Distribution network operation and maintenance system based on digital twinning

Abstract

The invention relates to the technical field of distribution network operation and maintenance, in particular to a distribution network operation and maintenance system based on digital twinning, and the system executes the following steps: constructing a distribution network digital twinning body of a physical distribution network, driving the distribution network digital twinning body to update the electrical state of the distribution network digital twinning body and equipment microenvironment parameters, outputting a global electrical stress distribution diagram and a global microenvironment distribution diagram at the current moment; a virtual temperature sensing grid is generated, thermal dynamic simulation is carried out, and a global temperature field distribution cloud picture is output; and analyzing the global temperature field distribution cloud picture, identifying and tracking a dynamic overheating risk trajectory, predicting an evolution path of the overheating risk germination point in a subsequent prediction environment and a planned load change trend, generating a dynamic migration trajectory of the overheating risk germination point, and outputting targeted early warning and routing inspection positioning information. According to the method, a traditional hysteresis operation and maintenance mode taking'temperature overrun 'as a judgment basis is broken, and qualitative changes from static monitoring to dynamic prediction and from passive warning to active pre-control are truly realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of distribution network operation and maintenance, and particularly relates to a distribution network operation and maintenance system based on digital twinning. BACKGROUND

[0002] With the continuous expansion of the operation scale of the distribution network and the increasing volatility of the load, the overheating point problem caused by equipment overload, poor environment or aging and other factors is increasingly prominent, becoming one of the key hidden dangers affecting power supply safety and equipment life. Especially in densely populated urban areas and high-temperature seasons, some equipment may cause temperature to jump instantaneously due to external micro-environmental disturbances (such as sudden wind speed drop, sunlight direct radiation change, etc.), forming so-called "ghost hot spots" or "short-time thermal migration", which has high uncertainty in location and intensity, bringing great challenges to traditional thermal monitoring methods.

[0003] The existing distribution network operation and maintenance system mainly relies on physical temperature sensors (such as infrared temperature measurement, wireless thermometer) to monitor key nodes at a point. However, due to installation conditions and cost constraints, temperature monitoring points are usually only laid in typical equipment or key nodes, which is difficult to cover all conductors, contacts and connectors and other high-risk parts, leading to hidden dangers being easily missed, and the traditional system relies on static over-temperature threshold for alarm, which often triggers early warning only when the temperature reaches the critical value or even the fault state, lacking dynamic perception and early identification ability for temperature rise trend; most of the existing monitoring methods are used for state recording and abnormal alarm, lacking forward-looking modeling of future thermal risk, and unable to realize predictive inspection and active operation and maintenance; even if some systems consider the environmental temperature, they usually ignore the dynamic influence of complex micro-environmental factors such as wind speed, wind direction and solar radiation, so that the thermal simulation or risk assessment deviates from the actual operation conditions, and cannot accurately reflect the generation mechanism and evolution process of high-temperature risk. SUMMARY

[0004] The present application provides a distribution network operation and maintenance system based on digital twinning, which is a distribution network operation and maintenance system that integrates multi-source data and has high-precision thermal perception and forward-looking analysis capability, and can construct a virtual thermal perception framework covering the whole network based on digital twinning without relying on large-scale sensor deployment, identify the "germination stage" of the overheating risk, and conduct thermal risk dynamic evolution deduction and active early warning combined with predicted working conditions, thereby effectively solving the pain points of "many blind spots, slow response and inability to predict".

[0005] A distribution network operation and maintenance system based on digital twinning, the system performs the following: S1, constructing a power distribution network digital twin of a physical power distribution network, collecting real-time operation data of the physical power distribution network, the real-time operation data including load current data, network topology data, and micro-environment data of devices; injecting the real-time operation data into a preset power distribution network digital twin, driving the power distribution network digital twin to update its electrical state and device micro-environment parameters, and outputting a global electrical stress distribution map and a global micro-environment distribution map at a current time; S2, based on the global electrical stress distribution map and the global micro-environment distribution map, generating a virtual temperature sensor grid and performing thermal dynamic simulation, including generating a virtual temperature sensor grid covering the whole world on all devices and connection points of the power distribution network digital twin; calling an embedded thermodynamic simulation engine, inputting the global electrical stress distribution map as a heat source, and inputting the global micro-environment distribution map as a boundary condition, performing dynamic thermodynamic simulation calculation, and outputting a global temperature field distribution cloud map; S3, analyzing the global temperature field distribution cloud map, identifying and tracking a dynamic overheating risk trajectory, including performing spatio-temporal superposition analysis on the global temperature field distribution cloud map at the current and historical time sequences, identifying a region with a temperature change rate exceeding a preset threshold, and marking it as an overheating risk budding point; predicting the evolution path of the overheating risk budding point under the subsequent predicted environmental and planned load change trend, generating its dynamic migration trajectory, and outputting targeted early warning and inspection positioning information.

[0006] Optionally, the construction of the power distribution network digital twin includes constructing a geometric model and an electrical connection relationship model including all lines, switches, transformers, and load nodes in a virtual space based on a GIS system and a SCADA system of the physical power distribution network; giving each type of element in the model its inherent physical attribute parameters, including line impedance, transformer rated ratio, and load type, to form a static initial model of the power distribution network digital twin.

[0007] Optionally, in S1, the load current data of each node and line is collected through the SCADA system and smart meters; the position state signals of circuit breakers and disconnectors are collected through power distribution automation terminals to generate real-time network topology data; the micro-environment data of devices, including device surface solar radiation intensity, environmental temperature, wind speed, and wind direction, are collected through a micro-environment sensor group deployed on site.

[0008] Optionally, S1 further includes assigning the collected load current data as an excitation source to the current source model of the corresponding node of the power distribution network digital twin; updating the on-off state of the switch element in the power distribution network digital twin according to the network topology data to reconstruct its electrical connection relationship; and assigning the micro-environment data of the devices to the micro-environment parameter attribute of the corresponding device in the power distribution network digital twin; The power distribution network digital twin is driven to perform a power flow calculation, and a global electrical stress distribution map characterized by current density and load rate of each element is output. Meanwhile, the received micro-environment data is interpolated and rendered in three-dimensional space to generate a continuously distributed global micro-environment distribution map.

[0009] Optionally, the generation of the virtual temperature sensing grid comprises automatically generating dense grid nodes at a preset interval on the basis of the geometric topological structure of the three-dimensional geometric topological model of the power distribution network digital twin, on all conductors, connectors, switch contacts and transformer bushing surfaces, to form the virtual temperature sensing grid.

[0010] Optionally, the generation of the virtual temperature sensing grid further comprises locally encrypting and deploying grid nodes for historical heat points, connection points and fastener positions.

[0011] Optionally, the thermodynamic simulation engine is constructed on the basis of computational fluid dynamics and solid heat transfer principles; the heat power density data of each element in the global electrical stress distribution map is mapped to the corresponding nodes of the virtual temperature sensing grid and set as a steady-state heat source; and the environmental temperature, wind speed, wind direction and solar radiation intensity data in the global micro-environment distribution map are set as dynamic boundary conditions for the thermodynamic simulation engine to calculate.

[0012] Optionally, the thermodynamic simulation engine couples the calculation of conductive heat exchange, convective heat exchange and radiative heat exchange to solve the heat balance equation of each node in the virtual temperature sensing grid, perform transient thermal simulation, and simulate the thermal dynamic process of the physical power distribution network under the current electrical stress and micro-environment conditions; and the thermodynamic simulation engine visualizes and interpolates the temperature values of all nodes in the virtual temperature sensing grid calculated by the thermodynamic simulation engine in the three-dimensional space of the power distribution network digital twin, and outputs a global temperature field distribution cloud map reflecting the real-time temperature distribution of the entire power distribution network system.

[0013] Optionally, the S3 specifically comprises: retrieving the global temperature field distribution cloud map at the current time and the historical time series in the previous preset time period, performing spatio-temporal superposition comparison, calculating the temperature change rate of each node in the virtual temperature sensing grid within a preset time window, and identifying and marking the grid nodes whose temperature change rate continuously exceeds a first preset threshold and whose absolute temperature has not exceeded a second preset threshold as overheat risk budding points. The predicted environmental data and planned load data in a future preset period are obtained; the predicted environmental data and planned load data are input into the updated thermodynamic simulation engine, and the overheat risk budding points identified at the current time and the temperature field around them are taken as the initial state to perform forward-looking deduction calculation, simulate the evolution process of the temperature field under future working conditions, and obtain a future temperature field prediction sequence.

[0014] Optionally, the S3 further comprises extracting the continuous change information of the spatial position, temperature intensity and thermal influence range of the overheat risk budding point in the forward deduction calculation result, generating the evolution path of the overheat risk budding point in the spatial and time dimensions, i.e. the dynamic migration track; the dynamic migration track comprises the moving direction of the risk point, the predicted arrival area and the intensity change trend; based on the dynamic migration track, the early warning information including the current position of the overheat risk budding point, the predicted path and the predicted evolution time into the overtemperature fault is generated.

[0015] The beneficial effects of the present application are: 1. The present application reconstructs a three-dimensional dynamic temperature field by constructing a virtual temperature sensing grid covering the surface of the entire network device in the digital twin, combining electrical stress distribution and refined micro-environment data, and driving a thermodynamic simulation engine coupled with CFD+solid heat transfer+radiation model. It breaks the spatial distribution limitation of traditional reliance on physical measurement points, realizes the full coverage, spatial continuity and time dynamic of the thermal sensing capability of the device. Especially for dynamic overheat points caused by sudden wind speed, local shading, load fluctuation, etc., it can identify them when they have not yet shown high temperature but have entered the warming stage, improving the forward-looking and completeness of hidden danger discovery.

[0016] 2. The present application upgrades the thermal simulation engine from a static state reconstruction tool to a future thermal state deduction platform. By introducing scheduling plans and weather forecast data, it constructs the "future current load + environmental boundary" working condition, and performs predictive simulation with the current temperature field as the initial condition, outputting the temperature evolution track in the future continuous time period. This mechanism builds a complete link from current identification -> forward simulation -> risk trajectory prediction -> early intervention, breaks the traditional lagging operation and maintenance mode based on "temperature overrun" as the judgment basis, and truly realizes the qualitative change from static monitoring to dynamic prediction, from passive alarm to active pre-control.

[0017] 3. Based on the dynamic migration track of the risk point output by the simulation, the present application constructs a structured early warning information set and a patrol coordinate set by integrating multi-dimensional data such as temperature change trend, spatial path, and predicted overtemperature time, providing three key decision elements of "when will the risk occur, where will the risk occur, and which point should be patrolled". Compared with the traditional way of only giving temperature overrun alarm, this output method is more operable and scenario adaptable, supporting functions such as automatic dispatching, patrol navigation, risk grading, etc., providing a solid support for intelligent and active operation and maintenance of distribution networks. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to make the technical solutions in the present application or the prior art clearer, the accompanying drawings needed in the embodiments or prior art description will be briefly introduced below. Obviously, the accompanying drawings in the following description only aim at the present application, and other accompanying drawings can also be obtained by those skilled in the art without any creative effort.

[0019] Fig. 1 A step schematic diagram for the system of the embodiment of the present application is shown. Fig. 2 A tracking and early warning schematic diagram of the embodiment of the present application is shown. DETAILED DESCRIPTION

[0020] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments. For some known technologies, other alternative ways can also be implemented by those skilled in the art; and the accompanying drawings are only used to more specifically describe the embodiments, and are not intended to specifically limit the present application.

[0021] As shown in Figs. 1-2 , a distribution network operation and maintenance system based on digital twinning, the system performs the following: S1, constructing a digital twinning of a physical distribution network, collecting real-time operation data of the physical distribution network, the real-time operation data including load current data, network topology data and device micro-environment data; injecting the real-time operation data into a preset digital twinning of the distribution network, driving the digital twinning of the distribution network to update its electrical state and device micro-environment parameters, and outputting a global electrical stress distribution map and a global micro-environment distribution map at the current time.

[0022] S11, constructing a digital twinning of the distribution network: Based on the geographic information system (GIS) and the supervisory control and data acquisition (SCADA) system data of the physical distribution network, a complete digital twinning of the distribution network is constructed in a virtual space, specifically including: geometric structure model , wherein, is a load node, is a line segment (including impedance parameters), is a transformer (including rated transformation ratio), is a control device such as a switch or circuit breaker.

[0023] electrical connection relationship topology model , the connection matrix represents the on-off relationship between nodes.

[0024] Each element gives a set of static physical property parameters: ; wherein, is the line impedance, is the transformer rated transformation ratio, is the node load type (constant power / constant impedance, etc.).

[0025] S12, collect real-time operating data set , including: S121, node / line load current data , obtained through SCADA system and smart meters, including: Each node's injection current: represents the real-time current value of user load or transformer and other equipment at a certain node; Current at both ends of each line: represents the real-time current size and direction flowing through each distribution line; Phase current / line current data: A, B, C three-phase current components commonly seen in three-phase systems; S122, network topology state data , switch / circuit breaker state collected by DTU / FTU, network topology state data is the on-off state information of all switch devices (such as circuit breakers, disconnectors) in the distribution network at time , reflecting which lines in the current distribution network are in a connected state and which are disconnected, thus determining the actual flow path of electric energy in the network; S123, micro-environment data set: ; wherein, is the device surface solar radiation intensity, is the ambient temperature, is the wind speed, is the wind direction.

[0026] S13, inject real-time operating data and drive twin body update: inject to the twin body current source model , wherein is the corresponding node; Use to update the topology model connection matrix , that is: ; Assign the micro-environment parameters in to the attribute domain of the corresponding device model: ; Indicates the numbers of two electrical nodes, for example is the starting node, is the end node, therefore, represents the flow from node to the node . The formula shows that at time , the collected set of micro-environmental parameters of the entire network is filtered and the part of the parameters corresponding to the location of the device is extracted and assigned to the micro-environment attribute field of the device , denoted as . It can also be understood that each device only needs to use the data related to its environment, such as its local temperature, wind speed, and solar intensity, to "take values by location" from the entire network data to form the exclusive environmental input of the device. This is a key step to ensure spatial distribution accuracy in subsequent thermal simulation.

[0027] Assignment process: the coordinate position (x, y, z) of each device in the physical space is taken as input and substituted into the constructed micro-environment continuous distribution field to obtain the specific environmental parameter values such as temperature, sunlight, and wind speed at that point, and assign them to the corresponding parameter variables inside the device model as its current micro-environmental attributes. In simple terms, it is "taking values by location and writing into the model", allowing each device in the twin to have exclusive parameters reflecting its actual environmental state on site, so that the device model can exhibit individual differences and spatial distribution effects in thermal simulation.

[0028] S14, output global distribution map: perform power flow calculation (using Newton-Raphson method) to solve current density distribution and element load rate , output global electrical stress distribution map: ; where, represents the current density at spatial position and time , represents the device load rate at spatial position and time , represents the global electrical stress distribution map output at time , consisting of the set of and ; Based on the micro-environmental parameter field , a three-dimensional interpolation function is used to reconstruct the continuous field: ; Output global micro-environment distribution map: .

[0029] where, represents the time The original micro-environment parameter set collected, including temperature, wind speed, wind direction, irradiance, represents a three-dimensional space interpolation function for constructing discrete micro-environment data into a continuous distribution field, represents the micro-environment interpolation result at the spatial position and time , represents the global micro-environment distribution map output at time , composed of .

[0030] In order to obtain the global electrical stress distribution map of the power distribution network, the first step is to perform power flow calculation, that is, to solve the distribution of node voltages and element currents in the power distribution network. Newton-Raphson method is used to solve the power flow equation, the specific process is as follows: 1. Construct the power flow equation set: based on the topology and parameters of the power distribution network digital twin, establish the power flow balance equation, that is, according to Kirchhoff's current law and Ohm's law, construct a nonlinear equation set about node voltage amplitude and phase angle. These equations represent the functional relationship between power injection and voltage, which is in the form of: active power error equation; reactive power error equation.

[0031] 2. Iterative solution using Newton-Raphson method: use Newton-Raphson method to solve the above nonlinear equation set: initialize the amplitude and phase angle of all node voltages; calculate the power error under the current iteration; construct the Jacobian matrix to reflect the influence of variable changes on the error; solve the linear equation set to get the voltage correction; update the node voltage and judge whether the error meets the convergence condition; If not converged, continue iteration until the error is small enough.

[0032] 3. Calculate the current density and element load rate: once the node voltage is solved, the current value can be calculated according to the impedance and connection relationship of each line in the network model, and then converted to: current density distribution J(x,y,z,t): represents the size of current per unit area at a certain point in space, which is used to identify hot lines; element load rate ρ(x,y,z,t): represents the ratio of current to device rated current, which is used to evaluate whether the device is overloaded.

[0033] 4. Output the electrical stress distribution map: map the current density and load rate results of all elements in three-dimensional space into a visual layer to form a global electrical stress distribution map, which can be used for subsequent thermal simulation or fault prediction analysis.

[0034] In the present application, the original micro-environment parameters are obtained by discrete collection of sensor points deployed near the equipment, that is, in the power distribution network, only a limited number of specific locations obtain these environmental data. To realize thermal simulation of any spatial point in the entire power distribution network, these discrete data need to be expanded into a continuous three-dimensional spatial distribution field. For this purpose, a three-dimensional interpolation function finterp is introduced, and the specific process is as follows: 1. Define the interpolation problem: the original micro-environment data set is composed of multiple discrete points, each point contains its spatial coordinates and the corresponding parameter value (such as temperature , irradiance , etc.), these points constitute the basis data source for interpolation. 2. Construct the three-dimensional interpolation function : according to the interpolation accuracy requirement and the data point distribution characteristics, select the three-dimensional spline interpolation or Gaussian process regression interpolation algorithm to construct , the input of the function is any three-dimensional coordinate point , and the output is the corresponding environmental parameter value.

[0035] 3. Reconstruct the continuous field: use to expand the original sampling data to all locations in the entire power distribution network space to obtain the micro-environment parameter value at each coordinate point and time : ; that is, generate a continuously changing micro-environment parameter field in the entire three-dimensional space, which changes continuously and smoothly with the change of spatial position.

[0036] S2, based on the global electrical stress distribution map and the global micro-environment distribution map, generate a virtual temperature sensor grid and perform thermal dynamic simulation, including generating a virtual temperature sensor grid covering the entire global on all devices and connection points of the power distribution network digital twin; calling the built-in thermodynamic simulation engine, inputting the global electrical stress distribution map as the heat source, and inputting the global micro-environment distribution map as the boundary condition, performing dynamic thermodynamic simulation calculation, and outputting the global temperature field distribution cloud map.

[0037] S21, generate the virtual temperature sensor grid: on the three-dimensional geometric topology model of the power distribution network digital twin, with a preset spatial resolution as the interval, automatically generate a dense set of grid nodes at key positions to form a virtual temperature sensor grid.

[0038] Device parts with thermal hazard history , adopt local grid encryption strategy, set smaller grid spacing , enhance the temperature resolution of the region.

[0039] represent the first virtual temperature sensor grid node, with specific spatial coordinates , used to carry local heat source input, environmental boundary conditions, and as a basic calculation unit during simulation, represent the first local area of the device with known or potential thermal hazards, such as joints, fasteners, connections, and in such areas, the system will perform "local grid encryption", that is, more nodes are arranged with smaller spacing to improve the accuracy and sensitivity of thermal simulation.

[0040] In S21, first according to the geometric structure of the power distribution network, in the following key parts: conductor (such as overhead line, busbar); switch contacts and connectors (high electrical connection accuracy, frequent thermal hazards); transformer bushings and fins (power-intensive, obvious heat accumulation); A series of grid nodes are generated with a uniform spatial spacing (called resolution) to form a three-dimensional discrete virtual temperature sensor grid, enabling the entire device to have "temperature sensing capability" in simulation.

[0041] At the same time, special parts such as bolts, welded joints, and terminal posts that have experienced high temperature, breakdown, and loosening risks in the history record will be targeted, and smaller spacing will be used for local encryption in these areas to achieve: Obtain more detailed temperature change curve in key parts; Capture "local overheating points" or "temperature mutation zones" to provide basis for subsequent early warning.

[0042] Breakthrough traditional thermal monitoring relying on a few temperature sensors, achieve "digital equivalent to cover temperature sensors", provide high-resolution basic data grid for subsequent calculation-based thermal field construction and risk trajectory identification.

[0043] S22, call the thermodynamic simulation engine and configure the calculation parameters: call the built-in thermodynamic simulation engine , which is based on the following physical modeling principles: Computational Fluid Dynamics (CFD): simulates convective heat transfer behavior in air; Solid heat transfer model: solve the heat conduction process in conductors; Radiation Model: Consider the radiation heating effect caused by sunlight.

[0044] I. Computational Fluid Dynamics (CFD): Simulate convective heat transfer behavior in air.

[0045] In open spaces, electrical equipment such as wires, switch contacts, or transformer bushings not only conduct heat outward through their own structures, but also exchange heat with the environment through air media. This part is mainly handled by the CFD module, which simulates: Natural convection: After the device heats up, the surface temperature rises, the surrounding air rises due to heating, forming a temperature-driven flow that carries away heat; Forced convection: When there is wind (wind speed and direction have been obtained through micro-environment data), air flows over the device surface at a faster speed, enhancing heat transfer effect.

[0046] The simulation engine will solve the fluid control equation set (continuity equation, momentum equation and energy equation), based on device surface shape, wind speed and direction, etc. Simulate air flow path and velocity field, and then calculate the heat transfer rate through convection.

[0047] II. Solid Heat Transfer Model: Solve the heat conduction process in conductors.

[0048] This is the basic part of thermal simulation. All conductive components in the power distribution network (copper wires, contacts, conductive connectors) will heat up due to Joule effect when current passes through, and this heat will conduct within the solid.

[0049] The simulation engine will take the heat power density of each device part obtained from the electrical stress distribution map as a heat source term, and substitute it into the three-dimensional solid heat conduction differential equation: This part of the model simulates how heat spreads in the conductor structure inside the device, and is the basis for evaluating heat accumulation, thermal gradient distribution, etc.

[0050] III. Radiation Model: Consider the radiation heating effect caused by sunlight.

[0051] Power distribution equipment operating outdoors will also be affected by solar radiation, especially in summer or sunny periods. This part is a heat source enhancement term. The simulation engine calculates the solar radiation power density Rsun received by each device surface, and combines its absorption rate, emissivity, and surface material reflection characteristics to convert it into temperature change: Surface receives sunlight → temperature rises; Temporary shadow blockage → temperature drops rapidly.

[0052] The simulation engine dynamically adjusts the radiant heat input based on the irradiance data provided in the microenvironment distribution map, combined with the solar altitude angle model and the spatial orientation of the equipment. This is the key to simulating the phenomenon of "sudden temperature changes".

[0053] The configuration input parameters are as follows: 1. Assigning values ​​to heat source terms: Assigning values ​​to the global electrical stress distribution map. Current density of each component With resistivity Mapped to heat source power density The formula is as follows: ;in, For the first The heat power density of each grid node The current density at this node, The resistivity of the material at that node.

[0054] First, the current density data obtained from the electrical stress distribution map is used. Combined with the resistivity of the equipment materials Calculate the heat power density per unit volume of each simulation grid node. The above is based on Joule's theorem, which describes the process by which energy is converted into heat due to resistance when current flows through a conductor. Each node independently calculates its corresponding heating intensity based on the current density and resistivity at its location. The calculation results will be used as a "steady-state heat source term" in the thermodynamic simulation engine to simulate the self-heating process generated by current flow inside the device.

[0055] 2. Boundary condition configuration: From the global microenvironment distribution map The following parameters are extracted and mapped to the boundary function of each boundary node in the three-dimensional spatial field: ;in, For ambient temperature, For wind speed, For wind direction, This represents the intensity of solar radiation.

[0056] The boundary conditions are configured to simultaneously extract key environmental parameters surrounding the power distribution equipment from the global microenvironment distribution map, including ambient temperature, wind speed, wind direction, and solar radiation intensity. These parameters are mapped to each boundary node in three-dimensional space and used as boundary input conditions for thermal simulation. Specifically: Ambient temperature affects the temperature difference between the equipment surface and the air, and is an important driving factor for convective and radiative heat transfer. Wind speed and direction determine the path and velocity of airflow, which in turn affect the intensity of convective heat transfer. The intensity of solar radiation determines the amount of solar energy absorbed by the surface of the equipment, forming an external radiative heat source.

[0057] These boundary conditions change dynamically and can be updated in real time according to time and spatial location, making the thermal simulation process closer to the real operating environment.

[0058] S23, Perform dynamic thermodynamic simulation calculations: for each grid node Thermodynamic simulation engine Using the transient thermal equilibrium equation: ; in, For material density, For specific heat capacity, Thermal conductivity, For the first The temperature of the node, This is an endogenous heat source term; For the convective heat transfer term (determined by wind speed / direction), local wind speed and direction data of each equipment surface are obtained through the microenvironment distribution map. Combined with the temperature difference between the surface temperature and the surrounding air temperature, the intensity of convective heat exchange between the air and the equipment is solved using a computational fluid dynamics (CFD) model.

[0059] The calculation of convective heat exchange is based on the Newsell number relation, using the following: a. Obtain local wind speed and equipment geometric characteristics; b. Calculate the local convective heat transfer coefficient ; c combined with temperature difference ;calculate: This value represents the heat dissipated into the air per unit area through convection; For the radiative heat transfer term (determined by irradiance intensity), the solar radiation intensity is collected from the microenvironment data. Combined with the radiation absorptivity of the equipment surface Emission rate The net radiative heat flux density absorbed by each surface node is calculated by considering the perspective factor relative to the sky or environment. This is primarily estimated using the radiative heat transfer equation: ;in, Here, represents the Stefan-Boltzmann constant; the first term represents the heat absorbed by shortwave solar radiation, and the second term represents the longwave radiation exchange between the device and the environment. The simulation engine adjusts the parameters based on the device's spatial orientation and the angle of sunlight received. The calculation results were corrected to make them more closely reflect reality.

[0060] The simulation solution uses finite volume or finite difference methods for discretization, progressively advancing the time axis. And calculate the heat transfer evolution process, get each node at each time point of temperature change results .

[0061] S24, output the global temperature field distribution cloud picture: all nodes The simulation temperature results , in the three-dimensional space of the twin body through the visualization interpolation function Space reconstruction, output continuous field: ; Generate intuitive high-precision global temperature field distribution cloud picture reflecting the real-time thermal state of the entire power distribution network, which is used for subsequent risk identification and early warning.

[0062] The core goal of S24 is to convert the large amount of discrete temperature data calculated by the thermal simulation engine into a continuous distribution three-dimensional temperature field that can be intuitively presented, thereby forming a "global temperature field distribution cloud picture" for displaying the current thermal operating state of the entire power distribution network. Specifically, at each node in the aforementioned virtual temperature sensor grid, the temperature simulation has been completed, and the temperature value at the current time has been obtained. Since these nodes are scattered point data, they cannot be directly used to construct a continuous spatial heat map, so visualization interpolation processing is needed. For this purpose, a visualization interpolation function is introduced to reconstruct a smooth and continuous temperature distribution field in the three-dimensional geometric space of the power distribution network twin body based on the temperature values and spatial coordinates of each node. This process can be understood as "filling the entire power distribution network structure surface and internal space with point temperature data", so that each location has a corresponding temperature value at any time. Finally, the interpolated temperature field data is rendered in three-dimensional space to generate a color cloud picture, with color representing temperature and shape fitting the device structure, which can clearly reflect: The distribution trend of heat in the equipment; The specific location of local high temperature or overheating point; Whether the overall temperature field change shows abnormal evolution characteristics.

[0063] This output is not only the basic data layer for analyzing overheating risk, but also facilitates the intuitive understanding of system operation status by operation and maintenance personnel, and is the key visualization result for realizing digital inspection and intelligent early warning.

[0064] S3, analyze the global temperature field distribution cloud picture, identify and track dynamic overheating risk trajectory, including spatiotemporal superposition analysis of the global temperature field distribution cloud picture at the current and historical time series, identifying areas with temperature change rate exceeding the preset threshold as overheating risk germination points; Predict the evolution path of the overheating risk germination points under the subsequent predicted environmental and planned load change trend, generate their dynamic migration trajectory, and output targeted early warning and inspection positioning information.

[0065] S31, Spatiotemporal Overlay Analysis and Incipient Temperature Identification: The aim is to identify potential overheating risk points that, although the current temperature has not yet exceeded the warning value, have already shown a rapid temperature rise trend; these are known as "incipient overheating risk points." The core idea is not to wait until the temperature exceeds the limit before issuing an alarm, but to identify the anomaly in its early stages, facilitating preventative measures. First, the system retrieves the global temperature field distribution cloud map for the current moment and a previous set time period. This map reflects the temperature distribution of each device surface node over time in three-dimensional space. Then, for each node in the virtual temperature sensing grid, the rate of temperature change within that time period is calculated, i.e., the rate of temperature increase per unit time. Next, the rate of change for each node is judged using two criteria: First, does the rate of temperature increase exceed a set threshold for the rate of change (indicating that the temperature rise is "fast"); Second, is the current temperature still within a safe range (meaning "not yet exceeded the temperature limit").

[0066] Only nodes that meet both of these conditions will be identified and marked as "early signs of overheating risk". These points are often in an early abnormal state and have not yet reached the triggering conditions of traditional temperature alarm systems, but if they continue to develop, they are very likely to evolve into actual overheating faults in the future.

[0067] The core advantage of this method is that it not only considers the "absolute value" of temperature, but also introduces the "trend" and "rate" of temperature change, which improves the timeliness and foresight of early warning and is a basic step in realizing predictive maintenance.

[0068] The specifics are as follows: Retrieve the current time and its previous preset time window Global temperature field distribution cloud map For each node in the virtual temperature sensing mesh Perform spatiotemporal analysis. For each node... Calculate its rate of temperature change within this time window. : ; The judgment criteria are as follows: If the following conditions are met: Then the node Marked as the incipient point of overheating risk ;in, The first preset threshold is the temperature change rate threshold. The second preset threshold is the absolute temperature safety upper limit. The set of identified germination points.

[0069] temperature change rate threshold The temperature change rate threshold can be set as 2.5 times the mean value of the normal fluctuation range according to the temperature rise rate history data during the normal operation of the reference device. Under stable load and environmental conditions, the temperature rise rate of the surface of the electrical device is usually slow and stable, and abnormal temperature rise is usually manifested as a sudden acceleration of the temperature rise rate.

[0070] absolute temperature safety upper limit The absolute temperature safety upper limit is usually set as a safety margin temperature slightly higher than the rated working temperature according to the temperature resistance level of the device material, industry standards or manufacturer given values, for example, for a device with a long-term working temperature of 85°C, it can be set to 90°C or 95°C. The rated insulation level of the device material or the limitation of the device working temperature in the national standard, the material of the electrical device such as wire, switch contact, transformer, etc. has a clear temperature resistance level (such as A class 105°C, B class 130°C, etc.), the performance of the device has begun to deteriorate before reaching this value.

[0071] S32, constructing a risk evolution path model: obtaining future prediction period , within: predicted environmental data ; planned load data ; represents the length of the prediction time period, i.e. the time range of the temperature field evolution deduced by the thermodynamic simulation engine into the future, which is used to simulate the overheating risk development trend in the future period of time.

[0072] connected to the environmental data service interface, weather bureau API or locally deployed environmental monitoring system, call the meteorological data in the future prediction period, the predicted environmental data specifically includes: environmental temperature, wind speed and direction, solar radiation intensity in the prediction period, these data can be spatially interpolated according to the area where the device is located, forming a predicted micro-environment parameter field with geographical distribution characteristics, and constructing into a time sequence form input. Synchronously call the planned load data provided by the power distribution dispatching system or energy consumption prediction platform, i.e. the expected power consumption, load curve and device working state arrangement of each load node in the future period. The planned load data is derived from: load prediction model (based on historical power consumption behavior, day type, holiday), short-term dispatching plan of SCADA system, user side operation scheduling plan, map these data to each node and line in the power distribution network digital twin, and construct the future current density prediction sequence.

[0073] input the above prediction data into the updated thermodynamic simulation engine , and the current temperature field distribution and the risk germination point set Using these as initial conditions, a dynamic thermal simulation is performed to predict the future temperature field sequence. .

[0074] Specifically, at the current moment The temperature field distribution map (i.e., the current three-dimensional temperature result) is used as the initial temperature condition for the thermal simulation, and the identified overheating risk initiation points are designated as key simulation areas. The surrounding areas are subjected to more detailed thermal response calculations at the mesh level. The following content is input into the updated thermodynamic simulation engine: Internal heat source term: The current density is calculated from the predicted load data, and then the predicted heating power of each grid point is calculated. Dynamic boundary conditions: continuously changing boundaries generated from predicted environmental data, including temperature, wind speed, wind direction, and sunshine. Initial temperature field state: the temperature distribution of the equipment at the current moment, ensuring that the simulation starts from the real working conditions.

[0075] The simulation engine then couples heat conduction, convection, and radiation mechanisms, progressively advancing the time dimension to simulate the dynamic thermal response of the equipment over future periods as load and environment change together. Once the simulation is complete, it outputs the results for each moment. The three-dimensional temperature distribution forms a set of continuous temperature field snapshots, collectively referred to as the future temperature field prediction sequence, denoted as... This sequence fully depicts the spatial and temporal evolution of temperature changes, providing a data foundation for subsequent dynamic risk trajectory generation and over-temperature early warning.

[0076] S33, Generating Dynamic Migration Trajectories: The aim is to extract the thermal evolution behavior of each identified risk initiation point over a future period based on the predicted future temperature field sequence, constructing its "dynamic trajectory" in space and time, i.e., dynamic migration path. This process not only reflects the location changes of the risk point but also reveals whether its temperature intensity and impact range are expanding, providing a basis for decision-making in subsequent early warning level classification and inspection deployment. For each initiation point... From the predicted temperature field Extract its spatial location, temperature intensity, and thermal influence range over time to construct its dynamic migration trajectory.

[0077] I. Extracting Spatial Location Changes: First, the temperature evolution around each risk initiation point is tracked in the predicted temperature field sequence. Since the initiation point may "move" spatially due to heat diffusion, load changes, or environmental disturbances, at each prediction time, the region around that point where the temperature is still above a certain relative threshold is searched, and the temperature extreme point or centroid location in that region is found as the "representative coordinates" of the risk point at that time. This method ensures that the risk trajectory is not only based on the original location but also dynamically reflects the migration direction and path of the actual hotspot.

[0078] II. Extracting temperature intensity variation: record the temperature value at the risk point coordinate at each time, forming a sequence of temperature intensity over time, which is used to assess whether the risk point is continuously warming up, approaching or exceeding the absolute safety threshold, and serves as the core indicator for risk level assessment.

[0079] III. Extracting thermal influence range variation: in the temperature field at each time, calculate the "isothermal area" around the risk point, i.e. the continuous spatial range with temperature greater than a reference temperature (90% of the current point temperature). Through three-dimensional space volume or surface area calculation, obtain the "thermal influence range" of the risk point, continuously record the variation of these ranges, and judge whether the risk has a "diffusion" trend, i.e. from a point-like local anomaly to a sheet-like or even regional high temperature zone.

[0080] The dynamic migration trajectory is represented as: ; where, represents the spatial position of the risk point over time, is the predicted temperature intensity, is the predicted thermal influence range, i.e. the isothermal area. According to the time sequence, the dynamic migration trajectory of the risk point is constructed, and the trajectory is visualized as a continuously extended path curve in space. Each point on the path carries the temperature value and influence range information at that time, similar to a "time-labeled thermal risk cloud cluster".

[0081] The output content includes: Migration direction (vector from current position to future predicted path), by calculating the tangent direction of the path curve, determine the direction the risk point will move in the future; Arrival area (spatial range where the risk point may migrate to in the future), combined with the spatial range expansion trend, predict the area that may be affected by the risk; Intensity trend (whether the temperature is continuously rising, whether the range is expanding), analyze the sequence variation of temperature values, and determine whether the temperature is rising, stable or falling.

[0082] S34, output warning and patrol positioning information: based on each migration trajectory , construct a multi-dimensional warning information set: Risk point current position: ; Dynamic path prediction map: Sequence trajectory; Over-temperature fault prediction time : when the first time, as the time window for predicting the occurrence of faults.

[0083] Generate the final warning information package: ; Meanwhile, the region coordinate set with the highest risk probability is screened out from all the migration trajectories , which is used to generate accurate inspection positioning information and guide the operation and maintenance personnel to quickly go to the potential hidden danger point.

[0084] The generation is as follows: For each overheat risk point migration trajectory , the following risk features corresponding to each trajectory point are analyzed step by step according to time slicing: Whether the temperature value is continuously rising and approaching the preset upper limit of temperature; Whether the temperature rise rate is rising, that is, whether the temperature change trend is intensifying; Whether the trajectory is close to the key equipment or the historical failure area; Whether the time difference between the current time and the predicted failure time point is short enough, that is, it will evolve into a failure.

[0085] Instead of generating a single comprehensive score, whether each trajectory point meets the above high-risk features is marked, such as: Marked as "high temperature approaching"; Marked as "rate anomaly"; Marked as "near failure"; Marked as "close to sensitive components".

[0086] These marks can be superimposed, indicating that the higher the risk composite degree is.

[0087] Among all the trajectory points, the node positions that meet the following conditions at the same time are preferentially screened out: Multiple risk feature marks appear at the same time; Closest to the current time (about to happen); Not repeated (avoid multiple risk points focusing on the same physical area); High spatial concentration (that is, multiple risk trajectories tend to the same small range area).

[0088] The spatial coordinates of these preferential points screened out are extracted to form the final region coordinate set with the highest risk probability.

[0089] The present application covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present application. In order to make the public have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can also be completely understood without the description of these details for those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits, etc. are not described in detail.

[0090] The above merely describes the preferred embodiments of the present application, and it should be pointed out that, for those skilled in the art, some improvements and refinements can be made without departing from the principles of the present application, and these improvements and refinements should also be considered as falling within the protection scope of the present application.

Claims

1. A distribution network operation and maintenance system based on digital twinning, characterized in that, The system performs the following: S1, constructing a power distribution network digital twin of a physical power distribution network, collecting real-time operation data of the physical power distribution network, the real-time operation data including load current data, network topology data, and device micro-environment data; injecting the real-time operation data into a preset power distribution network digital twin, driving the power distribution network digital twin to update its electrical state and device micro-environment parameters, and outputting a global electrical stress distribution map and a global micro-environment distribution map at the current time; S2, based on the global electrical stress distribution map and the global micro-environment distribution map, generating a virtual temperature sensor grid and performing thermal dynamic simulation, including generating a virtual temperature sensor grid covering the entire domain on all devices and connection points of the power distribution network digital twin; calling an embedded thermodynamic simulation engine, inputting the global electrical stress distribution map as a heat source, and inputting the global micro-environment distribution map as a boundary condition, performing dynamic thermodynamic simulation calculation, and outputting a global temperature field distribution cloud map; S3, analyzing the global temperature field distribution cloud map, identifying and tracking dynamic overheating risk trajectories, including performing spatio-temporal superposition analysis on the current and historical time series of the global temperature field distribution cloud map, identifying areas with temperature change rates exceeding a preset threshold, and marking them as overheating risk germination points; predicting the evolution path of the overheating risk germination points under the subsequent predicted environmental and planned load change trend, generating their dynamic migration trajectories, and outputting targeted early warning and inspection positioning information.

2. The distribution network operation and maintenance system based on digital twinning according to claim 1, characterized in that, The construction of the power distribution network digital twin includes constructing a geometric model and an electrical connection relationship model including all lines, switches, transformers, and load nodes in a virtual space based on the GIS system and the SCADA system of the physical power distribution network; giving each type of element in the model its inherent physical property parameters, including line impedance, transformer rated ratio, and load type, to form a static initial model of the power distribution network digital twin.

3. The distribution network operation and maintenance system based on digital twinning of claim 2, wherein, In S1, the load current data of each node and line is collected through the SCADA system and smart meters; the position state signals of circuit breakers and disconnectors are collected through power distribution automation terminals to generate real-time network topology data; the micro-environment data of the devices, including device surface solar radiation intensity, environmental temperature, wind speed, and wind direction, are collected through the micro-environment sensor group deployed on site.

4. The distribution network operation and maintenance system based on digital twinning of claim 3, wherein, S1 also includes assigning the collected load current data as excitation sources to the current source model of the corresponding nodes of the power distribution network digital twin; updating the on-off state of the switch elements in the power distribution network digital twin according to the network topology data to reconstruct the electrical connection relationship; assigning the device micro-environment data to the micro-environment parameter attributes of the corresponding devices in the power distribution network digital twin; The updated power distribution network digital twin is driven to perform power flow calculation, outputting a global electrical stress distribution map characterized by the current density and load rate of each element; at the same time, the received micro-environment data is interpolated and rendered in three-dimensional space to generate a continuously distributed global micro-environment distribution map.

5. The distribution network operation and maintenance system based on digital twinning of claim 1, wherein, The generation of the virtual temperature sensing grid includes automatically generating dense grid nodes at a preset interval on the basis of the geometric topological structure of the power distribution network digital twin, on all conductors, connectors, switch contacts, and transformer bushing surfaces, to form the virtual temperature sensing grid.

6. The distribution network operation and maintenance system based on digital twinning of claim 5, wherein, The generation of the virtual temperature sensing grid also includes locally encrypting and deploying grid nodes for historical heat points, connection points, and fastener locations.

7. The distribution network operation and maintenance system based on digital twinning of claim 5, wherein, The thermodynamic simulation engine is constructed on the basis of computational fluid dynamics and solid heat transfer principles; the heat power density data of each element in the global electrical stress distribution map is mapped to the corresponding nodes of the virtual temperature sensing grid and set as a steady-state heat source; and the environmental temperature, wind speed, wind direction, and solar radiation intensity data in the global microenvironment distribution map are set as dynamic boundary conditions for the thermodynamic simulation engine to calculate.

8. The distribution network operation and maintenance system based on digital twinning according to claim 7, characterized in that, The thermodynamic simulation engine couples the calculation of conductive heat exchange, convective heat exchange, and radiative heat exchange, solves the heat balance equation of each node in the virtual temperature sensing grid, performs transient thermal simulation, and simulates the thermal dynamic process of the physical power distribution network under the current electrical stress and microenvironment conditions. The thermodynamic simulation engine visualizes and renders the temperature values of all nodes in the virtual temperature sensing grid calculated by the thermodynamic simulation engine in the three-dimensional space of the power distribution network digital twin, and outputs a global temperature field distribution cloud map reflecting the real-time temperature distribution of the entire power distribution system.

9. The distribution network operation and maintenance system based on digital twinning of claim 1, wherein, The S3 specifically includes: retrieving the global temperature field distribution cloud map at the current time and the historical time series in the previous preset time period, performing spatio-temporal superposition comparison, calculating the temperature change rate of each node in the virtual temperature sensing grid within a preset time window, and identifying and marking the grid nodes with a temperature change rate continuously exceeding a first preset threshold and an absolute temperature not exceeding a second preset threshold as overheat risk budding points; predicted environmental data and planned load data in a future preset period are obtained; the predicted environmental data and planned load data are input into the updated thermodynamic simulation engine, and the overheat risk budding points currently identified and their surrounding temperature fields are taken as the initial state to perform forward-looking deduction calculation, simulate the evolution process of the temperature field under future working conditions, and obtain a future temperature field prediction sequence.

10. The distribution network operation and maintenance system based on digital twinning according to claim 9, characterized in that, The S3 also includes extracting the continuous change information of the spatial position, temperature intensity, and thermal influence range of the overheat risk budding points in the forward-looking deduction calculation result, generating an evolution path of the overheat risk budding points in the spatial and temporal dimensions, i.e., a dynamic migration trajectory; The dynamic migration trajectory includes the moving direction, the predicted arrival area, and the intensity change trend of the risk point; and based on the dynamic migration trajectory, warning information including the current position of the overheat risk budding point, the predicted path, and the predicted time of evolving into an overtemperature fault is generated.

Images