Safety early warning intelligent evaluation method and system for power distribution equipment

Abstract

This application provides a method and system for intelligent assessment of safety early warning for power distribution equipment, relating to the field of equipment safety early warning. It addresses the technical problem in existing technologies where the lack of systematic analysis of inter-equipment relationships in power distribution equipment safety assessments leads to weak forward-looking capabilities. The method includes: creating a digital twin for each power distribution device and constructing a dynamic causal graph based on the digital twin; each digital twin performing a status assessment based on operational status data and generating a status score; when the status score exceeds a first threshold, performing risk propagation deduction based on the dynamic causal graph and outputting a risk propagation path; and triggering a tiered early warning based on the risk propagation path. This application is used in the safety assessment and early warning process for power distribution equipment.

Description

Technical Field

[0001] This invention belongs to the field of equipment safety early warning, specifically a method and system for intelligent assessment of safety early warning of power distribution equipment. Background Technology

[0002] Power distribution equipment is a critical component of the power system, and its safe and stable operation directly affects the reliability of power supply. With the expansion of power system scale and the increase in complexity, the need for safety assessment of power distribution equipment is becoming increasingly urgent, and traditional safety assessment methods are gradually becoming inadequate to meet actual operation and maintenance requirements.

[0003] In existing technologies, safety assessments of power distribution equipment often focus on the condition monitoring and fault diagnosis of individual devices, lacking a systematic analysis of the interrelationships between devices, resulting in insufficient ability to predict risk propagation paths. Specifically, existing methods typically determine the existence of fault risks based solely on the operating parameters (such as temperature and current) of individual devices, neglecting the interrelationships between devices in the power distribution network, such as electrical connections and physical locations. When one device malfunctions, the risk may spread to other devices through these connections, and existing methods cannot clearly trace this propagation logic, let alone proactively predict the scope and impact of risk spread. Summary of the Invention

[0004] This application provides a method and system for intelligent assessment of safety early warning of power distribution equipment, which solves the technical problem of insufficient foresight in risk early warning of power distribution equipment in the prior art.

[0005] To achieve the above objectives, this application adopts the following technical solution: Firstly, it provides intelligent assessment methods for safety early warning of power distribution equipment, including: Create a digital twin for each power distribution device and build a dynamic cause-effect graph based on the digital twin; Each digital twin performs a status assessment based on operational status data and generates a status score; When the state score exceeds the first threshold, risk propagation deduction is performed based on the dynamic causal graph, and the risk propagation path is output. Trial-level early warnings are triggered based on the risk transmission path.

[0006] Based on the above technical solutions, the intelligent assessment method for safety early warning of power distribution equipment provided in this application creates a digital twin of the power distribution equipment, which can map the physical state and operating characteristics of the equipment in real time and accurately, ensuring the accuracy of the assessment results. Simultaneously, the construction of a dynamic cause-effect graph clearly outlines the causal relationships between various components of the equipment and related systems, providing a logical framework for risk propagation simulation. Furthermore, the threshold triggering mechanism based on state scoring achieves a quantitative standard for early warning activation, avoiding false alarms or missed alarms caused by subjective judgment, and improving the timeliness and targeting of early warnings. In addition, the simulation and output of risk propagation paths can proactively reveal the scope and impact of risk spread, predict risk diffusion trends in advance, and provide a basis for formulating precise prevention and control measures.

[0007] Furthermore, the creation of a digital twin for each power distribution device includes: Basic parameters of each power distribution device are collected and entered into the basic information database of the digital twin; the basic parameters include equipment model, equipment service life, equipment rated voltage and rated current; Sensors are deployed to collect operating status data of each power distribution device, and the operating status data is transmitted to the corresponding digital twin for storage through a data transmission channel; the operating status data includes current, voltage, temperature, vibration and discharge pulse.

[0008] Furthermore, the construction of a dynamic causal graph based on a digital twin includes: Obtain historical equipment fault records; for each fault event, sort the affected power distribution equipment by alarm timestamp and generate an equipment alarm sequence. The fault propagation direction is obtained by taking the device that first alarms in the device alarm sequence as the starting point and the devices that subsequently alarm as the propagation direction. The digital twins of each power distribution device are mapped to nodes, and the node attributes include device identifier, device type, and device status score; The initial edge connection relationship between nodes is established based on the electrical topology between devices, and the edge direction is the fault propagation direction; The edge initialization weight W is set according to the reciprocal of the physical distance between the devices. init ; Based on a preset time interval, or when the rate of change of the device's status score exceeds a threshold, the edge weights are dynamically updated to obtain a dynamic causal graph.

[0009] Furthermore, the formula for calculating the edge weight is: W = W init ×(1+θ×Score source ); where Score sourceθ represents the state score of the upstream node determined by the edge direction, and θ represents the weight correction coefficient.

[0010] Furthermore, the digital twin performs state assessment based on real-time data, including: The operating status data are preprocessed to obtain multidimensional signal features; the multidimensional signal features include the proportion of abnormal pulse energy E. abn Vibration steepness coefficient K ur and temperature gradient accumulation T acc ; Adjust the equipment health threshold H based on the equipment's service life Y. th ; Adjust the temperature threshold T based on the current load rate. th ; The device status score is calculated based on multidimensional signal characteristics, device health threshold, and temperature threshold.

[0011] Furthermore, the preprocessing of the operating status data includes: (1) Extraction of abnormal pulse energy ratio features: The current and voltage data are decomposed into sub-signals of multiple frequency bands using wavelet packets; The sum of energy within the preset frequency band after statistical decomposition is used to obtain the first energy sum. Calculate the total energy across all frequency bands to obtain the second total energy. Divide the sum of the first energy by the sum of the second energy to obtain the percentage of abnormal pulse energy. (2) Extraction of vibration steepness coefficient features: Acquire N time-domain sampling data of the vibration signal, and calculate the mean and standard deviation of the sampling data; The difference between the sampled value and the mean in each time domain is raised to the fourth power, summed, and then divided by the number of samples N to obtain the average value raised to the fourth power. The vibration steepness coefficient is obtained by dividing the fourth power of the average value by the fourth power of the standard deviation. (3) Extraction of cumulative temperature gradient features: Obtain continuous time series temperature detection values; Calculate the temperature difference between two adjacent time points and take the absolute value; The cumulative temperature gradient is obtained by summing the absolute values ​​of the temperature differences at all consecutive time points.

[0012] Furthermore, the formula for calculating the device health threshold is: H th =H initial ×(1-K year × Y / T year ; where H initial K represents the initial threshold of the device. year Y represents the annual adjustment factor, and T represents the service life. year Indicates the correction period. This indicates the floor function.

[0013] Furthermore, the formula for calculating the temperature threshold is: T th =T base +K load ×(LL base )×ΔT load Among them, T base Indicates the reference temperature, K. load L represents the load influence coefficient. base ΔT represents the baseline load rate, L represents the current load rate, and ΔT represents the baseline load rate. load This indicates the load temperature correction amount.

[0014] Furthermore, the formula for calculating the device's status score is: Where α represents the equipment insulation coefficient, β represents the sensitivity factor, and K ref ΔI represents the kurtosis reference value, γ represents the heat dissipation coefficient, λ represents the electrothermal coupling coefficient, ΔI represents the current change within the preset time window, and ΔT represents the temperature change within the preset time window.

[0015] Furthermore, the risk propagation simulation based on the dynamic causal graph includes: Initialize the dynamic cause-effect graph by taking the power distribution equipment whose status score exceeds the first threshold as the risk source node; Construct a propagation probability matrix P, and P ij =σ(W i,j ×Score i ); where P i,j W represents the propagation probability from node i to node j in the propagation probability matrix. i,j Score represents the edge weight between node i and node j. i σ represents the state score of the device at the i-th node in the dynamic causal graph, and σ represents the sigmoid function. The risk probability P of each neighboring node j is calculated using a Bayesian network model. risk,j ; Based on the propagation probability P ij and risk probability P risk,j Performing a Monte Carlo simulation yields multiple random propagation paths; Calculate the combined probability P of each random propagation path based on its occurrence frequency. path :P path =Pfreq ×∏P ij , where P freq P represents the frequency of occurrence of random propagation paths. freq =Number of path occurrences / Total number of simulations, ∏ (i->j)∈随机传播路径 P ij It represents the product of all propagation probabilities along a random propagation path; The combined probabilities of random propagation paths are sorted, and the top M random propagation paths are output to obtain the risk propagation paths.

[0016] Furthermore, the propagation probability P ij and risk probability P risk,j Performing a Monte Carlo simulation includes: A1. Starting from the risk source node, generate random numbers r∈[0,1]; A2. If r < P i,j The risk is then determined to have propagated from node i to node j, and the following action is taken: When P risk,j When the activation threshold is reached, node j is added to the current propagation path, and steps A1-A2 are recursively executed with node j as the new propagation starting point. A3, when P risk,j If the value is less than the activation threshold, a single simulation is completed, resulting in a single random propagation path. A4, repeat A1-A3, until the number of simulations is reached, resulting in multiple random propagation paths.

[0017] Furthermore, the triggering of tiered early warnings based on risk propagation paths includes: The risk is classified into levels based on the comprehensive probability of the risk transmission path, and the basis for the classification is as follows: Where Level represents the risk level and length represents the number of nodes in the risk propagation path; Based on the classification results, a warning signal corresponding to the level is sent.

[0018] Secondly, a smart assessment device for safety early warning of power distribution equipment is provided, comprising: a communication unit and a processing unit; The communication unit is used to acquire the operating status data of the power distribution equipment and send the graded early warning information generated by the processing unit to the terminal or system; The processing unit is used to create a digital twin for each power distribution device and construct a dynamic causal graph. Based on the operating status data, it performs status assessment through the digital twin and generates a status score. When the status score exceeds a first threshold, it performs risk propagation deduction based on the dynamic causal graph to output the risk propagation path and triggers a graded early warning according to the risk propagation path.

[0019] Thirdly, this application provides a smart assessment device for safety early warning of power distribution equipment, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is used to execute the instructions to implement the method described in the first aspect and any possible implementation thereof. This smart assessment device for safety early warning of power distribution equipment can be an electronic device or a chip within an electronic device.

[0020] Fourthly, this application provides a safety early warning intelligent assessment system for power distribution equipment, comprising: a data acquisition module, a risk assessment module, and a risk early warning module; wherein, The data acquisition module is used to collect real-time operating status data of the power distribution equipment and transmit it to the risk assessment module; The risk assessment module is used to create a digital twin for each power distribution device and construct a dynamic cause-effect graph based on the digital twin; each digital twin performs a status assessment based on the operating status data and generates a status score; when the status score exceeds a first threshold, risk propagation deduction is performed based on the dynamic cause-effect graph and the risk propagation path is output. The risk warning module is used to trigger and issue graded warning information based on the risk propagation path.

[0021] Fifthly, this application provides a computer-readable storage medium storing instructions that, when executed on a power distribution equipment safety early warning intelligent assessment device, cause the power distribution equipment safety early warning intelligent assessment device to perform the method described in the first aspect and any possible implementation thereof.

[0022] Sixthly, this application provides a computer program product containing instructions that, when the computer program product is run on a power distribution equipment safety early warning intelligent assessment device, causes the power distribution equipment safety early warning intelligent assessment device to perform the methods described in the first aspect and any possible implementation of the first aspect.

[0023] This application provides an intelligent assessment method and system for safety early warning of power distribution equipment, which can significantly improve the accuracy and effectiveness of safety early warning through multi-dimensional optimization design. In data processing, it extracts the proportion of abnormal pulse energy through wavelet packet decomposition, calculates the vibration steepness coefficient based on time-domain sampling, and accumulates the temperature gradient by summing the temperature difference, providing rich feature data for assessment. Regarding threshold setting, the equipment health threshold is dynamically adjusted according to the service life, and the temperature threshold is corrected in real time based on the current load rate, making the assessment standard more closely reflect the actual operating conditions of the equipment. The status score, by integrating multi-dimensional signal features and key parameters, provides a reliable basis for risk judgment.

[0024] The construction of a dynamic cause-effect graph, combined with historical fault records, determines the propagation direction. Edge weights are initialized based on electrical topology and physical distance, and dynamically updated according to equipment status scores, making the logical relationship of risk propagation more realistic. Risk propagation simulation enhances the scientific rigor of risk spread trend prediction by constructing a propagation probability matrix, calculating risk probabilities using Bayesian networks, generating random paths using Monte Carlo simulation, and selecting critical paths. Tiered early warning systems classify risk propagation paths based on their comprehensive probability and the number of nodes, enabling differentiated responses and facilitating rational resource allocation. This allows for early detection, early prediction, and effective prevention of safety risks in power distribution equipment, reducing the probability of fault occurrence and the scope of impact.

[0025] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0027] Figure 1 The system architecture diagram of the intelligent assessment system for safety early warning of power distribution equipment provided in the embodiments of this application is shown. Figure 2 A flowchart illustrating the intelligent assessment method for safety early warning of power distribution equipment provided in this application embodiment; Figure 3 A flowchart illustrating another intelligent assessment method for safety early warning of power distribution equipment provided in an embodiment of this application; Figure 4 A flowchart illustrating another intelligent assessment method for safety early warning of power distribution equipment provided in an embodiment of this application; Figure 5This is a schematic diagram of the structure of the intelligent assessment device for safety early warning of power distribution equipment provided in the embodiments of this application; Figure 6 This is a schematic diagram of the hardware structure of the intelligent assessment device for safety early warning of power distribution equipment provided in the embodiments of this application. Detailed Implementation

[0028] In the description of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B. The "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The terms "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0029] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0030] The intelligent safety early warning assessment method for power distribution equipment provided in this application embodiment can be applied to, for example... Figure 1 In the intelligent safety early warning assessment system for the power distribution equipment shown, such as Figure 1 As shown, the communication system includes: a data acquisition module, a risk assessment module, and a risk early warning module; wherein, The data acquisition module is used to collect real-time operating status data of power distribution equipment and transmit it to the risk assessment module; The risk assessment module is used to create a digital twin for each power distribution device and construct a dynamic cause-effect graph based on the digital twin; each digital twin performs a status assessment based on the operating status data and generates a status score; when the status score exceeds the first threshold, risk propagation simulation is performed based on the dynamic cause-effect graph and the risk propagation path is output. The risk warning module is used to trigger and issue tiered warning information based on the risk propagation path.

[0031] To address the technical problems of insufficient comprehensive assessment and inadequate risk propagation prediction in existing power distribution equipment safety early warning systems, this application provides an intelligent assessment method for power distribution equipment safety early warning, which includes: Create a digital twin for each power distribution device and build a dynamic cause-effect graph based on the digital twin; Each digital twin performs a status assessment based on operational status data and generates a status score; When the status score exceeds the first threshold, risk propagation simulation is performed based on the dynamic cause-effect graph, and the risk propagation path is output. Trial-level early warnings are triggered based on the risk transmission path.

[0032] Based on this, the method combines digital twins with dynamic causal graphs to achieve dynamic perception of the status of power distribution equipment and early risk assessment, providing systematic technical support for safety early warning.

[0033] like Figure 2 As shown in the embodiments of this application, the intelligent assessment method for safety early warning of power distribution equipment includes: S1. Create a digital twin for each power distribution device and build a dynamic causal graph based on the digital twin.

[0034] Among them, digital twins refer to the digital mapping of the physical entities of power distribution equipment, which can reflect the attributes, status and behavior of the equipment; dynamic cause-effect graphs are used to characterize the causal relationships between various power distribution equipment, and these relationships can be dynamically adjusted as the equipment status changes.

[0035] In some implementations, digital twins can be constructed by collecting basic parameters of devices (such as model and specifications) and combining them with real-time sensor data; in dynamic causal graphs, each digital twin can be used as a node, and the electrical connection, physical location and other relationships between devices can be used as edges to construct the relationship between nodes.

[0036] It should be noted that digital twins can be constructed using various technologies such as 3D modeling and data integration, and the causal relationships in dynamic causal graphs can also be determined based on multi-dimensional information such as historical data and expert experience, and are not limited to a specific construction method.

[0037] For example, a digital twin is created for a switch cabinet in a certain area, and its rated current, installation location and other information are recorded. At the same time, the digital twins of the switch cabinet and related equipment such as transformers and cables are used as nodes, and the edge connection relationship of a dynamic causal graph is constructed based on the power supply line connection between them.

[0038] S2. Each digital twin performs a status assessment based on the operational status data and generates a status score.

[0039] Among them, the status score is used to represent the current operating health status of power distribution equipment, which is a quantitative representation of whether there are any abnormalities or risks in the equipment.

[0040] In some implementations, status assessment can be achieved by comparing operating status data, such as voltage, current, and temperature, with preset standard ranges, or by using preset assessment rules, such as weighted summation, to calculate multi-dimensional data and obtain a status score.

[0041] It should be noted that the type of operational status data can be flexibly selected according to the characteristics of the equipment, and the rules for status assessment can also be adjusted according to the actual application scenario to adapt to the assessment needs of different equipment.

[0042] For example, a digital twin of a transformer acquires its oil temperature and winding temperature data during operation, quantifies the degree of deviation of the two data points from their corresponding standard values, and then adds them together according to a certain weight to obtain the transformer's status score.

[0043] S3. When the status score exceeds the first threshold, perform risk propagation deduction based on the dynamic cause-effect graph and output the risk propagation path.

[0044] Among them, risk propagation simulation is used to simulate the process of risk spreading from the initial device to other related devices in order to identify the devices that may be affected and the order of propagation.

[0045] In some implementations, a first threshold can be set as the critical value for a device to enter a risk warning state. When the state score exceeds this threshold, the digital twin of the corresponding device is used as the starting point. Based on the causal relationship and edge weight between nodes in the dynamic causal graph, path search, probability calculation and other methods are used to perform risk propagation simulation and screen out possible risk propagation paths.

[0046] It should be noted that the first threshold can be adjusted according to factors such as equipment type and operating environment, and the risk propagation simulation algorithm can also adopt various methods such as breadth-first search and depth-first search.

[0047] For example, when the status score of a cable exceeds a first threshold, starting from the digital twin of the cable, based on its association with the distribution box and load switch in the dynamic causal graph, the path that the risk may first propagate to the distribution box and then from the distribution box to the load switch is deduced.

[0048] S4. Trigger tiered early warnings based on the risk propagation path.

[0049] In some implementation methods, early warnings can be divided into different levels, such as general warnings, important warnings, and emergency warnings, based on factors such as the number of devices involved in the risk propagation path, the importance of the devices, and the probability of propagation. The corresponding level of warning can be triggered through SMS, system pop-ups, and audible and visual alarms.

[0050] It should be noted that the classification criteria for tiered early warning can be set according to actual management needs, and the triggering method for early warning can also be selected in combination with the application scenario to ensure that relevant personnel can obtain early warning information in a timely manner.

[0051] For example, if the risk transmission path involves only a single secondary device, a general warning is triggered, and a prompt message is sent to the on-duty personnel through the system; if the path involves multiple critical devices, an emergency warning is triggered, and in addition to the system prompt, an on-site audible and visual alarm is activated simultaneously.

[0052] Based on the above technical solutions, the intelligent assessment method for safety early warning of power distribution equipment provided in this application realizes dynamic mapping of equipment status through digital twins, providing a reliable basis for assessment; dynamic cause-effect graphs clearly present the correlation between equipment, laying the foundation for risk propagation analysis; combined with status scoring and threshold judgment, timely identification of risks is realized; through risk propagation simulation and graded early warning, relevant personnel can grasp the risk spread trend in advance and take targeted measures, thereby improving the effectiveness of power distribution equipment safety management.

[0053] In one possible implementation of the embodiments of this application, combined with Figure 2 ,like Figure 3 As shown, the above S1 can be implemented through the following S101, S102 and S103, which are explained in detail below: S101. Create a digital twin for each power distribution device, including collecting basic parameters and operating status data of the device and storing them in the digital twin.

[0054] Among them, the basic parameters are data that characterize the inherent attributes of the power distribution equipment, including equipment model, equipment service life, rated voltage and rated current; the operating status data are data that reflect the real-time operating status of the equipment, including current, voltage, temperature, vibration and discharge pulse.

[0055] In some implementations, basic parameters can be collected from documents such as equipment factory data and installation records, or manually entered into the system; operating status data can be collected by deploying current sensors, voltage sensors, temperature sensors, vibration sensors, etc. in key parts of the equipment, and data transmission channels can adopt wired communication (such as Ethernet, RS485 bus) or wireless communication (such as LoRa, NB-IoT).

[0056] It should be noted that the basic parameters are the foundation for building a digital twin, and their accuracy must be ensured to reflect the inherent characteristics of the equipment. The collection frequency of operating status data can be set according to the equipment type and importance, and a higher frequency can be used for critical equipment to ensure real-time performance.

[0057] For example, when creating a digital twin for a 10kV switchgear, its model "KYN28-12", service life "5 years", rated voltage "10kV", and rated current "630A" are collected as basic parameters; temperature data is collected every 5 minutes by a temperature sensor deployed in the cabinet, and current data is collected in real time by a current transformer, and transmitted to the database of the digital twin via RS485 bus for storage.

[0058] S102. Obtain historical equipment fault records, generate equipment alarm sequences, and determine the direction of fault propagation.

[0059] The historical equipment fault records include the fault occurrence time, affected equipment, alarm information, etc.; the equipment alarm sequence is a sequence of affected equipment in the same fault event ordered by alarm timestamp; the fault propagation direction is the direction from the earliest alarm device in the sequence to the subsequent alarm devices.

[0060] In some implementations, historical equipment fault records can be extracted from databases such as power distribution automation systems and equipment operation and maintenance management systems; timestamps can be accurate to the second to ensure sorting accuracy.

[0061] It should be noted that the time range of historical fault records can be set according to the equipment's operating years, and generally needs to cover at least 3-5 years of records to ensure that the amount of data is sufficient to support the analysis of fault propagation patterns; for new equipment without historical fault records, the initial propagation direction can be determined by referring to the fault data of similar equipment.

[0062] For example, a transformer fault in a certain area causes related cables and distribution boxes to alarm successively. The timestamps of the fault records are extracted: the transformer alarm time is "2023-06-10 08:30:00", the cable alarm time is "2023-06-10 08:32:15", and the distribution box alarm time is "2023-06-10 08:35:30". The generated equipment alarm sequence is [transformer, cable, distribution box], and the fault propagation direction is transformer → cable → distribution box.

[0063] S103. Construct a dynamic causal graph based on the digital twin and the direction of fault propagation, including node mapping, initial edge connection, edge weight setting and dynamic updating.

[0064] Among them, a node is a mapping of the digital twin of each power distribution equipment, and its attributes include equipment identifier (such as unique code), equipment type and equipment status score; the initial edge connection is the node association established based on the electrical topology between the equipment (such as power supply line connection, bus tie relationship), and the edge direction is the fault propagation direction determined by S102; the edge weight is used to characterize the association strength of risk propagation between nodes, including the initial weight and the dynamically updated weight.

[0065] In some implementations, the electrical topology can be obtained from the power grid GIS system or the distribution network topology map; the edge initialization weights can be calculated based on the reciprocal of the physical distance between devices, which can be obtained through GPS positioning or on-site measurement; the trigger condition for dynamically updating edge weights can be set to a preset time interval (e.g., 1 hour) or the rate of change of the device status score being greater than the rate of change threshold (e.g., 10% / minute), and the edge weight update formula is W=W init ×(1+θ×Score source ), where θ is a preset weight correction coefficient, which can be determined based on historical data fitting, and generally ranges from 0.01 to 0.1. Score source Assign a status score to the upstream node.

[0066] It should be noted that the state score in the node attributes needs to be updated in real time to support the dynamic adjustment of edge weights; when the electrical topology changes, such as the addition or deletion of equipment or the reconnection of lines, the initial edge connection needs to be re-established to ensure the accuracy of the cause-effect graph.

[0067] For example, digital twins of transformers, cables, and distribution boxes are mapped as nodes with node attributes of [ID:B01, Type: Transformer, Score: 1.2], [ID:D03, Type: Cable, Score: 0.8], and [ID:P05, Type: Distribution Box, Score: 0.6], respectively. Based on the electrical topology, the transformer is directly connected to the cable, and the cable is directly connected to the distribution box. Edge connections are established based on the fault propagation direction: B01→D03, D03→P05. The measured physical distance between B01 and D03 is 50 meters, and the physical distance between D03 and P05 is 30 meters. The edge weight W is initialized. init The weights are 1 / 50 = 0.02 and 1 / 30 ≈ 0.033 respectively; if θ = 0.1, then the current edge weights are 0.02 × (1 + 0.1 × 1.2) = 0.02 × 1.12 = 0.0224 and 0.033 × (1 + 0.1 × 0.8) = 0.033 × 1.08 ≈ 0.0356 respectively.

[0068] Based on the above technical solutions, by collecting data, analyzing fault patterns, and dynamically constructing cause-effect graphs, digital twins can accurately map equipment status, and dynamic cause-effect graphs can truly reflect the correlation of risk propagation between equipment, laying a reliable foundation for subsequent status assessment and risk simulation.

[0069] In one possible implementation of the embodiments of this application, combined with Figure 2 ,like Figure 3 As shown, the above S2 can be implemented through the following S201, S202 and S203, which are explained in detail below: S201. Preprocess the operating status data to obtain multidimensional signal characteristics.

[0070] Among them, the multidimensional signal characteristics include the proportion of abnormal pulse energy E abn Vibration steepness coefficient K ur and temperature gradient accumulation T acc The percentage of abnormal pulse energy is a characteristic quantity used to characterize the degree of abnormal discharge of the equipment. The vibration steepness coefficient is used to reflect the impact characteristics of the equipment vibration signal. The cumulative temperature gradient is used to reflect the cumulative trend of temperature change of the equipment.

[0071] In some implementations, preprocessing may include the following steps: (1) Extraction of abnormal pulse energy ratio features: For current and voltage signals, wavelet packet decomposition (WPD) is performed to decompose the signal into sub-signals in a preset frequency band; Extract the sum of energy within the preset frequency band after wavelet packet decomposition, and denote it as the first energy sum. Calculate the total energy across all frequency bands and denot it as the second energy sum. According to the formula: Abnormal pulse energy percentage = Sum of first energy / Sum of second energy, the abnormal pulse energy percentage E can be calculated. abn .

[0072] (2) Extraction of vibration steepness coefficient features: For the vibration signal, acquire its N time-domain sampling point data x i (i=1,2,…,N), calculate the mean μ and standard deviation σ of the signal, and calculate the kurtosis coefficient K according to the following formula. ur : ; (3) Extraction of cumulative temperature gradient features: For the temperature signal, acquire the continuous time series temperature monitoring value T. t (t is a time point), calculate the absolute value of the temperature difference between adjacent time points, and calculate the gradient accumulation using the following formula. T acc : ; The number of samples N can be determined according to the frequency characteristics of the equipment vibration signal, and is generally 1024 or 2048 to cover 2-3 vibration cycles. In the calculation of the cumulative temperature gradient, the interval of the continuous time series can be set to 5-15 minutes, which can be adjusted according to the heat dissipation characteristics of the equipment.

[0073] It should be noted that the preprocessing process requires noise reduction of the data, such as by using mean filtering or wavelet thresholding, to avoid noise interfering with the accuracy of feature extraction. Parameters such as the preset frequency band and the number of samples can be determined through experimental testing according to the equipment type (e.g., transformer, switchgear). In this embodiment, the preset frequency band is set as the 6th to 12th frequency bands after decomposition, corresponding to the 5kHz-10MHz frequency band, used to capture high-frequency pulse signals generated by power distribution equipment during operation due to abnormal conditions such as insulation defects and partial discharge. This frequency band covers the frequency characteristics of pulse signals under most typical faults of power distribution equipment, and can effectively distinguish between normal operation signals and fault pulse signals, providing a reliable frequency band basis for accurately extracting the proportion of abnormal pulse energy.

[0074] S202. Adjust the equipment health threshold H according to the equipment's service life. th And adjust the temperature threshold T based on the current load rate. th .

[0075] Among them, the equipment health threshold is a critical value for measuring the overall health status of the equipment, which decreases dynamically with the increase of service life; the temperature threshold is a benchmark for judging whether the equipment temperature is abnormal, which is adjusted in real time with the change of load rate.

[0076] In some implementations, the health threshold is calculated using the formula H. th =H initial ×(1-K year × Y / T year H initial The initial threshold for the equipment can be set according to the factory standard, such as 0.8 for transformers, K year This is the annual correction factor, obtained by fitting lifespan degradation data of similar equipment, and is generally taken as 0.02-0.05, T. year The correction period is typically set to one year. The formula for calculating the temperature threshold is T. th =T base +K load ×(LL base )×ΔT load T base Reference temperature, representing the baseline temperature reference value of the equipment at a reference load rate (such as normal operating load level), K. load The load impact factor is determined based on the equipment's heat dissipation curve; for example, 0.5 is used for switchgear. base The baseline load rate is set at 50%, ΔT load This is the load temperature correction amount, representing the temperature rise under full load, such as 40℃ for a transformer.

[0077] It should be noted that the reference temperature is usually taken as the ambient temperature of the equipment plus a certain basic temperature rise value (such as ambient temperature + 20℃), which is determined based on the heat dissipation characteristics of the equipment under normal load and safe operation experience. Alternatively, the average temperature of the equipment under long-term normal operation can be collected as the reference temperature to ensure that it matches the actual operating conditions of the equipment. The health threshold and temperature threshold of the equipment need to be calibrated regularly, and the parameters should be corrected in combination with the actual operating conditions of the equipment to ensure matching with equipment aging and load changes. For newly commissioned equipment, the average threshold of the same batch of equipment can be used as the initial value.

[0078] For example, a switch cabinet that has been in use for 3 years, H initial =0.8, K year =0.03, T year =1 year, calculated as H th =0.8×(1-0.03×3)=0.8×0.91=0.728; Current load rate L=70%, T base =40℃, K load =0.5, L base =50%, ΔT load =20℃, then T th =40+0.5×(0.7-0.5)×20=40+2=42℃.

[0079] S203. Calculate the equipment status score based on multidimensional signal characteristics, equipment health threshold, and temperature threshold.

[0080] Among them, the status score is a quantitative indicator that comprehensively reflects the operating status of the equipment. By integrating information from multiple dimensions such as discharge, vibration, and temperature, it achieves a comprehensive assessment of the health of the equipment.

[0081] In some implementations, the state score is calculated using the following formula: ; Where α is the equipment insulation coefficient, β is the sensitivity factor, used to reflect the degree of vibration impact on the equipment, and is taken as 0.01 for mechanical equipment, K ref γ is the kurtosis baseline value, i.e. the vibration kurtosis under normal conditions, which is obtained through historical normal data statistics and is generally taken as 3. γ is the heat dissipation coefficient, which is set according to the heat dissipation method of the equipment. For example, when the heat dissipation method is forced air cooling, γ is taken as 0.2. λ is the electrothermal coupling coefficient, which is used to reflect the correlation between current and temperature. The default value is 0.1. ΔI and ΔT are the changes in current and temperature within the preset time window.

[0082] It should be noted that α is used to quantify the response characteristics of equipment insulation performance to abnormal energy. It is typically determined based on the type of insulation material (e.g., oil-paper insulation, silicone insulation), combined with insulation aging test data. The relationship between insulation failure probability and energy impact is fitted using linear regression or a neural network, and finally calibrated by experts. For example, for oil-paper insulated distribution transformers, α can be taken as 1.2 after experimental fitting and calibration; for silicone insulated switchgear, α is often taken as 1.5. β can be categorized according to equipment type (0.01 for mechanical equipment such as circuit breakers and disconnectors; and as low as 0.005 for non-mechanical equipment such as transformers). Vibration table tests are used to simulate the equipment failure rate under different vibration intensities. The correlation curve between vibration kurtosis deviation and failure probability is statistically analyzed to obtain a baseline β value for different equipment types, which is then fine-tuned based on the actual proportion of vibration failures in on-site maintenance. K ref The vibration kurtosis level, representing the normal operating level of the equipment, can be calculated by collecting historical normal operating data from the initial 3-6 months of equipment commissioning, and then taking the statistical mean as K. ref Under normal conditions, the kurtosis of the vibration signal is concentrated around 3, so the default value is 3.

[0083] γ is used to reflect the adaptability of the equipment's heat dissipation method to temperature changes. Initial values ​​can be set according to the heat dissipation method, such as 0.2 for forced air cooling, 0.15 for natural air cooling, and 0.1 for oil-immersed self-cooling. These initial values ​​can then be corrected through temperature rise tests and finite element thermal simulations to simulate temperature distribution under abnormal conditions such as blocked heat dissipation channels or fan failures. For example, if there is severe dust accumulation in the air ducts of a forced air-cooled device, γ can be increased to 0.25.

[0084] λ is used to measure the correlation between current change and temperature change. It is usually based on the electromagnetic-thermal coupling simulation model of the equipment. Different current impact scenarios are input, and the distribution of the ratio of temperature change rate to current change rate is statistically analyzed. The mean of the 95% confidence interval is taken as the default value of 0.1. If the equipment has obvious electrothermal synergistic faults, such as electrothermal feedback caused by poor contact of the contacts, the sudden change characteristics of current and temperature before and after the fault can be extracted by analyzing the fault waveform data, and then λ is increased to 0.12.

[0085] In addition, the coefficients in the formula need to be optimized through training with a large amount of historical data. The least squares method can be used to select the coefficient that best fits the score to the actual fault state of the equipment as the final value. The time windows of ΔI and ΔT can be adjusted according to the response speed of the equipment, such as 30 minutes for cables and 1 hour for transformers.

[0086] In the formula for calculating the equipment condition score, the first term The proportion of abnormal pulse energy increases non-linearly—the increase is slow during minor anomalies and rapid during severe anomalies, consistent with the cumulative characteristics of insulation degradation: "slow at the beginning and accelerated later." This avoids excessive influence of minor fluctuations on the score while accurately detecting severe insulation problems; the second item This item is used to reflect the mechanical condition of the equipment. When the vibration kurtosis exceeds the vibration reference value, this item increases linearly, intuitively reflecting the aggravating trend of mechanical loosening, wear, and other faults, which is consistent with the direct correlation between mechanical faults and vibration characteristics; the third item By standardizing the ratio of accumulated temperature to a threshold and adjusting the weights based on the heat dissipation coefficient, the system reflects both the cumulative effect of temperature anomalies and the differences in equipment heat dissipation capabilities, making the scoring more closely reflect the heat dissipation characteristics of different devices. (Fourth item) It is used to reflect the correlation between current and temperature. Under normal circumstances, the change in current should correspond to a reasonable change in temperature. When the ratio is abnormal, it often indicates problems such as poor contact or abnormal impedance. It can capture the early characteristics of abnormal electrothermal conversion.

[0087] Therefore, this calculation formula achieves a multi-dimensional and systematic assessment of the operating status of power distribution equipment by comprehensively quantifying the equipment's insulation, mechanical properties, heat dissipation, and electrothermal coupling. Each indicator functions independently for different aspects of the equipment's characteristics while also complementing each other to form a complete evaluation system that comprehensively reflects the equipment's potential failure risks.

[0088] Based on the above technical solution, through multi-dimensional feature extraction, dynamic threshold adjustment and comprehensive scoring calculation, the status assessment can fully reflect the actual operating status of the equipment. It takes into account both the inherent characteristics and aging trend of the equipment and the real-time operating conditions, providing a scientific and accurate quantitative basis for subsequent risk assessment.

[0089] In one possible implementation of the embodiments of this application, combined with Figure 2 ,like Figure 3 As shown, the above S3 can be implemented through the following S301, S302 and S303, which are explained in detail below: S301. Initialize the dynamic cause-effect graph by taking the power distribution equipment whose status score exceeds the first threshold as the risk source node.

[0090] Among them, the risk source node refers to the power distribution equipment whose current operating status has entered the risk warning range, and its status score is the key indicator for triggering risk propagation inference; the initialization of the dynamic cause-effect graph refers to updating the current status score and edge weight of all nodes in the graph to ensure that the information in the graph is consistent with the real-time operating status of the equipment.

[0091] In some implementations, the first threshold can be set according to the equipment type and importance: for critical equipment (such as pre-setting the main transformer as critical equipment), the first threshold can be set to 60% of the status score range (such as 1.2 in the [0-2] score range), and for general equipment (such as pre-setting branch cables as general equipment), it can be set to 70% (such as 1.4). The specific value is determined by fitting the "score-fault occurrence probability" curve in the historical fault data to ensure that the deduction can be effectively triggered before the fault occurs; the initialization of the dynamic cause-effect graph can be automatically executed when the status score exceeds the first threshold, or it can be triggered in combination with manual confirmation.

[0092] It should be noted that if multiple devices have status scores that exceed the first threshold at the same time, all eligible devices should be treated as risk source nodes, and their corresponding dynamic cause-effect graph branches should be initialized to avoid missing potential risk sources.

[0093] For example, in a certain regional power distribution system, if the status scores of both the transformer (Score=1.5) and the switchgear (Score=1.3) exceed the first threshold (1.2), then both will be set as risk source nodes, their digital twin status scores will be updated in the dynamic causal graph, and the edge weights connected to them will be updated synchronously to complete the initialization of the causal graph.

[0094] S302. Construct the propagation probability matrix and calculate the risk probability of adjacent nodes using a Bayesian network model.

[0095] Wherein, the propagation probability matrix P is a matrix characterizing the probability of risk propagation between nodes in a dynamic causal graph, P i,j P represents the probability that the risk of node i propagates to node j; the risk probability P of adjacent nodes. risk,j This refers to the probability that node j will enter a risky state after being affected by upstream node i, which is calculated by combining a Bayesian network model with historical fault data.

[0096] In some implementations, the elements of the propagation probability matrix are calculated using the formula P. i,j =σ(W i,j ×Score i ), where σ is the sigmoid function: σ(x) = 1 / (1+e -x (Map the calculation results to the [0,1] interval), W i,j Score represents the edge weight between node i and node j. iRepresents the state score of the device for the \(i\)-th node in the dynamic causality diagram; the structure of the Bayesian network model is constructed based on the electrical associations between devices and the historical fault propagation rules. The conditional probability table of the nodes is determined by statistically calculating the conditional probability of "upstream node risk occurrence → downstream node risk occurrence" in historical data, and the model needs to be updated regularly with new fault data to ensure accuracy.

[0097] It should be noted that the dimension of the propagation probability matrix is consistent with the number of nodes in the dynamic causality diagram (assuming there are \(N\) nodes, the matrix is \(N\times N\) order). For nodes without direct association (i.e., edge weight \(W\) i,j = 0), its propagation probability \(P\) i,j= 0.

[0098] Exemplarily, based on the initialized dynamic causality diagram, including node A - transformer, B - switchgear, C - cable, given \(W\) A,B = 0.05, \(W\) B,C = 0.03, \(Score\) A = 1.5, \(Score\) B = 1.3, then the propagation probability \(P\) A,B = \(\sigma(0.05\times1.5)=\sigma(0.075)\approx0.519\), \(P\) B,C = \(\sigma(0.03\times1.3)=\sigma(0.039)\approx0.509\), construct the propagation probability matrix ; Assume that through Bayesian network calculation, the risk probability \(P\) risk,B of adjacent node A of B is 0.6, and the risk probability \(P\) risk,C of adjacent node B of C is 0.4.

[0099] S303. Perform Monte Carlo simulation based on the propagation probability and risk probability, calculate the comprehensive probability of the random propagation path and output the risk propagation path.

[0100] Among them, Monte Carlo simulation is a method of generating multiple possible risk propagation paths through multiple random trials; the comprehensive probability \(P\) path is an index to measure the possibility of the random propagation path appearing, which is jointly determined by the path occurrence frequency \(P\) freq and the product \(\prod P\) ij of all propagation probabilities on the path; the output risk propagation path is the top \(M\) random paths in terms of comprehensive probability ranking, and \(M\) is the preset number of paths, usually taking 3 - 5.

[0101] In some implementation manners, the steps of Monte Carlo simulation are: Generate a random number \(r\in[0,1]\) starting from the risk source node. If \(r < P_{i,j}\), it is determined that the risk propagates to node \(j\). When \(P_{risk,j}\geq\) the activation threshold, recursively simulate with \(j\) as the new starting point until \(P\) risk,jA simulation ends when the value is less than 0.3. The number of simulations is typically set to 1000-10000; more simulations result in more stable results. The activation threshold is usually set to 0.3, determined by the critical value of "risk probability - actual failure occurrence" in historical data. It should be noted that the calculation of the overall probability should take into account the product of the frequency of path occurrence and the propagation probability, and avoid relying solely on frequency or a single probability, which may lead to biased results. If the overall probabilities of multiple paths differ slightly (e.g., the difference is <0.05), the value of M can be appropriately increased to retain more potential paths.

[0102] For example, if 1000 Monte Carlo simulations are performed on a system containing risk source A, with an activation threshold of 0.3: in 600 simulations, the risk propagates from A to B, then the risk probability P of node B is... risk,B =0.6≥0.3, in 400 simulations, the risk propagates from B to C, then the risk probability P of node C is... risk,C =0.4≥0.3), forming path 1: A→B→C; in 300 simulations, the risk only propagated from A to B, forming path 2: A→B; and in 100 simulations, the risk did not propagate. Therefore, we calculate: Path 1: P freq(A→B→C) =600 / 1000=0.6, ∏P i,j =0.519×0.509≈0.264, the overall probability P path =0.6×0.264≈0.158; Path 2: P freq(A→B) =0.3, ∏P i,j =0.519, overall probability P path ≈0.3×0.519≈0.156. If M=2, output the top 2 paths: A→B→C (0.158), A→B (0.156).

[0103] Based on the above technical solution, through risk source initialization, probability matrix construction, Monte Carlo simulation and comprehensive probability ranking, risk propagation simulation can combine the real-time status of equipment and historical patterns to accurately capture potential risk spread paths. It takes into account the randomness of propagation and ensures the reliability of the results through quantitative probability, providing a clear basis for risk diffusion for subsequent graded early warning.

[0104] In one possible implementation of the embodiments of this application, combined with Figure 2 ,like Figure 3 As shown, the above S4 specifically includes the following S401 to S403: S401. Determine the overall probability and number of nodes of the risk propagation path.

[0105] Among them, the overall probability P pathIt is a quantitative indicator that measures the probability of a risk propagation path occurring. It is calculated by multiplying the frequency of path occurrence by the probability of all propagation along the path. The number of nodes (length) refers to the total number of digital twin nodes of power distribution equipment involved in the risk propagation path, reflecting the scope of risk spread.

[0106] In some implementations, the overall probability can be directly used as the result calculated in S303 without secondary calculation; the number of nodes is determined by traversing the nodes in each risk propagation path and counting them. When counting, duplicate nodes need to be excluded, that is, the case of circular propagation in the path needs to be excluded.

[0107] It should be noted that if multiple risk propagation paths partially overlap, such as sharing some nodes, the number of independent nodes for each path should be counted separately to avoid deviations in the scope assessment due to overlap.

[0108] For example, for the risk propagation path A→B→C, the overall probability is calculated to be 0.65, involving nodes A, B, and C, with a total of 3 nodes; for the path A→D, the overall probability is 0.42, involving nodes A and D, with a total of 2 nodes.

[0109] S402. Risk levels are classified based on overall probability and number of nodes.

[0110] Among them, the risk level is a classification standard based on the severity and scope of the risk, used to determine the urgency of the warning and the response level.

[0111] In some implementations, the classification of levels can be based on a multi-dimensional rule that combines comprehensive probability and the number of nodes: for example, when 0.7 ≤ P path When 0.5 ≤ P and length ≥ 5, it is defined as Level 1 (emergency warning); when 0.5 ≤ P path <0.7 and 3≤length<5, or 0.7≤P path When 0.3 ≤ P and length < 3, it is defined as Level 2 (Important Warning); when 0.3 ≤ P path <0.5 and length≥3, or 0.5≤P path When <0.7 and length<3, it is defined as Level 3 (general warning), or: The threshold values ​​for each interval can be determined by analyzing the correlation between the extent of losses caused by historical faults and the corresponding path parameters. For example, the path parameters corresponding to faults that cause large-scale power outages can be used as the critical values ​​for Level 1.

[0112] It should be noted that the classification rules need to be adjusted periodically based on actual operation and maintenance experience. For example, when the actual risk impact of a certain type of path exceeds expectations, its corresponding level can be appropriately increased.

[0113] For example, based on the above partitioning rules, the path A→B→C (P path =0.65, length=3) satisfies 0.5≤P path If the length is less than 0.7 and 3 ≤ length < 5, it is classified as Level 2 (Important Warning); Path A → D (P path =0.42, length=2) satisfies 0.3≤P path If the value is less than 0.5 and the length is less than 3, it is defined as Level 3 (general warning).

[0114] S403. Trigger and send a warning signal of the corresponding level based on the risk level.

[0115] Among them, early warning signals refer to information used to notify relevant personnel or systems to take prevention and control measures, including key information such as risk level, equipment involved, and transmission route.

[0116] In some implementation methods, different levels of early warning signals can adopt different sending methods and response mechanisms: Level 1 (emergency early warning) can be sent to the operation and maintenance manager and emergency team simultaneously through system pop-ups, SMS, and telephone voice notifications, and trigger on-site audible and visual alarms; Level 2 (important early warning) notifies the operation and maintenance team through system pop-ups and SMS, requiring a response within 2 hours; Level 3 (general early warning) is only displayed through system records and team dashboards, requiring verification within 24 hours.

[0117] For example, for a Level 2 warning path A→B→C, the system automatically generates the warning message "[Level 2 Warning] Risk path: A→B→C, overall probability: 0.65, involved equipment: transformer A, switch cabinet B, cable C, please have the maintenance team conduct on-site verification within 2 hours", and sends it to the team leader via SMS, while also displaying it in a pop-up window in the maintenance system; for a Level 3 warning path A→D, the information is only updated on the system dashboard, prompting "[Level 3 Warning] Risk path: A→D, overall probability: 0.42, involved equipment: transformer A, distribution box D, please complete the status check within 24 hours".

[0118] Based on the above technical solutions, by quantifying path parameters, classifying levels in multiple dimensions, and differentiating early warning responses, the graded early warning can not only accurately reflect the severity and scope of the risk, but also guide resources to be prioritized for high-risk scenarios, thereby improving the pertinence and efficiency of safety early warning and effectively reducing losses caused by power distribution equipment failures.

[0119] The above primarily describes the solutions of the embodiments of this application from the perspective of device implementation. It is understood that each device, such as a smart assessment device for safety early warning of power distribution equipment, includes at least one of the hardware structures and software modules corresponding to the execution of each function in order to achieve the above-mentioned functions. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] This application embodiment can divide the intelligent assessment device for safety early warning of power distribution equipment into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or software functional units. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0121] When using integrated units, Figure 5 A possible structural schematic diagram of the intelligent safety early warning assessment device for power distribution equipment (referred to as intelligent safety early warning assessment device 50 for power distribution equipment) involved in the above embodiments is shown. The intelligent safety early warning assessment device 50 for power distribution equipment includes a processing unit 501 and a communication unit 502, and may also include a storage unit 503. Figure 5 The structural diagram shown can be used to illustrate the structure of the intelligent assessment device for safety early warning of power distribution equipment involved in the above embodiments.

[0122] For example, communication unit 502 is used to acquire the operating status data of power distribution equipment and send the graded early warning information generated by the processing unit to relevant terminals or systems; The processing unit 501 is used to create a digital twin for each power distribution device and construct a dynamic cause-effect graph. Based on the operating status data, it performs status assessment through the digital twin and generates a status score. When the status score exceeds the first threshold, it performs risk propagation deduction based on the dynamic cause-effect graph to output the risk propagation path, and then triggers a graded early warning according to the risk propagation path.

[0123] In one possible implementation, the processing unit 501 is further configured to preprocess the operating status data to extract multidimensional signal features, adjust the equipment health threshold according to the equipment's service life, correct the temperature threshold in combination with the current load rate, and calculate the equipment's status score based on these parameters.

[0124] In one possible implementation, the communication unit 502 is further configured to receive historical equipment fault records from other systems, and the processing unit 501 is further configured to determine the fault propagation direction based on the historical fault records in order to optimize the construction of the dynamic cause-effect graph, and at the same time dynamically update the edge weights of the dynamic cause-effect graph according to a preset time interval or the rate of change of the equipment status score.

[0125] The processing unit 501 can be a processor or a controller, and the communication unit 502 can be a communication interface, transceiver, transceiver circuit, transceiver device, etc. The term "communication interface" is a general term and may include one or more interfaces. The storage unit 503 can be a memory. When the intelligent safety early warning assessment device 50 for power distribution equipment is a chip, the processing unit 501 can be a processor or a controller, and the communication unit 502 can be an input interface and / or an output interface, pins, or circuits, etc. The storage unit 503 can be a storage unit within the chip (e.g., a register, cache, etc.) or a storage unit located outside the chip (e.g., read-only memory (ROM), random access memory (RAM, etc.)).

[0126] The communication unit can also be called a transceiver unit. The antenna and control circuit with transceiver functions in the intelligent safety early warning assessment device 50 for power distribution equipment can be considered as the communication unit 502 of the intelligent safety early warning assessment device 50 for power distribution equipment, and the processor with processing functions can be considered as the processing unit 501 of the intelligent safety early warning assessment device 50 for power distribution equipment. Optionally, the device in the communication unit 502 used to implement the receiving function can be considered as a communication unit, which is used to execute the receiving steps in the embodiments of this application. The communication unit can be a receiver, a receiver circuit, etc. The device in the communication unit 502 used to implement the transmitting function can be considered as a transmitting unit, which is used to execute the transmitting steps in the embodiments of this application. The transmitting unit can be a transmitter, a transmitter, a transmitting circuit, etc.

[0127] Figure 5If the integrated units in the process are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. Storage media for storing computer software products include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0128] Figure 5 The units in the process can also be called modules; for example, a processing unit can be called a processing module.

[0129] This application embodiment also provides a hardware structure diagram of a smart assessment device for safety early warning of power distribution equipment (denoted as smart assessment device 60 for safety early warning of power distribution equipment), see [link to relevant documentation]. Figure 6 The intelligent safety early warning assessment device 60 for the power distribution equipment includes a processor 601, and optionally, a memory 602 connected to the processor 601.

[0130] In the first possible implementation, see Figure 6 The intelligent safety early warning assessment device 60 for power distribution equipment also includes a transceiver 603. The processor 601, memory 602, and transceiver 603 are connected via a bus. The transceiver 603 is used to communicate with other devices or communication networks. Optionally, the transceiver 603 may include a transmitter and a receiver. The device in the transceiver 603 that implements the receiving function can be considered as a receiver, which is used to perform the receiving steps in the embodiments of this application. The device in the transceiver 603 that implements the transmitting function can be considered as a transmitter, which is used to perform the transmitting steps in the embodiments of this application.

[0131] Based on the first possible implementation method Figure 6 The structural diagram shown can be used to illustrate the structure of the intelligent assessment device for safety early warning of power distribution equipment involved in the above embodiments.

[0132] in, Figure 6 The diagram can also illustrate the system chip in the intelligent safety early warning assessment device for power distribution equipment. In this case, the actions performed by the aforementioned intelligent safety early warning assessment device for power distribution equipment can be implemented by this system chip. The specific actions performed can be found above and will not be repeated here.

[0133] In implementation, each step of the method provided in this embodiment can be completed by integrated logic circuits in the processor or by instructions in software form. The steps of the method disclosed in the embodiments of this application can be directly manifested as being executed by a hardware processor, or being executed by a combination of hardware and software modules in the processor.

[0134] The processor in this application may include, but is not limited to, at least one of the following: a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a microcontroller unit (MCU), or an artificial intelligence processor, etc., which are various computing devices that run software. Each computing device may include one or more cores for executing software instructions to perform calculations or processing. The processor may be a separate semiconductor chip or integrated with other circuits into a single semiconductor chip. For example, it may be integrated with other circuits (such as encoding / decoding circuits, hardware acceleration circuits, or various bus and interface circuits) to form a SoC (System-on-a-Chip), or it may be integrated as a built-in processor within an ASIC. The ASIC with the integrated processor may be packaged separately or together with other circuits. In addition to the cores for executing software instructions to perform calculations or processing, the processor may further include necessary hardware accelerators, such as field-programmable gate arrays (FPGAs), PLDs (programmable logic devices), or logic circuits that implement dedicated logic operations.

[0135] The memory in the embodiments of this application may include at least one of the following types: read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions; random access memory (RAM) or other types of dynamic storage devices capable of storing information and instructions; or electrically erasable programmable-only memory (EEPROM). In some scenarios, the memory may also be a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures that can be accessed by a computer, but is not limited thereto.

[0136] This application also provides a computer-readable storage medium including instructions that, when run on a computer, cause the computer to perform any of the methods described above.

[0137] This application also provides a computer program product containing instructions that, when run on a computer, cause the computer to perform any of the methods described above.

[0138] This application also provides a chip including a processor and an interface circuit. The interface circuit is coupled to the processor. The processor is used to run computer programs or instructions to implement the above-described method. The interface circuit is used to communicate with other modules outside the chip.

[0139] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software programs, implementation can be, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device containing one or more servers, data centers, etc., that can be integrated with the medium. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0140] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0141] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of this application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and modifications of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and modifications.

Claims

1. A smart assessment method for safety early warning of power distribution equipment, characterized in that, include: Create a digital twin for each power distribution device and build a dynamic cause-effect graph based on the digital twin; Each digital twin performs a status assessment based on operational status data and generates a status score; When the state score exceeds the first threshold, risk propagation deduction is performed based on the dynamic causal graph, and the risk propagation path is output. Trial-level early warnings are triggered based on the risk transmission path.

2. The intelligent assessment method for safety early warning of power distribution equipment according to claim 1, characterized in that, Creating a digital twin for each power distribution device includes: Basic parameters of each power distribution device are collected and entered into the basic information database of the digital twin; the basic parameters include equipment model, equipment service life, equipment rated voltage and rated current; Sensors are deployed to collect operating status data of each power distribution device, and the operating status data is transmitted to the corresponding digital twin for storage through a data transmission channel; the operating status data includes current, voltage, temperature, vibration and discharge pulse.

3. The intelligent assessment method for safety early warning of power distribution equipment according to claim 2, characterized in that, The construction of a dynamic causal graph based on a digital twin includes: Obtain historical equipment fault records; for each fault event, sort the affected power distribution equipment by alarm timestamp and generate an equipment alarm sequence. The fault propagation direction is obtained by taking the device that first alarms in the device alarm sequence as the starting point and the devices that subsequently alarm as the propagation direction. The digital twins of each power distribution device are mapped to nodes, and the attributes of the nodes include device identifier, device type and device status score; The initial edge connection relationship between nodes is established based on the electrical topology between devices, and the edge direction is the fault propagation direction; The edge initialization weights are set based on the reciprocal of the physical distance between the devices; Based on a preset time interval, or when the rate of change of the device's status score exceeds a threshold, the edge weights are dynamically updated to obtain a dynamic causal graph.

4. The intelligent assessment method for safety early warning of power distribution equipment according to claim 1, characterized in that, The digital twin performs state assessment based on real-time data, including: The operating status data are preprocessed to obtain multidimensional signal features; the multidimensional signal features include the proportion of abnormal pulse energy E. abn Vibration steepness coefficient K ur and temperature gradient accumulation T acc ; Adjust the equipment health threshold H based on the equipment's service life Y. th ; Adjust the temperature threshold T based on the current load rate. th ; The device status score is calculated based on multidimensional signal characteristics, device health threshold, and temperature threshold.

5. The intelligent assessment method for safety early warning of power distribution equipment according to claim 4, characterized in that, The preprocessing of the operating status data includes: (1) Extraction of abnormal pulse energy ratio features: The current and voltage data are decomposed into sub-signals of multiple frequency bands using wavelet packets; The sum of energy within the preset frequency band after statistical decomposition is used to obtain the first energy sum. Calculate the total energy across all frequency bands to obtain the second total energy. Divide the sum of the first energy by the sum of the second energy to obtain the percentage of abnormal pulse energy. (2) Extraction of vibration steepness coefficient features: Acquire N time-domain sampling data of the vibration signal, and calculate the mean and standard deviation of the sampling data; The difference between the sampled value and the mean in each time domain is raised to the fourth power, summed, and then divided by the number of samples N to obtain the average value raised to the fourth power. The vibration steepness coefficient is obtained by dividing the fourth power of the average value by the fourth power of the standard deviation. (3) Extraction of cumulative temperature gradient features: Obtain continuous time series temperature detection values; Calculate the temperature difference between two adjacent time points and take the absolute value; The cumulative temperature gradient is obtained by summing the absolute values ​​of the temperature differences at all consecutive time points.

6. The intelligent assessment method for safety early warning of power distribution equipment according to claim 4, characterized in that, The formula for calculating the device's status score is: Where α represents the equipment insulation coefficient, β represents the sensitivity factor, and K ref ΔI represents the kurtosis reference value, γ represents the heat dissipation coefficient, λ represents the electrothermal coupling coefficient, ΔI represents the current change within the preset time window, and ΔT represents the temperature change within the preset time window.

7. The intelligent assessment method for safety early warning of power distribution equipment according to claim 6, characterized in that, The risk propagation simulation based on dynamic causal graphs includes: Initialize the dynamic cause-effect graph by taking the power distribution equipment whose status score exceeds the first threshold as the risk source node; Construct a propagation probability matrix P, and P ij =σ(W i,j ×Score i ); where P i,j W represents the propagation probability from node i to node j in the propagation probability matrix. i,j Score represents the edge weight between node i and node j. i σ represents the state score of the device at the i-th node in the dynamic causal graph, and σ represents the sigmoid function. The risk probability P of each neighboring node j is calculated using a Bayesian network model. risk,j ; Based on the propagation probability P ij and risk probability P risk,j Performing a Monte Carlo simulation yields multiple random propagation paths; Calculate the combined probability P of each random propagation path based on its occurrence frequency. path :P path =P freq ×∏P ij , where P freq P represents the frequency of occurrence of random propagation paths. freq =Number of path occurrences / Total number of simulations, ∏ (i->j)∈随机传播路径 P ij It represents the product of all propagation probabilities along a random propagation path; The combined probabilities of random propagation paths are sorted, and the top M random propagation paths are output to obtain the risk propagation paths.

8. The intelligent assessment method for safety early warning of power distribution equipment according to claim 7, characterized in that, The propagation probability P ij and risk probability P risk,j Performing a Monte Carlo simulation includes: A1. Starting from the risk source node, generate random numbers r∈[0,1]; A2. If r < P i,j The risk is then determined to have propagated from node i to node j, and the following action is taken: When P risk,j When the activation threshold is reached, node j is added to the current propagation path, and steps A1-A2 are recursively executed with node j as the new propagation starting point. A3, when P risk,j If the value is less than the activation threshold, a single simulation is completed, resulting in a single random propagation path. A4, repeat A1-A3, until the number of simulations is reached, resulting in multiple random propagation paths.

9. The intelligent assessment method for safety early warning of power distribution equipment according to claim 7, characterized in that, The triggering of tiered early warnings based on risk propagation paths includes: The risk is classified into levels based on the comprehensive probability of the risk transmission path, and the basis for the classification is as follows: Where Level represents the risk level and length represents the number of nodes in the risk propagation path; Based on the classification results, a warning signal corresponding to the level is sent.

10. A safety early warning intelligent assessment system for power distribution equipment, characterized in that, include: The module includes a data acquisition module, a risk assessment module, and a risk early warning module; among them, The data acquisition module is used to collect real-time operating status data of the power distribution equipment and transmit it to the risk assessment module; The risk assessment module is used to create a digital twin for each power distribution device and construct a dynamic cause-effect graph based on the digital twin; each digital twin performs a status assessment based on the operating status data and generates a status score; when the status score exceeds a first threshold, risk propagation deduction is performed based on the dynamic cause-effect graph and the risk propagation path is output. The risk warning module is used to trigger and issue graded warning information based on the risk propagation path.

Images