A coal mine user power supply risk analysis method based on digital twinning

Abstract

The application discloses a coal mine user power supply risk analysis method based on digital twinning, comprising the following steps: collecting coal mine power supply node current, voltage and switch state data, obtaining digital twin simulation value, calculating deviation sequence along time sequence; reconstructing the deviation curve, calculating the segmented derivative to generate the deviation evolution trajectory and the change rate sequence; extracting the switch state event to form the event sequence, generating the equidistant time grid data, normalizing and aligning the deviation trajectory; constructing the node connection relationship matrix to map the risk intensity vector, expanding the risk propagation chain along the topological path, accumulating calculation, applying the line impedance and load disturbance attenuation; weighting and summarizing the risk value at the end of the topological link to generate the whole network risk curve; arranging the node risk index sequence to form the structured risk table, and outputting the whole network risk curve and the node risk index. The application realizes dynamic quantification of the whole network power supply risk along the node deviation evolution, and improves the risk analysis precision and the visual decision value.

Description

Technical Field

[0001] This invention relates to the field of power system monitoring and risk analysis, and in particular to a method for analyzing power supply risks for coal mine users based on digital twins. Background Technology

[0002] In the operation and monitoring of coal mine power supply systems, real-time monitoring of the current, voltage, and switch status of each power supply node is crucial for ensuring power supply safety and optimizing dispatch. Current technologies typically rely on single measurement data or simple statistical analysis to assess the power supply status. While these methods are capable of detecting obvious anomalies, they are prone to overlooking the correlation between nodes and potential risk transmission paths when faced with multi-source data coupling, complex power supply network topologies, and load disturbances. This results in incomplete identification of power supply risks and an inability to achieve refined dynamic monitoring and quantification of network-wide risks.

[0003] In recent years, some studies have attempted to introduce digital twin technology or simulation models into power supply system analysis, using the comparison of virtual node states with actual measurement data to identify deviations. However, existing methods generally only perform static comparisons of single nodes, failing to form a continuous time series deviation evolution trajectory. They also lack a unified mapping mechanism for deviation change rate and asynchronous sampling data, resulting in insufficient capture of dynamic characteristics of node deviations. Furthermore, it is difficult to effectively integrate the attenuation correction of risk propagation by the topological relationship between nodes and line parameters, thus limiting the accuracy of network-wide risk propagation and node-level risk quantification.

[0004] In the abnormal risk prediction and decision support stage, existing technologies mostly rely on fixed thresholds or static rules, lacking dynamic accumulation and deviation evolution modeling across nodes and time. They cannot achieve multi-layer risk propagation and sensitive node tracing under complex load disturbances or topology changes. At the same time, they fail to structure and output the risk curves of the entire network and node risk indicators, which limits the operability and decision reference value of risk assessment results in actual operation and maintenance scheduling.

[0005] Therefore, how to provide a method for analyzing the power supply risks of coal mine users based on digital twins is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] One objective of this invention is to propose a method for analyzing power supply risks for coal mine users based on digital twins. This invention generates deviation sequences, constructs deviation evolution trajectories and event sequences, maps risk intensity vectors, and propagates risks along topological paths to achieve dynamic modeling and continuous risk analysis of the entire power supply status, improves the accuracy of risk quantification, and enhances the visualization and decision-making reference value of the entire network.

[0007] A method for analyzing power supply risks for coal mine users based on digital twins according to an embodiment of the present invention includes the following steps: Collect current, voltage, and switch status data of power supply nodes in coal mines, obtain simulated current, voltage, and status of corresponding nodes in the digital twin system, calculate the deviation vector of each node along the time series, perform sliding integral accumulation, and generate a deviation sequence. The deviation sequence is reconstructed as a continuous curve, and piecewise derivative calculations are performed along the curve to obtain the deviation evolution trajectory and rate of change sequence. The power supply node switch state change events are extracted to form an event time series. Asynchronous sampled current and voltage data are mapped to the event time interval. Interpolation reconstruction is performed to generate equidistant time grid data. The data is then normalized and the units are converted to obtain time-aligned data. Match time-aligned data with the corresponding deviation evolution trajectory and rate of change sequence; Construct a power supply node connection matrix, map the matched node deviation trajectory into a risk intensity vector, expand the risk propagation chain of each node along the topology path, perform cumulative calculations in sequence, and introduce line impedance and load disturbance coefficient at each node for attenuation correction; The risk values ​​at the end of each topology link are time-weighted and aggregated to generate a network-wide risk curve. Perform time series labeling and node index mapping on the risk curve of the entire network, organize the risk indicator sequence of each node, generate a structured risk table, and output the risk curve of the entire network and the risk indicators of each node.

[0008] Optionally, the data acquisition and deviation calculation steps include: Collect the current, voltage and switch status signals of each power supply node, obtain the simulated current, simulated voltage and simulated status signals of the corresponding node in the digital twin system, calculate the difference between the measured value and the simulated value node by node along the time series, and generate the deviation vector. The deviation vector is accumulated by performing a sliding integral along the time series, and the deviation vectors of consecutive time points of each node are superimposed in time order to form a cumulative deviation sequence. A piecewise moving average is applied to the cumulative deviation sequence to eliminate instantaneous fluctuations and maintain sequence continuity; The processed deviation sequences are sorted and labeled according to node index and acquisition time to generate a set of deviation sequences that are continuous in time and have consistent node correspondence. Perform a normalization mapping on the set of deviation sequences to standardize the deviations of each node under a unified dimension, so as to facilitate subsequent curve reconstruction; An outlier check is performed on the normalized deviation sequence, and points with deviations exceeding the threshold are included in the sliding integral accumulation calculation for correction to ensure the continuity of the sequence and the consistency of the deviation accumulation. Output the set of deviation sequences for all power supply nodes, providing input data for the calculation of deviation evolution trajectory and rate of change.

[0009] Optionally, the deviation sequence curve reconstruction step includes: The node deviation sequence is mapped onto a continuous time axis in the order of acquisition time to construct a continuous node deviation curve. Piecewise interpolation is performed on the node deviation curve to smoothly connect the differences between discontinuous points and generate a continuous deviation curve. The derivative is calculated piecewise along the continuous deviation curve according to the time series to obtain the vector of the rate of change of deviation at each time point; Smoothing filters are applied to the deviation rate of change vector to eliminate instantaneous abnormal fluctuations while preserving the deviation evolution trend; Organize the deviation change rate vectors of all nodes according to their node indices to form a set of deviation evolution trajectories and a set of change rate sequences; In the set of deviation evolution trajectories, the slope of the curve is marked, and the time period exceeding the preset change range is recorded as the high evolution rate interval. The output node deviation evolution trajectory set and change rate sequence provide input data for time alignment and risk mapping.

[0010] Optionally, the time alignment and event interpolation steps include: The power supply node switch status change events are arranged in chronological order of acquisition time to generate a node event time series. The asynchronously sampled current and voltage data from each node are mapped to the corresponding time intervals according to the event time sequence; Perform piecewise interpolation within the mapped time interval to interpolate non-equidistant sampling points into equidistant time grid data; Unit unification and normalization are performed on the equidistant time grid data to ensure that the current and voltage data of each node are processed under a unified dimension; The normalized time grid data is organized according to the node index to form a node time-aligned data set. Smoothing filtering is performed on the interpolated data in the node time-aligned dataset to eliminate sampling noise and transient outliers, thus maintaining data continuity; The output is a set of time-aligned node data, which provides input for deviation evolution trajectory matching and risk mapping.

[0011] Optionally, the step of matching the time-aligned data with the deviation evolution trajectory includes: Arrange the time-aligned data sets and the deviation evolution trajectory sets according to the node index; Based on the timestamp, the time-aligned data of each node is mapped to consecutive time points of the deviation evolution trajectory, so as to achieve a one-to-one correspondence between data points; Perform linear or higher-order interpolation on node data with time differences or incomplete sampling to ensure synchronization between the biased sequence and the time-aligned data; The matched node time-aligned data is combined with the deviation evolution trajectory and rate of change sequence to form a node deviation mapping matrix; Smoothing is performed on abnormal deviation points in the node deviation mapping matrix to eliminate the impact of instantaneous abnormal fluctuations on risk calculation. Organize the data into a complete matching dataset according to the node index and time series, and record the deviation value, deviation change rate and alignment current and voltage data for each time point; The output includes a matched node deviation mapping matrix and a time-aligned dataset, providing input for risk mapping and topology link calculation.

[0012] Optionally, the risk mapping and topology link accumulation calculation steps include: Construct a power supply node connection relationship matrix, and generate an adjacency matrix based on the electrical topology and line connection information between nodes; The matched node deviation trajectory is mapped to a risk intensity vector, and the deviation value and rate of change at each time point are recorded. The risk propagation chain is sequentially extended from the source node to the downstream node along each topological path. The risk value of each node is cumulatively calculated, and the risk contribution of the preceding node is superimposed level by level according to the topological order. By introducing line impedance coefficient and load disturbance coefficient at each node, the cumulative risk value is attenuated and adjusted to reflect the impact of power transmission loss and load fluctuation. A weighted summation is performed on the risk values ​​accumulated by the same node across multiple topology links to generate a comprehensive risk value for the node; The overall risk values ​​of the nodes are arranged in topological order to form a network-wide risk matrix, and the risk status of each node at each time point is recorded. The output node risk intensity vector set and the network-wide risk matrix provide a data foundation for calculating the network-wide risk curve and generating structured risk tables.

[0013] Optionally, the step of generating the network-wide risk curve includes: Extract the cumulative risk value sequence of each topology link terminal node, and arrange them according to the time series to form a terminal risk matrix; Based on the time weighting factor, perform time-by-time weighting on the terminal risk matrix to calculate the weighted risk value of each node at different time points; A sliding accumulation process is performed on the weighted risk values ​​of each node along the time series to generate a time-continuous risk curve segment. The continuous risk curve segments of each node are superimposed and merged in topological order to form the risk curve of the entire network; The risk curve of the entire network is smoothed to eliminate instantaneous spikes while retaining dynamic trends and abnormal fluctuation characteristics; Record the relationship between the node index and risk value of the network risk curve at each time point, and generate a node risk indicator mapping table; Output the network-wide risk curve and node risk indicator mapping table to provide complete data input for the generation of structured risk tables.

[0014] Optionally, the steps of organizing the network-wide risk curves and generating structured risk tables include: Node indexing is performed on the time series of the risk curve of the entire network, and a mapping relationship is established between the risk value of each node and the corresponding time point; Based on node index and time series, the risk curve of the entire network is divided into risk index sequences of each node, and the risk value and risk change trend of each node are recorded at each time point. Based on node type, power supply level, and electrical topology, the risk index sequences of each node are classified and sorted to form a hierarchical node risk matrix; Numerical normalization and interval mapping are performed on the hierarchical node risk matrix to generate a structured risk table with a uniform scale; Add node identifier, risk level, and time series index fields to the structured risk table to form a complete set of node risk indicators; Output structured risk tables and corresponding network-wide risk curves to provide quantitative basis for risk analysis, early warning and decision-making.

[0015] The beneficial effects of this invention are: (1) This invention collects current, voltage and switch status data of coal mine power supply nodes and combines them with digital twin simulation values ​​to calculate deviation sequences, construct deviation evolution trajectories and event sequences along the time series, realizes topological association modeling of power supply nodes in the whole network, effectively characterizes the dynamic deviation and risk propagation relationship between nodes, and improves the completeness and accuracy of risk expression under complex power supply networks.

[0016] (2) This invention generates equidistant time grid data and performs normalization and alignment, maps the node deviation trajectory into a risk intensity vector, unfolds the risk propagation chain along the topology path, applies line impedance and load disturbance attenuation, and performs weighted summarization of the risks of the entire network, thereby realizing continuous characterization and dynamic quantification of risk evolution and improving the ability to capture risk mutations and cumulative effects.

[0017] (3) This invention forms a structured risk table by organizing the node risk index sequence, outputs the network risk curve and node risk index, realizes the joint tracking and visualization of risks in time and space dimensions, improves the accuracy of network power supply risk analysis and decision reference value, and enhances the interpretability and practicality of the analysis results. Attached Figure Description

[0018] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is an overall flowchart of a digital twin-based coal mine user power supply risk analysis method proposed in this invention. Figure 2 This is a schematic diagram illustrating the generation and evolution trajectory of the deviation sequence of a digital twin-based coal mine user power supply risk analysis method proposed in this invention. Figure 3 This diagram illustrates the network-wide risk calculation and node risk indicators for a digital twin-based coal mine user power supply risk analysis method proposed in this invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0020] refer to Figure 1-3 A method for analyzing power supply risks for coal mine users based on digital twins includes the following steps: Collect current, voltage, and switch status data of power supply nodes in coal mines, obtain the simulated current, voltage, and status of corresponding nodes in the digital twin system, calculate the deviation vector of each node along the time series, and perform sliding integral accumulation to generate the deviation sequence. The deviation sequence is reconstructed as a continuous curve, and piecewise derivative calculations are performed along the curve to obtain the deviation evolution trajectory and rate of change sequence. The power supply node switch state change events are extracted to form an event time series. Asynchronous sampled current and voltage data are mapped to the event time interval, and interpolation reconstruction is performed to generate equidistant time grid data. The data is then normalized and the units are converted to obtain time-aligned data. Match time-aligned data with the corresponding deviation evolution trajectory and rate of change sequence; Construct a power supply node connection matrix, map the matched node deviation trajectory into a risk intensity vector, expand the risk propagation chain of each node along the topology path, perform cumulative calculations in sequence, and introduce line impedance and load disturbance coefficient at each node for attenuation correction; The risk values ​​at the end of each topology link are time-weighted and aggregated to generate a network-wide risk curve. Perform time series labeling and node index mapping on the risk curve of the entire network, organize the risk indicator sequence of each node, generate a structured risk table, and output the risk curve of the entire network and the risk indicators of each node.

[0021] In this embodiment, the data acquisition and deviation calculation steps include: The current, voltage, and switch status information of each node in the coal mine power supply system are collected in real time through current sensors, voltage sensors, and switch status acquisition devices. The acquisition interval can be set according to the load fluctuation frequency to ensure that the data covers the entire power supply cycle. During the data acquisition process, the actual current, voltage, and switch status values ​​are compared with the simulated values ​​of the corresponding nodes in the digital twin system, and the node deviation vector is calculated. The deviation vector is accumulated and statistically analyzed along the time series, and integral accumulation is performed within the sliding time window to form a continuous deviation sequence; During the integration process, considering data noise and instantaneous fluctuations, the deviation data is preprocessed using filtering and smoothing algorithms to ensure that the deviation sequence reflects the long-term deviation trend of the nodes rather than instantaneous fluctuations. Deviation sequences can accurately describe the dynamic deviation of each node during operation, providing basic data for subsequent deviation evolution trajectories and risk calculations; After data collection, a preliminary quality check is performed on the deviation data of each node, including data integrity, continuity, and outlier identification. Missing data are filled in using linear or spline interpolation, and outliers are corrected by thresholding or neighborhood averaging to ensure that the deviation sequence is continuous and smooth on the time axis. Meanwhile, the deviation data is weighted in the calculation process by taking into account the node's operating status and load characteristics, so as to reflect the node's risk sensitivity under different load conditions.

[0022] In this embodiment, the deviation sequence curve reconstruction step includes: The collected deviation sequence is reconstructed into a continuous curve, and the discrete deviation points are connected into a continuous curve using a smooth interpolation method. During curve reconstruction, polynomial fitting or spline interpolation methods can be used to smooth the trend of node deviation over time while retaining the characteristics of abrupt change points to reflect potential abnormal events. Piecewise derivatives are calculated along the curve to obtain a sequence of deviation change rates, which is used to quantify the instantaneous rate of change of nodal deviations. The piecewise derivative is calculated to locally fit the curve within each time period in order to capture local fluctuations and trend changes in the deviation. Deviation evolution trajectory and rate of change sequence can intuitively present the dynamic characteristics of node deviation, provide input data for risk propagation modeling, and support subsequent time alignment and event matching.

[0023] In this embodiment, the time alignment and event interpolation steps include: Extract the switch state change events of each power supply node to form an event time series, which is used to identify key operation nodes of the system; Asynchronously sampled current and voltage data are mapped to event time intervals, and interpolation is performed on missing or irregular sampled data to generate equidistant time grid data. During the interpolation process, linear interpolation, spline interpolation, or local weighted regression methods are used to fill in data points with uneven sampling intervals, so that the time series data is continuously available at the grid points; Normalization is performed on the generated equidistant time grid data to unify data from different nodes and different units into a comparable numerical range, and unit conversion is performed to ensure data consistency. During the normalization and unit conversion process, a node attribute weighting method is adopted to incorporate node importance and load sensitivity into time-aligned data, so that subsequent risk calculations can reflect the different impact levels of nodes.

[0024] In this embodiment, the step of matching time-aligned data with deviation evolution trajectory includes: The normalized time-aligned data is matched one-to-one with the deviation evolution trajectory and the rate of change sequence, and the node state data at each time point is precisely aligned with the deviation vector. During the matching process, timestamps are strictly mapped to ensure that the event sequence, collected data, and deviation sequence are synchronized on the same timeline; For cases with missing data or deviation trajectory delays, linear interpolation or neighborhood weighted filling is used to ensure that the matching results are continuous and complete. After matching is completed, a deviation evolution matrix with multiple nodes and multiple time steps is formed, which provides a reliable data foundation for subsequent risk mapping and topology link accumulation calculation, and supports the modeling of dynamic dependencies between nodes.

[0025] In this embodiment, the risk mapping and topology link accumulation calculation steps include: Based on the topology of the coal mine power supply system, a node connection matrix is ​​constructed, where the elements in the matrix represent the strength and path dependency of the electrical connections between nodes. Time-aligned data and deviation evolution trajectories are mapped to node risk intensity vectors, and the risk propagation chain of each node is unfolded along the topological path; Cumulative calculations are performed sequentially along the propagation chain, and line impedance and load disturbance coefficients are introduced at each node for attenuation correction to reflect the physical constraints of risk transmission and the attenuation effect of inter-node influence in the actual power grid. The cumulative calculation of the risk propagation chain is strictly performed in the order of the topological path to ensure that the risk accumulates gradually from the upstream node to the downstream node; For nodes with loops or multiple paths, the cumulative risk value is calculated using a weighted average method to ensure the continuity and consistency of risk transmission across the entire network. During the cumulative calculation process, the risk impact weights between nodes can be dynamically adjusted based on the node load level, line impedance, and topological location, enhancing the refined expressive capability of the risk propagation model.

[0026] In this embodiment, the step of generating the network-wide risk curve includes: The risk values ​​of the terminal nodes of each topology link are time-weighted and aggregated to generate a risk curve for the entire network. Time-weighted aggregation is achieved by assigning different weights to the historical risk values ​​of nodes. The weights can be determined based on changes in node load, fluctuations in deviation, and the importance of switching events. The generated network-wide risk curve presents the overall risk level of the entire network at various points in time, and smoothing processes are used to avoid the interference of instantaneous fluctuations on the overall trend. The network-wide risk curve can be used to display the evolution of network-wide risks, providing dispatchers with an intuitive reference and enabling dynamic risk visualization.

[0027] In this embodiment, the steps of organizing the network-wide risk curves and generating structured risk tables include: Perform node index mapping and time series labeling on the risk curve of the entire network, and organize the risk value of each node at each time point into a structured table to form a node risk indicator sequence; The structured risk table includes fields such as node identifier, time step, deviation value, rate of change, and risk value, which facilitates subsequent analysis and visualization. By using structured tables, statistical analysis and trend prediction of risk evolution at each node can be performed, enabling hierarchical output of risk levels. The output network-wide risk curves and node risk index tables can serve as a reference for power supply dispatch risk management, early warning, and decision-making, improving the accuracy and visualization of risk quantification analysis.

[0028] Example 1: To verify the feasibility of this invention in practice, it was applied to a multi-node collaborative risk analysis scenario in a coal mine user power supply system. This scenario involves the collection and analysis of power supply operation data for various types of nodes, including substation nodes, distribution nodes, and power consumption terminal nodes. In actual operation, the current, voltage, and switch status of each node are affected by load fluctuations, equipment operation, and power consumption behavior, exhibiting dynamic fluctuation characteristics. Simultaneously, different nodes are coupled through power supply topology and links. When a local node experiences abnormal deviations or load anomalies, the risk may propagate along the topology to other nodes. Traditional single-node threshold analysis methods are insufficient to characterize cross-node risk propagation, resulting in incomplete risk identification, unobservable propagation paths, and delayed response.

[0029] In this scenario, real-time data on current, voltage, and switch status of each node are collected, and deviation vectors are calculated by combining the node simulation values ​​in the digital twin system. After time alignment, denoising, and normalization, the collected data generates multi-node state vectors to characterize the overall operating status of each node. Based on the digital twin simulation model and node topology, neighborhood-weighted risk propagation calculations are performed. Each node acquires and weights the state information of its neighboring nodes to form a node risk representation. Further spatiotemporal joint analysis of the risk evolution sequence is conducted, and multi-scale collaborative risk trajectories are generated by combining deviation integral curves and risk potential field constraints, achieving dynamic modeling and visual tracking of network-wide risks.

[0030] To evaluate the technical effectiveness of this invention, typical operating conditions such as sudden load surges, load declines, and sudden power fluctuations were selected. Under the same data acquisition conditions, the traditional threshold method, the comparative method based on graph structure modeling, and the method of this invention were compared and analyzed. The evaluation focused on indicators such as risk identification accuracy, false alarm rate, false negative rate, average response time, and observability of risk propagation paths. Experimental results show that in the scenario of sudden load surges, the traditional method achieved a risk identification accuracy of only 79%, with false alarm and false negative rates of 12% and 15%, respectively, and could not effectively display the risk propagation path. The comparative method improved the accuracy to 87%, with false alarm and false negative rates of 8% and 10%, respectively, but the propagation path remained incomplete. The method of this invention achieved an accuracy of 95%, with false alarm and false negative rates of 4% and 5%, respectively, and could fully visualize the risk propagation link along the topology. In the scenarios of load declines and sudden power fluctuations, the method of this invention also demonstrated high accuracy and low false alarm rate, and could dynamically track risk propagation between nodes, achieving a unified expression of risk across the entire network.

[0031] The following are quantitative evaluation results from multiple experiments under different power supply node operating conditions, verifying that the present invention has good stability and applicability under various typical risk propagation and load fluctuation conditions.

[0032] Table 1: Performance Comparison of Multi-Node Risk Analysis for Coal Mine User Power Supply Systems

[0033] As shown in Table 1, the traditional threshold method performs the worst in terms of risk identification accuracy, false alarm rate, and false negative rate, and it cannot provide the risk propagation path, indicating its insufficient adaptability in multi-node coupled disturbance scenarios. The graph structure method improves accuracy, but the risk propagation path is still incomplete, only achieving weak observability analysis. The method of this invention performs excellently under various operating conditions, with a stable risk identification accuracy of 93-95%, significantly reduced false alarm and false negative rates, the shortest average response time, and the ability to fully visualize the risk propagation path along the topology, enabling dynamic modeling and real-time tracking of risks across the entire network.

[0034] The above results further demonstrate that this invention, combining digital twin simulation, deviation vector analysis, and a neighborhood-weighted risk propagation mechanism, can not only accurately identify potential risk nodes in coal mine user power supply systems but also clearly depict the risk propagation path within the topology, improving the real-time performance and stability of risk identification. Compared to traditional threshold analysis methods and partial graph model methods, this invention has significant advantages in multi-node collaborative risk modeling, propagation path visualization, and dynamic tracking of network-wide risks, providing reliable data support and technical assurance for safety monitoring, risk warning, and scheduling decisions in coal mine user power supply systems.

[0035] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for analyzing power supply risks to coal mine users based on digital twins, characterized in that, Includes the following steps: Collect current, voltage, and switch status data of power supply nodes in coal mines, obtain simulated current, voltage, and status of corresponding nodes in the digital twin system, calculate the deviation vector of each node along the time series, perform sliding integral accumulation, and generate a deviation sequence. The deviation sequence is reconstructed as a continuous curve, and piecewise derivative calculations are performed along the curve to obtain the deviation evolution trajectory and rate of change sequence. The power supply node switch state change events are extracted to form an event time series. Asynchronous sampled current and voltage data are mapped to the event time interval. Interpolation reconstruction is performed to generate equidistant time grid data. The data is then normalized and the units are converted to obtain time-aligned data. Match time-aligned data with the corresponding deviation evolution trajectory and rate of change sequence; Construct a power supply node connection matrix, map the matched node deviation trajectory into a risk intensity vector, expand the risk propagation chain of each node along the topology path, perform cumulative calculations in sequence, and introduce line impedance and load disturbance coefficient at each node for attenuation correction; The risk values ​​at the end of each topology link are time-weighted and aggregated to generate a network-wide risk curve. Perform time series labeling and node index mapping on the risk curve of the entire network, organize the risk indicator sequence of each node, generate a structured risk table, and output the risk curve of the entire network and the risk indicators of each node.

2. The method for analyzing power supply risks for coal mine users based on digital twins according to claim 1, characterized in that, The data acquisition and deviation calculation steps include: Collect the current, voltage and switch status signals of each power supply node, obtain the simulated current, simulated voltage and simulated status signals of the corresponding node in the digital twin system, calculate the difference between the measured value and the simulated value node by node along the time series, and generate the deviation vector. The deviation vector is accumulated by performing a sliding integral along the time series, and the deviation vectors of consecutive time points of each node are superimposed in time order to form a cumulative deviation sequence. A piecewise moving average is applied to the cumulative deviation sequence to eliminate instantaneous fluctuations and maintain sequence continuity; The processed deviation sequences are sorted and labeled according to node index and acquisition time to generate a set of deviation sequences that are continuous in time and have consistent node correspondence. Perform a normalization mapping on the set of deviation sequences to standardize the deviations of each node under a unified dimension, so as to facilitate subsequent curve reconstruction; An outlier check is performed on the normalized deviation sequence, and points with deviations exceeding the threshold are included in the sliding integral accumulation calculation for correction to ensure the continuity of the sequence and the consistency of the deviation accumulation. Output the set of deviation sequences for all power supply nodes, providing input data for the calculation of deviation evolution trajectory and rate of change.

3. The method for analyzing power supply risks for coal mine users based on digital twins according to claim 2, characterized in that, The deviation sequence curve reconstruction step includes: The node deviation sequence is mapped onto a continuous time axis in the order of acquisition time to construct a continuous node deviation curve. Piecewise interpolation is performed on the node deviation curve to smoothly connect the differences between discontinuous points and generate a continuous deviation curve. The derivative is calculated piecewise along the continuous deviation curve according to the time series to obtain the vector of the rate of change of deviation at each time point; Smoothing filters are applied to the deviation rate of change vector to eliminate instantaneous abnormal fluctuations while preserving the deviation evolution trend; Organize the deviation change rate vectors of all nodes according to their node indices to form a set of deviation evolution trajectories and a set of change rate sequences; In the set of deviation evolution trajectories, the slope of the curve is marked, and the time period exceeding the preset change range is recorded as the high evolution rate interval. The output node deviation evolution trajectory set and change rate sequence provide input data for time alignment and risk mapping.

4. The method for analyzing power supply risks of coal mine users based on digital twins according to claim 3, characterized in that, The time alignment and event interpolation steps include: The power supply node switch status change events are arranged in chronological order of acquisition time to generate a node event time series. The asynchronously sampled current and voltage data from each node are mapped to the corresponding time intervals according to the event time sequence; Perform piecewise interpolation within the mapped time interval to interpolate non-equidistant sampling points into equidistant time grid data; Unit unification and normalization are performed on the equidistant time grid data to ensure that the current and voltage data of each node are processed under a unified dimension; The normalized time grid data is organized according to the node index to form a node time-aligned data set. Smoothing filtering is performed on the interpolated data in the node time-aligned dataset to eliminate sampling noise and transient outliers, thus maintaining data continuity; The output is a set of time-aligned node data, which provides input for deviation evolution trajectory matching and risk mapping.

5. The method for analyzing power supply risks for coal mine users based on digital twins according to claim 4, characterized in that, The time-aligned data and deviation evolution trajectory matching steps include: Arrange the time-aligned data sets and the deviation evolution trajectory sets according to the node index; Based on the timestamp, the time-aligned data of each node is mapped to consecutive time points of the deviation evolution trajectory, so as to achieve a one-to-one correspondence between data points; Perform linear or higher-order interpolation on node data with time differences or incomplete sampling to ensure synchronization between the biased sequence and the time-aligned data; The matched node time-aligned data is combined with the deviation evolution trajectory and rate of change sequence to form a node deviation mapping matrix; Smoothing is performed on abnormal deviation points in the node deviation mapping matrix to eliminate the impact of instantaneous abnormal fluctuations on risk calculation. Organize the data into a complete matching dataset according to the node index and time series, and record the deviation value, deviation change rate and alignment current and voltage data for each time point; The output includes a matched node deviation mapping matrix and a time-aligned dataset, providing input for risk mapping and topology link calculation.

6. The method for analyzing power supply risks for coal mine users based on digital twins according to claim 5, characterized in that, The risk mapping and topology link accumulation calculation steps include: Construct a power supply node connection relationship matrix, and generate an adjacency matrix based on the electrical topology and line connection information between nodes; The matched node deviation trajectory is mapped to a risk intensity vector, and the deviation value and rate of change at each time point are recorded. The risk propagation chain is sequentially extended from the source node to the downstream node along each topological path. The risk value of each node is cumulatively calculated, and the risk contribution of the preceding node is superimposed level by level according to the topological order. By introducing line impedance coefficient and load disturbance coefficient at each node, the cumulative risk value is attenuated and adjusted to reflect the impact of power transmission loss and load fluctuation. A weighted summation is performed on the risk values ​​accumulated by the same node across multiple topology links to generate a comprehensive risk value for the node; The overall risk values ​​of the nodes are arranged in topological order to form a network-wide risk matrix, and the risk status of each node at each time point is recorded. The output node risk intensity vector set and the network-wide risk matrix provide a data foundation for calculating the network-wide risk curve and generating structured risk tables.

7. The method for analyzing power supply risks of coal mine users based on digital twins according to claim 6, characterized in that, The steps for generating the network-wide risk curve include: Extract the cumulative risk value sequence of each topology link terminal node, and arrange them according to the time series to form a terminal risk matrix; Based on the time weighting factor, perform time-by-time weighting on the terminal risk matrix to calculate the weighted risk value of each node at different time points; A sliding accumulation process is performed on the weighted risk values ​​of each node along the time series to generate a time-continuous risk curve segment. The continuous risk curve segments of each node are superimposed and merged in topological order to form the risk curve of the entire network; The risk curve of the entire network is smoothed to eliminate instantaneous spikes while retaining dynamic trends and abnormal fluctuation characteristics; Record the relationship between the node index and risk value of the network risk curve at each time point, and generate a node risk indicator mapping table; Output the network-wide risk curve and node risk indicator mapping table to provide complete data input for the generation of structured risk tables.

8. The method for analyzing power supply risks of coal mine users based on digital twins according to claim 7, characterized in that, The steps for compiling the network-wide risk curve and generating the structured risk table include: Node indexing is performed on the time series of the risk curve of the entire network, and a mapping relationship is established between the risk value of each node and the corresponding time point; Based on node index and time series, the risk curve of the entire network is divided into risk index sequences of each node, and the risk value and risk change trend of each node are recorded at each time point. Based on node type, power supply level, and electrical topology, the risk index sequences of each node are classified and sorted to form a hierarchical node risk matrix; Numerical normalization and interval mapping are performed on the hierarchical node risk matrix to generate a structured risk table with a uniform scale; Add node identifier, risk level, and time series index fields to the structured risk table to form a complete set of node risk indicators; Output structured risk tables and corresponding network-wide risk curves to provide quantitative basis for risk analysis, early warning and decision-making.

9. A method for analyzing power supply risks for coal mine users based on digital twins according to claim 8, characterized in that, The risk indicator sequence analysis steps include: Perform sliding window statistics on the node risk indicator sequence in the structured risk table to calculate the average risk value and volatility of each node within the time window; Based on the average risk value and volatility, a node risk level sequence is calculated to form a dynamic node risk curve; Perform threshold filtering on the dynamic curve of node risk, mark high-risk fluctuation ranges, and record the corresponding time period and node index; Generate a risk change trend matrix for all network nodes, sort the matrix by risk level and time series, and form a risk trend set that can be used for risk evolution analysis; The system outputs dynamic curves of node risk, sequences of node risk levels, and sets of risk trends, providing data support for deviation evolution and risk prediction.

10. A method for analyzing power supply risks for coal mine users based on digital twins according to claim 9, characterized in that, The steps for visualizing and outputting the risk evolution include: The node risk dynamic curve, the network-wide risk curve, and the structured risk table are mapped to each other at a unified time scale to generate a comprehensive risk dataset. Perform visualization mapping on the comprehensive risk dataset according to topological relationships and node risk levels to generate two-dimensional or three-dimensional risk evolution charts; Record node indexes, risk values, risk levels, and corresponding time points in the comprehensive risk dataset to form a traceable data output file; The generated risk evolution charts are provided together with traceable data output files, providing an intuitive presentation and data storage of the entire power supply risk analysis.

Images