a simulation residual scoring function is introduced to determine the deviation score. Based on the disturbance driving factor and the physical connection relationship of the transformer area, a disturbance propagation modeling mechanism is introduced to calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral through temporal coupling processing. Based on the deviation score and the path impact integral, an active early warning judgment mechanism is defined to determine whether an early warning is triggered. After the warning is triggered, the node risk score is calculated based on the path impact integral of all nodes to determine the source of the fault.

[0007] As a preferred scheme for a simulation-based active early warning and location method for transformer area faults, wherein: The process involves acquiring multidimensional electrical data and performing interpolation and alignment to obtain aligned time-series data. Based on this aligned time-series data, a structural interlocking perturbation driving factor construction method is introduced to construct the perturbation driving factor for each node, including: Based on the aligned time-series data, a comprehensive disturbance driving factor is constructed for each node. Specifically, for each type of electrical data of the node, a single-channel disturbance factor that combines instantaneous acceleration and deviation relative to historical steady state is calculated. The single-channel disturbance factors corresponding to all electrical data categories of the node are weighted and fused to generate a disturbance driving factor that reflects the overall dynamic instability of the node.

[0008] As a preferred scheme for a simulation-based active early warning and location method for transformer area faults, wherein: Based on the topology parameters and electrical equipment parameters of the transformer substation, a digital twin transformer substation simulation model is constructed to obtain simulation time-series data. This data is then compared with aligned time-series data to obtain residual values. Based on these residual values, a simulation residual scoring function is introduced to determine the deviation score, which includes: Run the digital twin simulation model to generate simulation time series data of each key node in the transformer area under normal operating conditions; The simulation time series data is compared with the aligned time series data of the corresponding nodes and time points to calculate the residual sequence; Based on the residual sequence, a simulation residual scoring function is introduced to determine the deviation score of the system's macroscopic state. The simulation residual scoring function accumulates the residuals at each time point within a predefined time window, with the residuals closer to the current time point having a larger weight.

[0009] As a preferred scheme for a simulation-based active early warning and location method for transformer area faults, wherein: Based on the disturbance driving factor and the physical connection relationship of the transformer area, a disturbance propagation modeling mechanism is introduced to calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral through temporal coupling processing, including: Based on the comprehensive disturbance driving factor of each node and the physical connection relationship of the transformer area, the disturbance propagation intensity coefficient between any two key nodes is calculated. The disturbance propagation intensity coefficient is at least related to the comprehensive disturbance driving factor of the source node, the physical distance between the nodes, the line impedance characteristics, the electrical phase difference, and the measured disturbance propagation time delay.

[0010] The beneficial effects of this preferred technical solution are: by combining the disturbance state of the node itself with the physical connection characteristics of the network, the instantaneous intensity of the disturbance propagation between nodes can be quantified, providing a basis for analyzing the spread of anomalies on the spatial network.

[0011] As a preferred scheme for a simulation-based active early warning and location method for transformer area faults, wherein: The method of introducing a disturbance propagation modeling mechanism based on disturbance driving factors and physical connection relationships of transformer areas, calculating disturbance propagation intensity coefficients between nodes, constructing a multi-physical quantity coupling intensity tensor, and obtaining the path impact integral through temporal coupling processing also includes: For each type of electrical data between each pair of nodes, a multi-physical quantity coupling strength tensor is constructed based on the disturbance propagation strength coefficient, the instantaneous difference of electrical data, and the steady-state fluctuation characteristics. The path impact integral is calculated by performing attenuation integral processing on the coupling strength tensor of multiple physical quantities along the time dimension.

[0012] As a preferred scheme for a simulation-based active early warning and location method for transformer area faults, wherein: The aforementioned proactive early warning judgment mechanism, based on deviation score and path impact integral, determines whether an early warning is triggered by: An active warning is triggered if and only if the deviation score exceeds a preset deviation threshold and there is at least one pair of nodes whose path impact integral exceeds a preset path integral threshold for the pair of nodes.

[0013] The beneficial effects of this preferred technical solution are as follows: by adopting the dual criteria that macroscopic state anomalies and microscopic path impacts must be met simultaneously, false alarms caused by single-point data anomalies or local instantaneous interference can be avoided, thereby improving the reliability of the early warning signal.

[0014] As a preferred scheme for a simulation-based active early warning and location method for transformer area faults, wherein: After the warning is triggered, the node risk score is calculated based on the path impact integral of all nodes, and the fault source points are determined, including: After the warning is triggered, the fault source is traced based on the path impact integral of all node pairs; For each node, the risk score is calculated by combining the path impact integral with each neighboring node, the physical distance between nodes, the directional difference of the node's state vector, and the pre-set path importance weight. Based on the risk scores of all nodes, the node with the highest score is identified as the suspected source of the fault.

[0015] Secondly, the present invention provides a simulation-based active early warning and location system for transformer area faults, comprising: The multi-source fusion and feature extraction module is used to acquire multi-dimensional electrical data and perform interpolation and alignment to obtain aligned time-series data. Based on the aligned time-series data, a structural interlocking perturbation driving factor construction method is introduced to construct the perturbation driving factor for each node. The simulation benchmark comparison module is used to construct a digital twin simulation model of the transformer substation based on the topology parameters and electrical equipment parameters of the substation, obtain simulation time series data, compare it with the aligned time series data, obtain residual values, and introduce a simulation residual scoring function based on the residual values ​​to determine the deviation score value. The path energy accumulation module is used to introduce a disturbance propagation modeling mechanism based on the disturbance driving factor and the physical connection relationship of the transformer area, calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral quantity through temporal coupling processing. The early warning decision module is used to define an active early warning judgment mechanism based on the deviation score and path impact integral, and to determine whether to trigger an early warning. The network source tracing and location module is used to calculate the node risk score and determine the source of the fault based on the path impact integral of all nodes after an early warning is triggered.

[0016] Thirdly, the present invention provides a computer device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the simulation-based active early warning and location method for transformer area faults are implemented.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of a simulation-based active early warning and location method for transformer area faults.

[0018] The beneficial effects of this invention are as follows: By constructing a disturbance propagation coefficient, this invention introduces multiple physical parameters such as distance between nodes, impedance, phase difference, and delay, and superimposes a periodic modulation function to form a high-dimensional disturbance reachability expression. Combined with the subsequently constructed multi-physical quantity coupling tensor, it comprehensively captures the directionality, amplitude, and temporality of the disturbance propagation path, thus addressing the problems of independent node modeling and neglect of propagation mechanisms in traditional methods. By integrating the coupling tensor on the time axis to construct the path impact integral, it achieves modeling of the energy of the continuous impact of disturbances, distinguishing the fundamental differences between short-term disturbances and long-term evolutionary anomalies. This effectively isolates sporadic disturbances and focuses on systemic and cumulative fault risks, solving the problem of insufficient ability to distinguish between transient and steady-state anomalies in existing methods. By calculating the root cause score of the fault, and comprehensively considering factors such as the integral energy of the disturbance propagation path, the directional differences of the node state vector, physical distance attenuation, and path importance weights, a highly interpretable and accurate root cause inversion inference model is constructed. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is an overall flowchart of a simulation-based active early warning and location method for transformer area faults provided by the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0022] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a simulation-based active early warning and location method for transformer area faults, including: S1: Acquire multidimensional electrical data and perform interpolation and alignment to obtain aligned time series data. Based on the aligned time series data, introduce a structural interlocking disturbance driving factor construction method to construct the disturbance driving factor for each node. S2: Based on the topology parameters and electrical equipment parameters of the transformer area, a digital twin transformer area simulation model is constructed to obtain simulation time series data. This data is then compared with the aligned time series data to obtain residual values. Based on these residual values, a simulation residual scoring function is introduced to determine the deviation score. S3: Based on the disturbance driving factor and the physical connection relationship of the transformer area, a disturbance propagation modeling mechanism is introduced to calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral through temporal coupling processing; S4: Based on the deviation score and path impact integral, define an active early warning judgment mechanism to determine whether an early warning is triggered; S5: After triggering the warning, calculate the node risk score based on the path impact integral of all nodes and determine the source of the fault.

[0023] It should be noted that through steps S1-S5, real-time multi-source data is deeply integrated with the digital twin simulation model, and combined with disturbance propagation dynamics modeling based on network topology, a closed-loop intelligent diagnostic system of "monitoring-simulation-propagation analysis-decision-source tracing" is constructed. This not only enables early perception of latent faults and slowly changing anomalies, but also significantly improves the accuracy and interpretability of locating the root cause of faults in complex distribution networks by quantifying disturbance propagation paths and energy accumulation. This overcomes the technical shortcomings of traditional methods, such as reliance on static thresholds, high misjudgment rates, and inability to effectively trace fault propagation paths.

[0024] Example 2, refer to Figure 1 As one embodiment of the present invention, based on the previous embodiment, a simulation-based active early warning and location method for transformer area faults is provided, including: In this embodiment, in step S1 above, multidimensional electrical data is acquired and interpolated and aligned to obtain aligned time-series data. Based on the aligned time-series data, a structural interlocking disturbance driving factor construction method is introduced to construct the disturbance driving factor for each node, including: Multiple heterogeneous sensors are installed at key nodes in the distribution substation (such as transformer outlets, main branches, and load concentration points) to collect multi-dimensional electrical data, including voltage, current, temperature, power factor, and power quality.

[0025] The multidimensional electrical data was interpolated and aligned using cubic spline interpolation to obtain aligned time series data. ; Furthermore, to reduce computational resource pressure and preserve anomalous signal characteristics, a perturbation factor is constructed for the aligned time-series data: a structurally integrated perturbation driving factor construction method is introduced, fusing the variation characteristics of multi-dimensional electrical data into a single perturbation index. The perturbation driving factor for each key node at each time step is... The definition is as follows: in, It is the first Key nodes at time The perturbation driving factor is expressed in the response index of a single node; It is the first The first key node Electrical data at time The perturbation driving factor quantifies the dynamic instability of the channel at the current moment; This is the disturbance weight of the i-th electrical data, reflecting the importance of this electrical data in the comprehensive judgment (e.g., voltage is more sensitive than temperature). It is determined by the information entropy weighting method, and in this embodiment, the preferred value range is [value range missing]. And the sum is 1; It is the first The first key node Electrical data at time The aligned data values; It is the first The first key node The steady-state average of electrical data (historical average with a sliding window) is determined according to specific requirements; in this embodiment, it is preferably 5 to 30 seconds. It is a regularization factor used to prevent numerical instability issues caused by the denominator being zero. In this embodiment, it is preferably used. ; It is the first The first key node The aligned electrical data at time The second derivative, i.e. the change in acceleration and rate of change, is obtained by approximation using the three-point central difference. It represents the total number of electrical data.

[0026] In this embodiment, in step S2 above, a digital twin transformer area simulation model is constructed based on the topology parameters and electrical equipment parameters of the transformer area to obtain simulation time series data. This data is then compared with the aligned time series data to obtain residual values. Based on these residual values, a simulation residual scoring function is introduced to determine the deviation score, which includes: To describe the propagation trend of anomalies identified based on peeling structure, a digital twin transformer area simulation model is constructed. Furthermore, a digital twin transformer area simulation model is constructed based on a distributed power flow simulation method using topology modeling, and the modeling is implemented using power system simulation platforms such as OpenDSS (Open Distribution System Simulator) or DIgSILENTPowerFactory.

[0027] Specifically, the input data includes the actual transformer substation's GIS geographic information, electrical topology map, transformer parameters, conductor impedance, and load configuration. All input data is retrieved from the existing database of the transformer substation. This data is imported into the simulation platform to construct an equivalent circuit model of the node-branch structure. Steady-state and transient power flow calculations are performed using the Newton-Raphson method or DC power flow algorithm. This generates the simulation trajectory of multi-dimensional electrical data (voltage, current, active power, reactive power, etc.) at each node of the transformer substation under normal operation; this simulation data is denoted as […]. ; and with the aligned timing data By comparing the results and subtracting them, we can obtain the F2 norm at time [time value missing]. residual value .

[0028] To accurately evaluate the temporal consistency between the simulated trajectory and the aligned time-series data, a simulation residual scoring function is introduced. This is used to measure the strength of the trend of a transformer area deviating from the normal simulation trajectory. The simulation residual scoring function is specifically expressed as: in, It is the simulation residual scoring function, representing the current time step. The deviation score; It is the window width, which is the historical time interval used for cumulative scoring, and is a parameter pre-configured by the staff; It is a cumulative time variable; It is the squared residual, representing the instantaneous energy of the state deviation; It is the time decay factor, an adjustment coefficient that controls the rate of decrease of the time weight, determined through experimental fitting. In this embodiment, the preferred value range is [value range missing]. ; It is a time weighting item, which adjusts the weight of each moment within the window, making the current moment have a greater weight and the historical moments have a decreasing weight; It is a weighting adjustment term that ensures the denominator of the square root is always positive, avoiding zero values.

[0029] like Exceeding the stability tolerance limit given by the simulation model If so, the transformer area is judged to have an abnormal macroscopic state.

[0030] In this embodiment, step S3 above introduces a disturbance propagation modeling mechanism based on the disturbance driving factor and the physical connection relationship of the transformer area, calculates the disturbance propagation intensity coefficient between nodes, constructs a multi-physical quantity coupling intensity tensor, and obtains the path impact integral quantity through temporal coupling processing, including: To further confirm whether an early warning has been triggered and to enhance the response capability of the transformer substations to potential propagation trends, a disturbance propagation modeling mechanism based on the interconnection relationship of transformer substations is introduced.

[0031] Specifically, calculate the perturbation propagation coefficient between each key node at the current moment. , is represented as: in, Represents a node To the node The disturbance propagation intensity coefficient describes the disturbance from... arrive Instantaneous accessibility; It is a node and The physical distance between them, such as cable length or equivalent electrical distance; It is the line impedance between nodes (including the combined value of resistance and reactance). It is a node and The mixed phase sum of multiple electrical parameters between nodes, i.e. and The sum of the phase differences of all electrical parameters between them; This is the slope control coefficient of the Sigmoid function, which adjusts the sensitivity of the propagation threshold change range. Based on expert experience, the preferred value range in this embodiment is [value range missing]. ; The delay is represented by the following method: by using a multi-source synchronous measurement device (such as a PMU or a high-precision edge sampling device) to capture and record the absolute time of occurrence of the characteristic point (such as voltage change, harmonic spike) of the same disturbance event at node i and node j respectively, the difference between the two absolute times is taken to obtain the propagation delay of the disturbance. It is an electrical constant, and the angular frequency corresponding to the power grid frequency. Reflecting periodic modulation factors, It is the power frequency of the power system in the distribution area, i.e., the grid frequency, which is taken as 50Hz.

[0032] To further couple the propagated information with the nonlinear relationships between multiple physical quantities (i.e., multidimensional electrical data), a multi-physical quantity coupling strength tensor is constructed for each propagation path. Each element in the multi-physical quantity coupling strength tensor... Indicates the first A physical quantity at time... node To the node The strength coupling of propagating disturbances is defined as follows: in, It is the first The first key node Electrical data at time The aligned data values; It is used for amplitude normalization, enabling comparison of differences across different dimensions while avoiding division by zero, based on physical quantities. The minimum stable value for measurement is empirically set, and in this embodiment, the preferred value range is [value range missing]. ; It is a sensitivity adjustment factor used to enhance or suppress the contribution of different physical quantities to the tensor. It is obtained through experimental fitting, and in this embodiment, the preferred value range is [0.5, 2]. , It is a node and nodes physical quantity The historical standard deviation reflects its steady-state fluctuation characteristics; Used to avoid division by zero when the denominator is zero, and to ensure the stability of the logarithmic term; the reference range of values ​​is [range to be specified]. .

[0033] Furthermore, a unified model is used to model the temporal coupling changes of all physical quantities between nodes. The specific implementation formula is as follows: in, From node To the node The path impact integral represents the cumulative disturbance energy intensity of the path over a certain period of time. It is the first The time memory decay factor of an electrical data represents the persistence of the physical quantity's response to disturbances over time. It is matched according to the physical characteristics of the physical quantity. Examples include: voltage: 0.3~0.5, temperature: 0.01~0.1, current: 0.2~0.4, and harmonics: 0.05~0.2.

[0034] It is a node Upper The first derivative of each electrical data is calculated using first-order differences.

[0035] In this embodiment, step S4 above defines an active early warning judgment mechanism based on the deviation score and path impact integral, and the judgment on whether to trigger an early warning includes: By combining the simulation residual scoring function with the path impact integral, an active early warning judgment mechanism is formed, and an early warning trigger function is defined. : in, The node-to-path integral threshold is set dynamically using an empirical quantile method based on skewed distributions (such as P90); once a judgment is made... If the current state is considered to have significantly deviated from the simulation trajectory and formed a disturbance path impact, then the early warning mechanism is actively triggered.

[0036] In this embodiment, after the early warning is triggered in step S5 above, the node risk score is calculated based on the path impact integral of all nodes, and the fault source is determined, including: After the warning is triggered, the source tracing phase begins, and the node risk score is calculated based on the path impact integral of all nodes: in, It is a node As a probability score for the root cause of transformer area failure; It is a node The set of adjacent nodes; It is the index of the adjacent node; It is a node To the neighbor The path importance weights are determined using statistical regression analysis based on fault occurrence frequency and other parameters from historical power distribution area operation data extracted from existing databases. The reference value range is... ; From node To the node The path impact integral; It is a node At any moment Aligned electrical data and nodes At any moment The dot product of the aligned electrical data.

[0037] The node with the highest final score is the suspected source of the fault, thus enabling its location.

[0038] Example 3: The above is an illustrative scheme of a simulation-based active early warning and location method for transformer area faults according to this embodiment. It should be noted that the technical solution of a simulation-based active early warning and location system for transformer area faults and the technical solution of the simulation-based active early warning and location method for transformer area faults described above belong to the same concept. Details not described in detail in the simulation-based active early warning and location system for transformer area faults in this embodiment can be found in the description of the simulation-based active early warning and location method for transformer area faults described above.

[0039] This embodiment also provides a simulation-based active early warning and location system for transformer area faults, including: The multi-source fusion and feature extraction module is used to acquire multi-dimensional electrical data and perform interpolation and alignment to obtain aligned time-series data. Based on the aligned time-series data, a structural interlocking perturbation driving factor construction method is introduced to construct the perturbation driving factor for each node. The simulation benchmark comparison module is used to construct a digital twin simulation model of the transformer substation based on the topology parameters and electrical equipment parameters of the substation, obtain simulation time series data, compare it with the aligned time series data, obtain residual values, and introduce a simulation residual scoring function based on the residual values ​​to determine the deviation score value. The path energy accumulation module is used to introduce a disturbance propagation modeling mechanism based on the disturbance driving factor and the physical connection relationship of the transformer area, calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral quantity through temporal coupling processing. The early warning decision module is used to define an active early warning judgment mechanism based on the deviation score and path impact integral, and to determine whether to trigger an early warning. The network source tracing and location module is used to calculate the node risk score and determine the source of the fault based on the path impact integral of all nodes after an early warning is triggered.

[0040] This embodiment also provides an electronic device applicable to a simulation-based active early warning and location method for transformer area faults, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement a simulation-based active early warning and location method for transformer area faults, as proposed in the above embodiments.

[0041] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements a simulation-based active early warning and location method for transformer area faults as proposed in the above embodiment.

[0042] The storage medium proposed in this embodiment belongs to the same inventive concept as the simulation-based active early warning and location method for transformer area faults proposed in the above embodiment. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0043] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A simulation-based active early warning and location method for transformer area faults, characterized in that, include: Multidimensional electrical data is acquired and interpolated and aligned to obtain aligned time-series data. Based on the aligned time-series data, a structural interlocking disturbance driving factor construction method is introduced to construct the disturbance driving factor for each node. Based on the topology parameters and electrical equipment parameters of the transformer substation, a digital twin transformer substation simulation model is constructed to obtain simulation time series data. This data is then compared with the aligned time series data to obtain residual values. Based on these residual values, a simulation residual scoring function is introduced to determine the deviation score. Based on the disturbance driving factor and the physical connection relationship of the transformer area, a disturbance propagation modeling mechanism is introduced to calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral through temporal coupling processing. Based on the deviation score and the path impact integral, an active early warning judgment mechanism is defined to determine whether an early warning is triggered. After the warning is triggered, the node risk score is calculated based on the path impact integral of all nodes to determine the source of the fault.

2. The simulation-based active early warning and location method for transformer area faults as described in claim 1, characterized in that, The process involves acquiring multidimensional electrical data and performing interpolation and alignment to obtain aligned time-series data. Based on this aligned time-series data, a structural interlocking perturbation driving factor construction method is introduced to construct the perturbation driving factor for each node, including: Based on the aligned time-series data, a comprehensive disturbance driving factor is constructed for each node. Specifically, for each type of electrical data of the node, a single-channel disturbance factor that combines instantaneous acceleration and deviation relative to historical steady state is calculated. The single-channel disturbance factors corresponding to all electrical data categories of the node are weighted and fused to generate a disturbance driving factor that reflects the overall dynamic instability of the node.

3. The simulation-based active early warning and location method for transformer area faults as described in claim 2, characterized in that, Based on the topology parameters and electrical equipment parameters of the transformer substation, a digital twin transformer substation simulation model is constructed to obtain simulation time-series data. This data is then compared with aligned time-series data to obtain residual values. Based on these residual values, a simulation residual scoring function is introduced to determine the deviation score, including: Run the digital twin simulation model to generate simulation time series data of each key node in the transformer area under normal operating conditions; The simulation time series data is compared with the aligned time series data of the corresponding nodes and time points to calculate the residual sequence; Based on the residual sequence, a simulation residual scoring function is introduced to determine the deviation score value of the system's macroscopic state; The simulation residual scoring function is a weighted accumulation of the residuals at each time point within a predefined time window.

4. The simulation-based active early warning and location method for transformer area faults as described in claim 3, characterized in that, Based on the disturbance driving factor and the physical connection relationship of the transformer area, a disturbance propagation modeling mechanism is introduced to calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral through temporal coupling processing, including: Based on the comprehensive disturbance driving factor of each node and the physical connection relationship of the transformer area, the disturbance propagation intensity coefficient between any two key nodes is calculated. The disturbance propagation intensity coefficient is at least related to the comprehensive disturbance driving factor of the source node, the physical distance between nodes, the line impedance characteristics, the electrical phase difference, and the measured disturbance propagation time delay.

5. The simulation-based active early warning and location method for transformer area faults as described in claim 4, characterized in that, The method, based on disturbance driving factors and the physical connection relationship of transformer areas, introduces a disturbance propagation modeling mechanism to calculate the disturbance propagation intensity coefficient between nodes, constructs a multi-physical quantity coupling intensity tensor, and obtains the path impact integral through temporal coupling processing. It also includes: For each type of electrical data between each pair of nodes, a multi-physical quantity coupling strength tensor is constructed based on the disturbance propagation strength coefficient, the instantaneous difference of electrical data, and the steady-state fluctuation characteristics. The path impact integral is calculated by performing attenuation integral processing on the coupling strength tensor of multiple physical quantities along the time dimension.

6. The simulation-based active early warning and location method for transformer area faults as described in claim 5, characterized in that, The aforementioned proactive early warning judgment mechanism, based on deviation score and path impact integral, determines whether an early warning is triggered, including: An active warning is triggered if and only if the deviation score exceeds a preset deviation threshold and there is at least one pair of nodes whose path impact integral exceeds a preset path integral threshold for the pair of nodes.

7. The simulation-based active early warning and location method for transformer area faults as described in claim 6, characterized in that, After the warning is triggered, the node risk score is calculated based on the path impact integral of all nodes to determine the fault source, including: After the warning is triggered, the fault source is traced based on the path impact integral of all node pairs; For each node, the risk score is calculated by combining the path impact integral with each neighboring node, the physical distance between nodes, the directional difference of the node's state vector, and the pre-set path importance weight. Based on the risk scores of all nodes, the node with the highest score is identified as the suspected source of the fault.

8. A simulation-based active early warning and location system for transformer area faults, using the method described in any one of claims 1 to 7, characterized in that, include: The multi-source fusion and feature extraction module is used to acquire multi-dimensional electrical data and perform interpolation and alignment to obtain aligned time-series data. Based on the aligned time-series data, a structural interlocking perturbation driving factor construction method is introduced to construct the perturbation driving factor for each node. The simulation benchmark comparison module is used to construct a digital twin simulation model of the transformer substation based on the topology parameters and electrical equipment parameters of the substation, obtain simulation time series data, compare it with the aligned time series data, obtain residual values, and introduce a simulation residual scoring function based on the residual values ​​to determine the deviation score value. The path energy accumulation module is used to introduce a disturbance propagation modeling mechanism based on the disturbance driving factor and the physical connection relationship of the transformer area, calculate the disturbance propagation intensity coefficient between nodes, construct a multi-physical quantity coupling intensity tensor, and obtain the path impact integral quantity through temporal coupling processing. The early warning decision module is used to define an active early warning judgment mechanism based on the deviation score and path impact integral, and to determine whether to trigger an early warning. The network source tracing and location module is used to calculate the node risk score and determine the source of the fault based on the path impact integral of all nodes after an early warning is triggered.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images