Power transmission line operation and maintenance analog simulation system based on digital twinborn technology

Abstract

The invention discloses a power transmission line operation and maintenance analog simulation system based on a digital twinborn technology, and relates to the technical field of intelligent operation and maintenance and simulation of a power system, and the system comprises a data collection module which is used for collecting real-time operation state data of a power transmission line and establishing a data link, the data comprises meteorological information, current-carrying capacity, temperature, tension and conductor sag multi-source heterogeneous operation state data, the model building and adjusting module is used for building or adjusting virtual simulation model parameters based on the real-time operation state data, and the comparative analysis module is used for analyzing the real-time operation state data. The analysis module is used for carrying out multi-dimensional comparison analysis on the updated virtual simulation model and an actual power transmission line, and comparison dimensions comprise a tension change trend, wire vibration response and thermal stability distribution; according to the power transmission line operation and maintenance analog simulation system based on the digital twinborn technology, the automation, intelligence and scientization levels of power transmission line operation and maintenance management are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance and simulation technology of power systems, specifically to a transmission line operation and maintenance simulation system based on digital twin technology. Background Technology

[0002] In existing power systems, transmission lines serve as critical channels for power transmission, and their operational status directly impacts the safety and stability of the power grid. Traditional operation and maintenance methods primarily rely on manual inspections, periodic maintenance, and partial sensor data acquisition, which suffer from low efficiency, information lag, and insufficient status coverage. With the increasing complexity of transmission line operating environments, such as climate change, icing, and wind-induced vibration, existing operation and maintenance methods struggle to comprehensively and promptly grasp the status of lines throughout their entire lifecycle. This results in limited fault identification capabilities and delayed early warning mechanisms, ultimately affecting the scientific rigor and accuracy of operation and maintenance decisions.

[0003] The introduction of digital twin technology has provided a new direction for transmission line operation and maintenance simulation. By constructing a virtual mirror of the physical line, it enables the mapping and prediction of its state, behavior, and evolution process. However, existing digital twin systems for transmission lines are mostly static models, failing to effectively integrate real-time operational status data. This leads to discrepancies between the virtual model and the actual operating environment, especially when dynamic parameters such as wind speed, temperature, icing thickness, and current load change significantly, resulting in a substantial reduction in simulation accuracy. Furthermore, the current lack of efficient data acquisition, processing, and modeling mechanisms prevents dynamic updates and closed-loop optimization of the virtual model, thus hindering its online monitoring, fault simulation, and predictive analysis capabilities. Summary of the Invention

[0004] The purpose of this invention is to provide a transmission line operation and maintenance simulation system based on digital twin technology, which dynamically updates the virtual simulation model according to the real-time operating status data of the transmission line to solve the problem of line parameters deviating from the actual operating environment.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a transmission line operation and maintenance simulation system based on digital twin technology, the system comprising: The data acquisition module is used to collect real-time operating status data of the transmission line and establish a data link. The data includes meteorological information, current carrying capacity, temperature, tension, conductor sag and other multi-source heterogeneous operating status data. The model building and adjustment module is used to build or adjust the parameters of the virtual simulation model based on real-time running status data. The comparative analysis module is used to perform multi-dimensional comparative analysis between the updated virtual simulation model and the actual transmission line. The comparative dimensions include tension change trend, conductor vibration response and thermal stability distribution. The dynamic optimization module is used to dynamically optimize and iteratively update the virtual simulation model based on the results of comparative analysis. The data fusion processing module communicates with the model building and adjustment module. It is used to preprocess multi-source operating status data and adopts an adaptive weight fusion algorithm to integrate different data sources to generate model input parameters, thereby improving the update accuracy and stability of the simulation model.

[0006] Preferably, the data acquisition module includes a meteorological sensor, a conductor status sensor, a tower tilt angle monitor, and an infrared image acquisition device, which are used to collect data on wind speed, temperature, humidity, rainfall intensity, conductor temperature, conductor tension, sag curvature change, tower displacement, and thermal anomaly characteristics of the line surface in the environment where the transmission line is located.

[0007] Preferably, the data fusion processing module includes a time synchronization unit, a data cleaning unit, and a weight dynamic calculation unit. The time synchronization unit is used to calibrate data from multiple different data sources according to a unified timestamp. The data cleaning unit is used to remove missing values ​​and abnormal jump values. The weight dynamic calculation unit dynamically adjusts the fusion weight coefficients based on the source stability, historical volatility, and contextual relevance of each type of data.

[0008] Preferably, the model construction and adjustment module includes a parameter calibration submodule and a structural modeling submodule. The parameter calibration submodule automatically corrects the errors in the initial model input parameters by comparing them with historical operating condition data. The structural modeling submodule generates a virtual simulation space model based on 3D GIS information, tower physical structure data, and line path topology.

[0009] Preferably, the comparison analysis module employs a multi-feature fusion anomaly identification algorithm, a time series similarity algorithm for tension change trends, a frequency domain energy spectrum analysis method for conductor vibration response, and a distribution uniformity analysis and regional temperature rise gradient comparison for thermal stability distribution. It comprehensively evaluates the degree of deviation between the simulation model and the actual line in different dimensions and outputs a quantitative score.

[0010] Preferably, the dynamic optimization module includes a model error inversion submodule and a simulation strategy adaptation submodule. The model error inversion submodule analyzes the key input parameters that generate errors based on the comparative analysis results and corrects them. The simulation strategy adaptation submodule automatically selects different optimization strategies, such as linear fine-tuning optimization, local reconstruction, or overall reconstruction, according to different deviation levels.

[0011] Preferably, the comparison analysis module supports historical trend alignment, which can extract the characteristic curves of tension and temperature rise changes in different time periods and compare them with historical typical working condition curves to help identify potential structural aging or environmental stress accumulation.

[0012] Preferably, the data fusion processing module has a spatiotemporal consistency verification function, which performs structural similarity analysis on the data of adjacent towers or line segments in space, and makes a continuity judgment on the data fluctuation curve in time.

[0013] Preferably, the visualization module integrates a multi-level information view switching function, which supports quick switching from the panoramic view of the transmission line to a local view of a single tower or line segment, and dynamically annotates the model deviation heat map, risk level assessment map and operation and maintenance suggestion prompt box.

[0014] Preferably, the dynamic optimization module is equipped with a confidence interval rollback mechanism, which will automatically roll back to the previous stable model version when the simulation results fluctuate beyond the statistical confidence interval.

[0015] As can be seen from the above technical solution, the present invention has the following beneficial effects: This transmission line operation and maintenance simulation system based on digital twin technology, through the inclusion of a data acquisition module, can acquire multi-source heterogeneous data such as meteorological conditions, conductor tension, current carrying capacity, temperature, and sag during transmission line operation in real time. This forms a dynamic panoramic monitoring of the line's operating status, providing data support for the high fidelity of the simulation model. The model building and adjustment module maps the acquired real-time data to the virtual simulation model and updates it in conjunction with the structural and parametric models, effectively reducing the error between the model and the actual line operating status, significantly improving modeling accuracy and simulation fit. The comparative analysis module uses a multi-dimensional comparison method to compare the differences between the virtual model and the real line operation from the perspectives of tension change trends, conductor vibration response, and thermal stability distribution, enabling the identification of abnormal deviations. The rapid identification and location of potential risks helps to discover them in advance. By setting up a dynamic optimization module, the system can automatically optimize model parameters or structure based on comparison results and support model iterative evolution, thereby constructing a simulation system with adaptive capabilities to meet the long-term operation and maintenance simulation needs under complex working conditions. The data fusion processing module introduces an adaptive weight fusion algorithm to dynamically weight data from different sources, with different precision and time delays, improving the confidence and stability of the model input data, and thus enhancing the robustness and predictive reliability of the overall simulation system. This system has the ability to perceive, simulate, reconstruct, and predict the trend of line operation status in real time, and can provide operation and maintenance personnel with visualized early warning information and strategy suggestions, thereby improving the automation, intelligence, and scientific level of transmission line operation and maintenance management. Attached Figure Description

[0016] Figure 1 This is a system connection diagram of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, the present invention provides a technical solution: a transmission line operation and maintenance simulation system based on digital twin technology, the system comprising: The data acquisition module is used to collect real-time operating status data of the transmission line and establish a data link. The data includes meteorological information, current carrying capacity, temperature, tension, conductor sag and other multi-source heterogeneous operating status data. The model building and adjustment module is used to build or adjust the parameters of the virtual simulation model based on real-time running status data. The comparative analysis module is used to perform multi-dimensional comparative analysis between the updated virtual simulation model and the actual transmission line. The comparative dimensions include tension change trend, conductor vibration response and thermal stability distribution. The dynamic optimization module is used to dynamically optimize and iteratively update the virtual simulation model based on the results of comparative analysis. The data fusion processing module communicates with the model building and adjustment module. It is used to preprocess multi-source operating status data and adopts an adaptive weight fusion algorithm to integrate different data sources to generate model input parameters, thereby improving the update accuracy and stability of the simulation model.

[0019] This implementation method uses power transmission lines as the object, constructing a closed-loop simulation system with real-time monitoring, modeling, analysis, and optimization capabilities. First, various types of sensors deployed on the power transmission lines, such as tension sensors, infrared thermometers, anemometers, current transformers, and video monitoring equipment, collect real-time operational status data across multiple dimensions, including temperature, tension, conductor sag, wind speed, and current load. This raw data is then uploaded to the data fusion processing module via edge acquisition terminals through 4G / 5G networks or fiber optic communication links.

[0020] In the data fusion processing module, the system first performs time synchronization processing on data from different sources, using a linear interpolation algorithm to unify timestamps and align data of different frequencies to a unified sampling period. For example, temperature data and tension data are collected at 1-second and 2-second intervals respectively, and then unified to a 1-second period. Subsequently, the system performs noise reduction processing on the collected data, using wavelet transform to eliminate high-frequency noise and ensure data stability.

[0021] Next, the system introduces an adaptive weighted fusion algorithm, which dynamically calculates the weights of each data source based on the criterion of minimizing historical errors. Specifically, assuming that the average errors of temperature data from sensors A and B compared to the calibration values ​​over the past 10 minutes are 0.2 and 0.5 degrees Celsius, respectively, the system assigns a higher fusion weight to A, for example, A's weight is 0.71, and B's weight is 0.29. Then, the fused temperature input parameters are generated using a weighted average. All fusion results are transmitted as model input parameters to the model building and tuning module.

[0022] The model building and adjustment module uses a multibody dynamics method and a thermodynamic simulation model to collaboratively construct the simulation scenario. Taking a conductor as an example, the system establishes a mathematical model based on the conductor material parameters (such as elastic modulus, coefficient of thermal expansion, and resistivity) and input parameters such as tension, temperature, and current to calculate the conductor's thermal stability and sag response. The thermal stability calculation follows the principle of energy conservation and specifically includes the following processes: calculating the Joule heat Q based on the current I, Q = I²×R t, where R is the resistance and t is the unit time; the conductor temperature rise ΔT = Q / (m×c), where m is the conductor mass and c is the conductor specific heat capacity; the temperature rise causes the conductor expansion ΔL = α×L×ΔT, where α is the coefficient of thermal expansion and L is the initial length; the conductor sag is determined by ΔL and the tension change, and the system then derives a new sag value based on the equilibrium equation.

[0023] Once the model is built, the system compares the simulation results with the actual measured values. The comparison and analysis module employs a multi-dimensional index system, such as the similarity of tension change curves (using the Dynamic Time Warping (DTW) algorithm), the difference in vibration response spectrum (extracting frequency domain features through Fast Fourier Transform (FFT), and the deviation in the spatial distribution of thermal stability (evaluated through pixel differences in the thermal distribution map), to perform numerical comparisons and quantify deviations. For example, if the DTW distance between the actual tension change trend and the simulation output is less than a set threshold of 0.1, it is considered that the trend is consistent; if it exceeds this threshold, it is marked as a deviation that needs optimization.

[0024] Based on the comparative analysis results, the dynamic optimization module initiates an optimization strategy, automatically fine-tuning key parameters in the model using genetic algorithms or gradient descent methods. This includes adjusting conductor resistivity, initial tension values, or external wind speed influence factors, and then rerunning the simulation to converge to the optimal model state. This process is executed in a closed loop, with the results compared to actual values ​​after each optimization round until the error converges to an acceptable range (e.g., the error decreases to within 5%).

[0025] The core of the entire system lies in achieving a high degree of fit and prediction of the operating status of transmission lines through a continuous data acquisition-fusion-modeling-analysis-optimization cycle, so that the simulation model always maintains consistency with the actual operating conditions and is forward-looking.

[0026] The data acquisition module includes a meteorological sensor, a conductor status sensor, a tower tilt angle monitor, and an infrared image acquisition device, which are used to collect data on wind speed, temperature, humidity, rainfall intensity, conductor temperature, conductor tension, sag curvature change, tower displacement, and thermal anomaly characteristics of the line surface in the environment where the transmission line is located.

[0027] This implementation method integrates multiple types of high-precision sensors to achieve comprehensive and detailed monitoring of the transmission line's operating environment and conductor condition. First, meteorological sensors, including anemometers, temperature and humidity sensors, and rain gauges, are installed on the top or side arm of the tower to collect data on wind speed, air temperature, relative humidity, and rainfall intensity, respectively. The anemometer uses a cup-shaped structure to measure the number of rotations per unit time, which is then converted into a wind speed value, for example, in meters per second. The temperature and humidity sensor measures resistance and capacitance responses, converting them into temperature and humidity values, with a sampling period of every 60 seconds. The rain gauge uses a tipping bucket structure; each tipping represents 0.2 millimeters of rainfall. The system determines the rainfall intensity per unit time, measured in millimeters per hour, by accumulating the number of tipping buckets.

[0028] Conductor condition sensors, deployed near conductor clamps or within vibration dampers, include an infrared thermometer, a tension sensor, and a device for measuring sag. The infrared thermometer calculates the conductor surface temperature based on the infrared radiation energy emitted from the conductor surface using a pre-defined temperature-radiation relationship model. For example, when the radiation energy reaches a certain level, the system identifies the conductor surface temperature as 80 degrees Celsius according to the model. The tension sensor measures minute deformations in the conductor, converts them into electrical signals, and then converts them into tension values, expressed in Newtons. The sag measurement device calculates the actual sag of the conductor over a certain span using a combination of laser ranging and angle observation; for example, it determines the sag value by the height difference and horizontal distance between two measurement points.

[0029] The tower tilt angle monitor integrates devices that measure acceleration and angular velocity, collecting data in three dimensions every 5 seconds. The system calculates the tower tilt angle based on the distribution of gravity. For example, when the horizontal acceleration is relatively large and the vertical acceleration is decreasing, the system calculates the current tilt angle to be approximately 65 degrees. When the calculation shows that this angle exceeds the safety threshold of 10 degrees, the system automatically generates a warning record.

[0030] The infrared image acquisition device periodically acquires images using a fixed camera or a device mounted on a drone. Each image consists of 320 rows and 240 columns of pixels. The system determines the temperature represented by each pixel based on its grayscale value and a pre-calibrated temperature reference curve. For example, a pixel with a grayscale value of 200 corresponds to a temperature of 110 degrees Celsius. The system further identifies areas where the temperature exceeds a warning threshold, such as areas above 80 degrees Celsius, and marks these as thermal anomalies.

[0031] All the above-mentioned data acquisition devices are connected to the edge computing terminal via a local communication module. The edge terminal performs data caching, verification, compression, and unified time processing, and then uploads the data to the central system via network in the form of a data stream. The system organizes and archives various types of data according to a unified time label, forming structured data input, providing a comprehensive and accurate data foundation for subsequent simulation modeling and multidimensional analysis.

[0032] The data fusion processing module includes a time synchronization unit, a data cleaning unit, and a dynamic weight calculation unit. The time synchronization unit is used to calibrate data from multiple different data sources according to a unified timestamp. The data cleaning unit is used to remove missing values ​​and abnormal jump values. The dynamic weight calculation unit dynamically adjusts the fusion weight coefficients based on the source stability, historical volatility, and contextual relevance of each type of data.

[0033] This implementation method, by setting up a time synchronization unit, a data cleaning unit, and a weight dynamic calculation unit, enables the data fusion processing module to have the ability to unify, clean, and integrate multi-source data with adaptive weights, thereby improving the accuracy and stability of the input data of the simulation system.

[0034] First, the time synchronization unit performs time alignment on data from different sources. In actual operation, different sensor devices have inconsistent sampling frequencies; for example, a tension sensor samples every 5 seconds, an infrared image is captured every 30 seconds, and a meteorological sensor updates every 60 seconds. To unify the input format, the system sets a unified time reference sequence, such as constructing a standard time axis with a time step of 1 second. During actual processing, the system performs interpolation and completion processing on each type of data according to this standard axis. For example, if a sensor has data at 1 second, 3 seconds, and 5 seconds, the system can use linear interpolation to estimate the missing data value in the 2nd second based on the numerical trend between 1 second and 3 seconds, and continue to process the gaps to ensure that all data types have valid data items in every second, achieving strict time alignment.

[0035] Next, the data cleaning unit is responsible for identifying and processing missing values ​​and anomalous jumps in the data. Missing value identification checks for gaps in data records along the standard timeline. If a certain type of data is not uploaded at a particular moment, it is recorded as a missing item. The system fills in the missing items using interpolation or replaces them with the moving average of similar data, according to a predefined strategy. For example, if conductor tension data is missing at the 10th second, the system can use the average of the data from the 9th and 11th seconds as the filler data. For anomalous jump identification, the system uses a continuous data window analysis method, setting the window size to 5 time points. If the data at the current time point differs significantly from the previous 4 time points—for example, if a jump suddenly occurs that is more than 5 times larger than the past average—the system identifies it as an anomalous item and replaces it with the median of the preceding and following time periods to ensure data continuity and physical plausibility.

[0036] Finally, the dynamic weight calculation unit dynamically adjusts the weight coefficients of each data source based on multiple indicators to ensure that the fusion result reflects the true physical state as much as possible. Its core processing consists of three dimensions: First, assessing source stability. The system counts the number of successful uploads and packet losses for each type of sensor every 10 minutes. For example, a tension sensor should upload data 120 times in 10 minutes, but actually uploads 115 times, so the stability score is 115 divided by 120. Second, assessing historical volatility. The system counts the magnitude of numerical fluctuations of this type of data within a set time window; the smaller the fluctuation, the higher the score. Third, assessing contextual relevance, i.e., the degree of influence between this type of data and the target output variable. For example, if historical modeling reveals that conductor temperature has a significant impact on thermal stability, the system will assign a higher weight to conductor temperature data.

[0037] The three evaluation dimensions are all uniformly converted into scores ranging from 0 to 1, and then combined according to a set ratio. For example, if the three scores are 0.92, 0.85, and 0.88, the system calculates a weighted average as the final weight coefficient. This weight coefficient then participates in the weighted fusion processing of multi-source data. For example, temperature data from different sources at the same time are weighted and averaged to generate unified input parameters. The final fusion result serves as a key input for subsequent simulation modeling and analysis, improving the model's prediction accuracy, robustness, and environmental adaptability.

[0038] The model building and adjustment module includes a parameter calibration submodule and a structural modeling submodule. The parameter calibration submodule automatically corrects the errors in the initial model input parameters by comparing them with historical operating condition data. The structural modeling submodule generates a virtual simulation space model based on 3D GIS information, tower physical structure data, and line path topology.

[0039] This implementation method divides the model building and adjustment module into a parameter calibration sub-module and a structural modeling sub-module, thereby realizing the collaborative construction and dynamic optimization of the virtual simulation model of the transmission line at both the numerical parameter and spatial geometry levels.

[0040] First, the parameter calibration submodule automatically corrects the initial input parameters of the simulation model. This module first retrieves measured operational data from the historical operational database under similar meteorological conditions, load levels, and temperature conditions to the current line. This data includes actual conductor tension, temperature changes, current flow, and tower sway amplitude. Then, the system substitutes the initial parameters used in the current simulation model (such as conductor resistance, coefficient of thermal expansion, and mass per unit length) into the model, generates simulation output for the corresponding time period, and compares it point-by-point with historical measured data. The difference between the simulated value and the measured value at each time point is recorded as the error. The system accumulates all error values ​​over a period of time and takes the average as the global error level.

[0041] When the average error exceeds a set threshold, such as 5%, the system will automatically enter the parameter correction process. In this process, the system adopts a stepwise approximation strategy, adjusting the initial parameter values ​​one by one. After each adjustment, the simulation process is rerun, and the results are compared with the measured values ​​to update the error value. Through multiple iterations, the system selects the set of parameters that minimizes the error as the input to the new model and updates the current simulation model. This process requires no manual intervention and has autonomous learning and correction capabilities.

[0042] Meanwhile, the structural modeling submodule is used to construct a three-dimensional spatial model of the transmission line. The system first imports terrain and geomorphological data from a 3D geographic information platform, including basic information such as altitude, slope, and surface features. Then, it retrieves the actual tower layout information for the line from the power design database, including the tower's number, latitude and longitude coordinates, altitude, tower type (e.g., straight-line tower, tension tower), tower height, and crossarm structural dimensions. With the support of existing path topology data, the system automatically connects the tower locations sequentially, establishing the spatial relationship of the conductor path.

[0043] During the modeling process, the system calls upon a pre-defined 3D model component library to match the standard physical model and coordinate information of each type of tower. For example, a high-voltage tension tower model contains several component modules, including tower columns, crossarms, and insulator hanging points. These modules are assembled based on the tower's structural parameters and relative spatial positions to generate a complete solid model. The conductor section simulates its natural bending shape based on parameters such as span distance, sag curve, and tension, generating a realistic shape through geometric algorithms, and accurately reproducing the spatial form by considering the line's direction and height differences.

[0044] Ultimately, the system combines all tower and conductor models to form a complete 3D simulation of the power line, supporting structural analysis, operational simulation, and environmental impact assessment on a digital platform. For example, the system can simulate the bending deformation of conductors under temperature rise in a virtual scene, or the dynamic vibration response of towers under wind speed changes, providing maintenance personnel with an intuitive and interactive simulation environment to support decision-making.

[0045] The comparative analysis module employs a multi-feature fusion anomaly identification algorithm, a time series similarity algorithm for tension change trends, a frequency domain energy spectrum analysis method for conductor vibration response, and a distribution uniformity analysis and regional temperature rise gradient comparison for thermal stability distribution. It comprehensively evaluates the degree of deviation between the simulation model and the actual line in different dimensions and outputs a quantitative score.

[0046] In this implementation, the comparative analysis module uses a multi-dimensional feature fusion-based anomaly detection mechanism to achieve quantitative matching and difference identification between the simulation model and actual operating data. The system first unifies the time base between the simulation output results and the actual monitoring data, ensuring they are compared using the same time step. The time axis is typically constructed in seconds, for example, with a sampling period of once per second, ensuring a one-to-one correspondence between the simulated value and the measured value at each time point.

[0047] In terms of tension change trend analysis, the system matches and compares the simulated tension time series with historical measured tension data. During the comparison process, the system first re-aligns the two time series along their time axes and uses a point-by-point sliding method to find the maximum overlap. The comparison standard is the degree of consistency between the two curves in terms of peak positions, trough positions, and rates of change. For the tension change curves within each continuous time window, the system calculates the similarity score between the two curves segment by segment, in 10-second groups. If the direction of change and fluctuation amplitude of the simulated curve within a window are close to the measured curve, the system records it as a high similarity segment, and the score for this segment is set above 90 points; if there is a significant deviation, such as the simulated curve lagging by more than 5 seconds, or the fluctuation direction being opposite, the score for this segment drops to below 60 points. Finally, the average of the scores for multiple time windows is taken as the tension trend comparison score.

[0048] In conductor vibration response analysis, the system converts the acceleration or minute displacement data of each conductor into an energy distribution map using a frequency analysis algorithm. This map displays the energy distribution characteristics across different frequency bands. The system extracts the dominant frequency positions where energy concentration is most significant in both the simulated and measured conductors and compares the frequency difference between them. For example, if the simulated dominant frequency is 4 Hz and the measured dominant frequency is 3.8 Hz, the frequency difference is 0.2 Hz. The system sets a scoring range based on the size of the difference. A difference of less than 0.5 Hz is considered highly consistent, and the score for this item is above 90 points. If the difference is greater than 1 Hz, the simulation model is considered poorly fitted in that frequency band, and the score drops below 70 points. Simultaneously, the system also compares the frequency energy distribution width to determine if there is any energy spillover or improper concentration.

[0049] In terms of thermal stability distribution analysis, the system divides the temperature distribution of each conductor segment in the simulation model into regions and compares it spatially with the temperature distribution map in the thermal infrared image. The system divides the line into several segments, each 50 meters long, and calculates the average temperature value within each segment. Subsequently, the system compares the temperature differences between adjacent regions to form a regional temperature rise gradient map. For example, if the temperature of a conductor segment is 80 degrees Celsius and the adjacent region is 72 degrees Celsius, the system records the temperature rise gradient of that segment as 8 degrees Celsius and compares this value with the predicted value in the simulation model. If the deviation between the measured gradient and the simulated gradient is less than 5 degrees Celsius, the system considers the region to be a good match and scores 95 points; if it exceeds 10 degrees Celsius, it is considered to have an error, and the score drops to below 60 points.

[0050] The system aggregates the scores from the three analytical dimensions mentioned above and assigns weight coefficients to each dimension. For example, the tension trend score has a weight of 40%, the vibration response score has a weight of 30%, and the thermal stability score has a weight of 30%. By multiplying each score by its corresponding weight and then summing the weighted results, a comprehensive score for the simulation model is finally formed. For example, in one comparison, the tension trend score was 85, the vibration response score was 78, and the thermal stability score was 90. After weighted calculation, the comprehensive score was 83.3. The system uses this score as a quantitative expression of the simulation model's fitting effect in the current round and is used to determine whether the model needs further optimization.

[0051] The dynamic optimization module includes a model error inversion submodule and a simulation strategy adaptation submodule. The model error inversion submodule analyzes the key input parameters that generate errors based on the comparative analysis results and corrects them. The simulation strategy adaptation submodule automatically selects different optimization strategies, such as linear fine-tuning optimization, local reconstruction, or overall reconstruction, according to different deviation levels.

[0052] In this embodiment, the dynamic optimization module constructs a dynamic optimization mechanism for the simulation model by setting up a model error inversion submodule and a simulation strategy adaptation submodule, which automatically switches from error identification to response strategy, effectively ensuring simulation accuracy and operational stability.

[0053] The model error inversion submodule first receives deviation information between the simulation results and actual monitoring data from the comparative analysis module. This deviation information includes multiple dimensions such as the maximum deviation value of the tension curve, the offset value of the conductor's main oscillation frequency, and the temperature difference value of the conductor's temperature rise hotspot region. The system classifies and processes this deviation data and records its corresponding simulation model input parameters. For example, tension deviation corresponds to the initial tension value, frequency offset corresponds to structural stiffness or wind load coefficient, and temperature difference may be related to resistivity, heat capacity, or heat dissipation parameters.

[0054] Subsequently, the system prioritizes the adjustment of each input parameter according to a preset sensitivity scoring mechanism. This process relies on historical modeling data and sensitivity evaluation rules. For example, in the past 100 model runs, it was found that adjusting the initial tension value had the greatest impact on the accuracy of the tension curve, scoring 0.95, while adjusting the ambient temperature only produced an accuracy improvement of 0.15. Therefore, the system prioritizes adjusting the initial tension value.

[0055] The system iteratively corrects sensitive parameters in small steps. For example, when the system identifies a tension prediction error of 12%, it will tentatively modify the initial tension setting by increasing or decreasing it by 3% in each step. After each adjustment, a new round of simulation is immediately run, and the difference between the output result and the measured value is re-acquired. By recording the average deviation after each adjustment, the system can quickly determine whether the parameter adjustment direction is correct, and continue adjusting or revert to a better point accordingly. Typically, if the average error decreases by more than 5% in three consecutive rounds of adjustment, the optimization direction is considered effective, and the process continues until the error is below 3%.

[0056] The simulation strategy adaptation submodule is used to automatically match and optimize the scheme based on the current error level. The system sets three deviation levels: the first level is low, with a deviation of less than 5%; the second level is medium, with a deviation between 5% and 15%; and the third level is high, with a deviation exceeding 15%. After each simulation, the system calculates a comprehensive score, converting each error into a percentage score and then summing them according to their weights. For example, if tension accounts for 40%, vibration accounts for 30%, and thermal stability accounts for 30%, and the scores for the three indicators are 90, 85, and 60 respectively, the weighted comprehensive score is 79 points, corresponding to a deviation of approximately 21%, which the system classifies as a high-level error.

[0057] For Category I low-level errors, the system executes a linear fine-tuning optimization strategy, making minor adjustments only to parameters with the highest sensitivity scores within a range of 1% to 5%, without changing the model structure. For medium-level errors, the system selects the local area with the largest error and reconstructs the structural model of that area. For example, if the sag simulation result of a certain section of conductor deviates by 12%, the system will only reconstruct the tower location, conductor type, and environmental boundary settings for that section. For high-level errors, the system clears all simulation parameters and restarts the modeling process, starting from loading 3D terrain data, reconstructing all tower structures, conductor parameters, and boundary conditions to ensure that the simulation is restored to a controllable state from the source.

[0058] After each dynamic optimization process is completed, the optimization path and the final error change curve will be recorded and stored in the database to form a model learning closed loop, providing a basis for the initial parameter setting under similar working conditions in the future, thereby improving the initialization accuracy of the next round of simulation.

[0059] The comparative analysis module supports historical trend alignment, which can extract the characteristic curves of tension and temperature rise changes in different time periods and compare them with historical typical working condition curves to help identify potential structural aging or environmental stress accumulation.

[0060] In this embodiment, the comparative analysis module integrates the historical trend alignment function to realize the whole process modeling of the trend identification of tension and temperature rise data changes of transmission lines in different time periods, historical operating condition comparison and aging early warning analysis.

[0061] The system first receives real-time tension and conductor temperature data from the data acquisition module and then periodically segments the data over time. Common time periods include daily, 7-day, or 30-day periods. For example, the system divides the tension data of a certain line over the past 30 days into 30 segments, each covering 24 hours, while maintaining the original time series structure. The system generates a characteristic curve for each data segment, recording key change characteristics such as maximum and minimum values, fluctuation amplitude, and duration of continuous high values.

[0062] Subsequently, the system extracts typical operating samples from the historical operating condition database that are similar to the current line environmental conditions. The selection criteria include: the same geographical location, similar season, temperature range difference not exceeding 5 degrees Celsius, and load current level difference not exceeding 10%. The extracted historical sample curves are also segmented and feature extracted at the same time granularity as a reference.

[0063] To achieve trend comparison, the system performs unified processing on current cycle data and historical sample data, including data standardization. The absolute values ​​of tension and temperature rise are mapped to a range of 0 to 1 using minimum and maximum values ​​to eliminate the influence of numerical dimensions on curve shape comparison. After standardization, the system aligns the characteristic curves of the current and historical samples segment by segment, identifying similarities and differences in the changing trends of the two curves. For example, it determines whether peaks appear earlier, troughs are delayed, and whether there are abnormal plateau periods.

[0064] For example, in tension comparison, if the conductor tension remains 15% higher than the historical average for three consecutive hours on a given day, accompanied by a slowing rate of change, the system defines this phenomenon as an "abnormal high tension plateau." Similarly, if the temperature rise curve consistently shows a peak at the same time each day and rises daily, the system identifies it as a "slowly increasing temperature rise trend zone." The system sets multiple rules for each type of trend deviation; for example, a deviation lasting more than 2 hours or an accumulated abnormal time exceeding 6 hours within a day is recorded as a "suspicious trend segment."

[0065] The identified abnormal trends will be compared with the structural aging early warning model. This model includes multiple sets of reference conditions. For example, if a section of conductor experiences a high tension plateau for five consecutive days, and the plateau period increases by an average of no less than 20 minutes per day, it is preliminarily determined that there is a risk of structural relaxation or tension release failure in that area. If an abnormal temperature rise trend occurs for three consecutive days, and the local temperature is more than 5 degrees Celsius higher than the upper limit of the normal range, the system will determine that there is thermal stress accumulation or deterioration of cooling conditions.

[0066] Finally, the system matches the trend change identification results with historical deviation levels, outputting a trend change report and risk level score. The system provides a visual graph, showing a comparison view of tension and temperature rise curves between the current period and historical periods, helping maintenance personnel identify areas of long-term performance degradation that may be caused by structural aging, wire fatigue, or local environmental influences, and can further assist in scheduling maintenance plans or optimizing load distribution strategies.

[0067] The data fusion processing module has a spatiotemporal consistency verification function, which performs structural similarity analysis on the data of adjacent towers or line segments in space, and makes a continuity judgment on the data fluctuation curve in time.

[0068] In this embodiment, the data fusion processing module sets up a spatiotemporal consistency verification function to make a reasonable judgment on the spatial distribution and temporal evolution of different data sources, so as to ensure the logical consistency and dynamic continuity of the input data of the simulation model.

[0069] In terms of spatial consistency verification, the system first constructs the spatial topology of the entire line based on the tower coordinate information stored in the 3D geographic information system. The system calculates the straight-line distance and relative height difference between each pair of adjacent towers. If the horizontal distance between them is less than 500 meters and the height difference is less than 10 meters, the two lines are considered similar in structure and environment and should have similar operational status data. The system compares the tension value, conductor temperature, sag, and other key parameters collected at the same time for each pair of adjacent line segments. If the tension value of a certain conductor segment is more than 20% higher than that of the adjacent segment, the system marks the data as structurally inconsistent and pushes it to the data cleaning module for further analysis.

[0070] For example, the system records that the tension of a certain section of conductor is 1800 Newtons, while the tension of a nearby section is 1450 Newtons, a difference of 350 Newtons, which accounts for approximately 24% of the lower tension values. If the difference exceeds the set 20% similarity threshold, the system determines that the data may be abnormal, such as sensor measurement inaccuracy or incorrect tension input, and prompts for investigation or correction.

[0071] For time consistency verification, the system constructs a continuous data window for each type of time series data, such as conductor temperature or tension, according to a set time step, with a commonly used window length of 10 minutes. Within each time window, the system extracts the data fluctuation amplitude, rate of change, and direction of change. For example, if the conductor temperature rises from 60 degrees Celsius to 68 degrees Celsius within 10 minutes, the system records it as a continuous upward trend. If, in the subsequent 11th minute, the data suddenly drops to 52 degrees Celsius, a decrease of 16 degrees Celsius, the system determines that this jump exceeds the allowable fluctuation limit of 10 degrees Celsius and marks it as a time anomaly.

[0072] For segments with poor continuity, the system further checks whether the contextual data supports abrupt changes. For example, if the wind speed does not change drastically between adjacent time points, the current load does not fluctuate significantly, but the temperature suddenly drops, the system considers the temperature data to be of low reliability. According to the system configuration rules, abnormal data can be removed or replaced with the average value of adjacent time points to ensure that the entire time curve is smooth and continuous, and to avoid outliers affecting the stability of the simulation input.

[0073] After completing consistency checks in both space and time, the system labels the verified data as "trustworthy" and uses it as high-priority input for the model building module, ensuring that the simulation parameters have a sound physical basis and a continuous logical structure. Simultaneously, data identified as anomalous and its source are logged for data quality assessment and subsequent optimization of sensor maintenance strategies.

[0074] The visualization module integrates multi-level information view switching functionality, supporting quick switching from the panoramic view of the transmission line to a local view of a single tower or line segment, and dynamically annotating model deviation heatmaps, risk level assessment maps, and operation and maintenance suggestion prompts.

[0075] In this embodiment, the visualization module has a multi-level information view switching function, which allows users to quickly view the status information of transmission lines in different spatial ranges. It also uses intuitive graphic annotations to visualize model errors, risk levels and operation and maintenance suggestions, thereby providing effective support for line operation monitoring and decision-making.

[0076] First, the system constructs a panoramic view using line spatial modeling data. This view is based on a three-dimensional geometric model of the entire transmission line, integrating tower coordinates, conductor paths, elevation data, and surrounding environmental information. By default, users can see the entire transmission corridor upon entering the system. The main interface displays the location of all towers and line segments, overlaid with color-coded operational status overviews, such as whether conductor tension exceeds limits or temperature is abnormal. All status indicators are extracted from the real-time simulation model and collected data.

[0077] When a user clicks on a tower or a specific line segment in panoramic view, the system immediately switches to a partial view. This process is handled by the visualization engine, which locates the target node based on its 3D coordinates and renders the corresponding model unit at a magnified scale (e.g., increasing the original magnification to 10x). The system then reloads the high-precision model and detailed attribute information of the local structure. The partial view displays the tension curve, temperature rise graph, and vibration response curve of that line segment over the past 30 minutes, along with graphical annotations comparing simulation and measured data.

[0078] Regarding the model deviation heatmap, the system numerically calculates the simulation errors of tension, temperature, or vibration at each node and sets error level ranges. For example, errors less than 5% are Level 1, 5% to 15% are Level 2, and errors exceeding 15% are Level 3. The system maps the error of each calculation node to a color label, for example, green for Level 1, yellow for Level 2, and red for Level 3. These colors are then overlaid onto the surface of the graphic model of the tower or line segment, forming a visual heatmap. Users can easily identify areas of concentrated deviation through the color distribution.

[0079] The risk level assessment chart is based on a multi-indicator scoring logic preset in the system. For each section of the line, scores are calculated separately for factors such as the duration of tension exceeding the limit, the rate of temperature rise, and the vibration frequency deviation. For example, if the tension exceeds the upper limit for 30 minutes, 5 points are awarded for every 10 minutes, resulting in a total score of 15 points; a temperature rise rate exceeding the reference value by 50% is awarded 10 points; and a vibration frequency deviation of 0.8 Hz is awarded 20 points. All indicator scores are summed to obtain the total score, for example, 45 points. The system divides the score range into three levels: below 60 is high risk, 60 to 80 is medium risk, and above 80 is low risk, represented in red, orange, and green on the chart, respectively.

[0080] The maintenance suggestion dialog box pops up in the upper right corner of the partial view based on the above risk level and indicator performance. The dialog box contains specific anomaly descriptions and maintenance suggestions. For example, if the tension score is 15 points, the temperature rise score is 10 points, and the vibration score is 20 points, with a total score of 45 points, the system automatically identifies it as a high-risk level and displays the message "The current conductor tension is too high. It is recommended to check the tightening condition on-site and evaluate the tension compensation plan." If it is a medium-risk or low-risk level, the suggestion will recommend "increasing the monitoring frequency" or "regularly reviewing the data stability."

[0081] Throughout the operation, users can return to the panoramic view at any time via mouse operation or the menu bar, or switch target towers by clicking on the side selection list. The system supports switching local scenes without reloading the main model and ensures that the heat map, risk map, and prompt box are automatically updated after each switch, maintaining the timeliness and consistency of the data.

[0082] The dynamic optimization module includes a confidence interval rollback mechanism. When the simulation results fluctuate beyond the statistical confidence interval, it will automatically roll back to the previous stable model version.

[0083] In this embodiment, the dynamic optimization module introduces a confidence interval backoff mechanism to statistically control the fluctuations of the output results during the simulation model iteration process. When abnormal fluctuations exceeding the normal range are identified, the module automatically restores the previous stable model state, thereby ensuring the reliability and continuity of the simulation process.

[0084] During model optimization, the system generates a set of key output metrics for each iteration, such as predicted tension values, conductor temperature, and thermal stability assessment results, and records these results as a snapshot of the model version's output. The system is configured to retain data from at least the 10 most recent stable versions as a statistical reference sample. For example, in the tension dimension, the system organizes the tension output data from the 10 versions into a sequence, with a maximum value of 1850 Newtons, a minimum value of 1750 Newtons, and an average value of 1800 Newtons. Depending on the system configuration, if a confidence margin of ±5% is set, the tension confidence interval is set to 1700 to 1900 Newtons.

[0085] After each optimization, the system performs statistical verification on the simulation output of this round, comparing each indicator to see if it falls within the confidence interval. For example, if the tension simulation output of this round is 1950 Newtons, which is higher than the upper limit of 1900 Newtons, and the temperature output is 85 degrees Celsius, while the upper limit of the temperature confidence interval is 82 degrees Celsius, both key indicators exceed the confidence interval, and the system determines that the current model is an abnormal fluctuation model.

[0086] When the system detects that any key indicator exceeds its corresponding upper or lower confidence limit, it triggers a rollback mechanism. At this point, the system first suspends further transmission and updates of the current model parameters, switching the simulation state back to the previous model version that did not exceed the confidence range, including its parameter settings, input conditions, boundary constraints, etc. The system marks this abnormal model as an invalid version and archives the deviation values ​​and outlier information of this version for subsequent anomaly analysis and model evaluation.

[0087] After the rollback is completed, the system restarts the optimization process and performs fine-tuning again based on the previous stable model. To avoid frequent rollbacks caused by continuous fluctuations, the system records the number of recent consecutive rollbacks. For example, if three rollbacks occur within a short period of time, the system automatically reduces the parameter adjustment step size, such as from 5% to 2%, or switches to a more conservative optimization algorithm, extending the simulation time window used in each round of optimization to improve stability.

[0088] The entire confidence interval backoff mechanism is built based on the long-term statistical characteristics of the simulation model. It has the ability to adapt and prevent the spread of anomalies, and is particularly suitable for scenarios with highly sensitive parameter control, such as conductor tension, structural temperature rise or long span sag simulation.

[0089] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A transmission line operation and maintenance simulation system based on digital twin technology, characterized in that, The system includes: The data acquisition module is used to collect real-time operating status data of the transmission line and establish a data link. The data includes meteorological information, current carrying capacity, temperature, tension, conductor sag and other multi-source heterogeneous operating status data. The model building and adjustment module is used to build or adjust the parameters of the virtual simulation model based on real-time running status data. The comparative analysis module is used to perform multi-dimensional comparative analysis between the updated virtual simulation model and the actual transmission line. The comparative dimensions include tension change trend, conductor vibration response and thermal stability distribution. The dynamic optimization module is used to dynamically optimize and iteratively update the virtual simulation model based on the results of comparative analysis. The data fusion processing module communicates with the model building and adjustment module. It is used to preprocess multi-source operating status data and adopts an adaptive weight fusion algorithm to integrate different data sources to generate model input parameters, thereby improving the update accuracy and stability of the simulation model.

2. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The data acquisition module includes a meteorological sensor, a conductor status sensor, a tower tilt angle monitor, and an infrared image acquisition device, which are used to collect data on wind speed, temperature, humidity, rainfall intensity, conductor temperature, conductor tension, sag curvature change, tower displacement, and thermal anomaly characteristics of the line surface in the environment where the transmission line is located.

3. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The data fusion processing module includes a time synchronization unit, a data cleaning unit, and a weight dynamic calculation unit. The time synchronization unit is used to calibrate data from multiple different data sources according to a unified timestamp. The data cleaning unit is used to remove missing values ​​and abnormal jump values. The weight dynamic calculation unit dynamically adjusts the fusion weight coefficients based on the source stability, historical volatility, and contextual relevance of each type of data.

4. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The model building and adjustment module includes a parameter calibration submodule and a structural modeling submodule. The parameter calibration submodule automatically corrects the errors of the initial model input parameters by comparing them with historical operating condition data. The structural modeling submodule generates a virtual simulation space model based on 3D GIS information, tower physical structure data, and line path topology.

5. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The comparative analysis module employs a multi-feature fusion anomaly identification algorithm, a time series similarity algorithm for tension change trends, a frequency domain energy spectrum analysis method for conductor vibration response, and a distribution uniformity analysis and regional temperature rise gradient comparison for thermal stability distribution. It comprehensively evaluates the degree of deviation between the simulation model and the actual line in different dimensions and outputs a quantitative score.

6. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The dynamic optimization module includes a model error inversion submodule and a simulation strategy adaptation submodule. The model error inversion submodule analyzes the key input parameters that generate errors based on the comparative analysis results and corrects them. The simulation strategy adaptation submodule automatically selects different optimization strategies, such as linear fine-tuning optimization, local reconstruction, or overall reconstruction, based on different deviation levels.

7. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The comparative analysis module supports historical trend alignment, which can extract the characteristic curves of tension and temperature rise changes in different time periods and compare them with historical typical working condition curves to help identify potential structural aging or environmental stress accumulation.

8. The transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The data fusion processing module has a spatiotemporal consistency verification function, which performs structural similarity analysis on the data of adjacent towers or line segments in space, and makes a continuity judgment on the data fluctuation curve in time.

9. A transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The visualization module integrates a multi-level information view switching function, which supports quick switching from the panoramic view of the transmission line to a local view of a single tower or line segment, and dynamically annotates the model deviation heat map, risk level assessment map and operation and maintenance suggestion prompt box.

10. A transmission line operation and maintenance simulation system based on digital twin technology according to claim 1, characterized in that: The dynamic optimization module includes a confidence interval rollback mechanism. When the simulation results fluctuate beyond the statistical confidence interval, it will automatically roll back to the previous stable model version.

Images