Distribution network lightning monitoring method and system

Abstract

The present application provides a distribution network lightning monitoring method and system, which relates to the field of information technology, including: obtaining dynamic change data of environmental factors through real-time monitoring of the ambient temperature and humidity along the distribution line, and using it as the input of the mathematical model of the insulator performance parameters to obtain the final mathematical model of the insulator performance parameters that can adapt to changes in environmental conditions; obtaining and combining the insulator manufacturing database, determining the specific manufacturing process parameters of the insulator according to the optimal combination of insulator materials and structural parameters that meet the actual needs of the distribution line, producing and putting the distribution line into operation; if it is found that the insulator has a hidden fault, according to the location and model of the faulty insulator, matching the corresponding replacement product from the insulator manufacturing database, and triggering the instruction of insulator replacement to ensure the safe and stable operation of the distribution line. The precise optimization of insulator performance and fault prediction are achieved, and the lightning protection capability and operation reliability of the distribution line are improved.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a distribution network lightning monitoring method and system. Background Art

[0002] The selection of the model and quantity of insulators for distribution lines is a complex technical issue. In actual distribution lines, insulators are subjected to the impact of lightning currents, and the steepness of their rising edges directly affects the insulation performance and mechanical strength of the insulators. The steepness of the rising edge of lightning currents is affected by many factors, including the type of lightning discharge, the charge distribution of thunderclouds, and the distance between the lightning strike point and the distribution line. The complexity and variability of these factors pose challenges to the accurate measurement and analysis of the steepness of the rising edge of lightning currents. Therefore, how to establish a scientific and reasonable insulator selection method based on comprehensive consideration of multiple factors including the steepness of the rising edge of lightning currents, insulator performance parameters, and the actual operating environment of distribution lines is a technical issue that urgently needs to be studied and resolved. Summary of the invention

[0003] In order to solve the above technical problems, the present invention provides a distribution network lightning monitoring method, which mainly includes:

[0004] Obtain historical lightning current data at the geographical location of the distribution line, analyze the correlation between the steepness of the lightning current rising edge and the thundercloud charge distribution, and obtain the distribution characteristics of the steepness of the lightning current rising edge at the geographical location of the distribution line; determine the insulation performance and mechanical strength requirements that the distribution line insulators need to meet based on the obtained distribution characteristics of the steepness of the lightning current rising edge, and establish a mathematical model of insulator performance parameters in combination with the distance between the lightning strike point and the line, the material of the insulator, and the structural parameters; obtain dynamic change data of environmental factors through real-time monitoring of the ambient temperature and humidity along the distribution line, and use it as the input of the mathematical model of the insulator performance parameters to obtain the final mathematical model of insulator performance parameters that can adapt to changes in environmental conditions; take meeting the steepness requirements of the lightning current rising edge and environmental adaptability as the optimization goal, and take the insulator material and structural parameters as the optimization variables , using a multi-objective optimization method, based on the final insulator performance parameter mathematical model, the optimal combination of insulator materials and structural parameters that meet the actual needs of the distribution line is obtained; the insulator manufacturing database is obtained and combined, and the specific manufacturing process parameters of the insulator are determined according to the optimal combination of insulator materials and structural parameters that meet the actual needs of the distribution line, and the distribution line is produced and put into operation; after the distribution line is put into operation, the working status data of the insulator is collected in real time through the online monitoring device, and the data is correlated with the distribution characteristics of the steepness of the rising edge of the lightning current, and the use of the insulator is analyzed and predicted to discover potential insulator failure hazards; if the insulator is found to have a failure hazard, the corresponding replacement product is matched from the insulator manufacturing database according to the location and model of the faulty insulator, and the insulator replacement instruction is triggered to ensure the safe and stable operation of the distribution line.

[0005] The present invention provides a distribution network lightning monitoring system, which mainly includes: a lightning data analysis module, which is used to obtain historical lightning current data of the geographical location of the distribution line, analyze the correlation between the steepness of the rising edge of the lightning current and the charge distribution of the thundercloud, and obtain the distribution characteristics of the steepness of the rising edge of the current and the steepness of the rising edge of the lightning discharge current; an insulator performance modeling module, which is used to determine the insulation performance and mechanical strength requirements that the distribution line insulator needs to meet according to the obtained distribution characteristics of the steepness of the rising edge of the lightning current, and establish an insulator performance parameter mathematical model in combination with the distance between the lightning strike point and the line, the material of the insulator, and the structural parameters; an environmental adaptability optimization module, which is used to obtain dynamic change data of environmental factors through real-time monitoring of the ambient temperature and humidity along the distribution line, and use it as the input of the insulator performance parameter mathematical model to obtain the final insulator performance parameter mathematical model that can adapt to changes in environmental conditions;

[0006] The multi-objective optimization calculation module is used to take the steepness of the rising edge of lightning current and environmental adaptability as the optimization objectives, and the insulator material and structural parameters as the optimization variables. The multi-objective optimization method is used to obtain the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line based on the final insulator performance parameter mathematical model; the insulator parameter optimization module obtains and combines the insulator manufacturing database, determines the specific manufacturing process parameters of the insulator according to the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line, and produces and puts the distribution line into operation; the online monitoring and maintenance module is used to collect the working status data of the insulator in real time through the online monitoring device after the distribution line is put into operation, and correlates it with the distribution characteristics of the steepness of the rising edge of lightning current, analyzes and predicts the use of insulators, and discovers potential insulator failure hazards;

[0007] The fault diagnosis and maintenance instruction module is used to match the corresponding replacement product from the insulator manufacturing database according to the location and model of the faulty insulator if a potential fault is found in the insulator, and trigger the instruction to replace the insulator to ensure the safe and stable operation of the distribution line.

[0008] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0009] The present invention discloses a distribution network lightning monitoring method and system. The method obtains the steepness distribution characteristics of the rising edge of the lightning current by analyzing the historical lightning current data of a specific geographical location, and establishes a mathematical model of the insulator performance parameters in combination with the environmental temperature and humidity monitoring data. With the goal of meeting the lightning protection requirements and environmental adaptability, a multi-objective optimization method is used to determine the optimal material and structural parameters of the insulator. The specific process parameters are determined in combination with the manufacturing database and the insulator is produced. After commissioning, potential faults are predicted through online monitoring and analysis, and replacement products are matched from the database. The present invention realizes the precise optimization of insulator performance and fault prediction, improves the lightning protection capability and operational reliability of the distribution line, and provides a strong guarantee for the safe and stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 The present invention is a flow chart of the lightning monitoring method for distribution network. DETAILED DESCRIPTION

[0011] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0012] like Figure 1 The distribution network lightning monitoring method of this embodiment may specifically include:

[0013] S101. Obtain historical lightning current data at the geographical location of the distribution line, analyze the correlation between the steepness of the rising edge of the lightning current and the charge distribution of the thundercloud, and obtain the distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location of the distribution line.

[0014] The historical lightning current data of the geographical location of the distribution line is obtained, and the rising edge steepness characteristics are extracted from the historical lightning current data. The rising edge steepness characteristics are classified by using the K-means clustering algorithm to obtain the preliminary distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location. According to the thundercloud charge distribution data, a thundercloud charge distribution model is established by using the Gaussian mixture model, and the preliminary distribution characteristics of the steepness of the rising edge are corrected by using the thundercloud charge distribution model. The Morlet wavelet transform is used to perform time-frequency analysis on the corrected preliminary distribution characteristics of the steepness of the rising edge to obtain the steepness distribution characteristics at different time scales; a feature vector matrix is ​​constructed based on the steepness distribution characteristics at different time scales, and the feature vector matrix is ​​used to perform dimensionality reduction processing on the historical lightning current data to obtain a lightning current feature data set after dimensionality reduction. A long short-term memory network is constructed based on the lightning current feature data set after dimensionality reduction as a lightning current rising edge steepness prediction model, and the long short-term memory network is used to predict the steepness of the rising edge of the newly collected lightning current data, and the distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location are output.

[0015] For example, historical lightning current data of the geographical location of the distribution line is obtained, and the steepness characteristics of the rising edge are extracted from it. The steepness characteristics are classified by the K-means clustering algorithm to obtain the preliminary distribution characteristics of the steepness of the rising edge of the lightning current in the area. According to the thundercloud charge distribution data, a thundercloud charge distribution model is established using the Gaussian mixture model, and the preliminary distribution characteristics of the steepness of the rising edge are corrected using the model. The Morlet wavelet transform is used to perform time-frequency analysis on the corrected preliminary distribution characteristics of the steepness of the rising edge to obtain the steepness distribution characteristics at different time scales. For the obtained multi-scale steepness distribution characteristics, the principal component analysis method is used to extract the main eigenvectors, construct the eigenvector matrix, select the principal components with a cumulative contribution rate of 95%, and use the eigenvector matrix to reduce the dimensionality of the original lightning current data to obtain the lightning current characteristic data set after dimensionality reduction. The lightning current characteristic data set after dimension reduction is normalized and preprocessed by time series segmentation, and a long short-term memory network LSTM is constructed as a prediction model for the steepness of the rising edge of the lightning current. The trained LSTM model is used to predict the steepness of the rising edge of the newly collected lightning current data. Specifically, the newly collected lightning current data is subjected to the above-mentioned dimension reduction processing, normalization and time series segmentation preprocessing, and then input into the trained LSTM model to output the final distribution characteristics of the steepness of the rising edge of the lightning current in the area. Using the measurement data of the three-dimensional electric field instrument, a three-dimensional model of the thundercloud charge distribution is constructed, and the three-dimensional model is correlated with the distribution characteristics of the steepness of the rising edge of the lightning current to obtain the relationship between the thundercloud charge distribution and the steepness of the rising edge of the lightning current. The historical lightning current data of the geographical location of the distribution line is obtained, for example, lightning current data of a certain area in the past five years are collected, including timestamps, current peaks and rise times. The steepness characteristics of the rising edge are extracted from these data, and the calculation formula is steepness = current peak / rise time. For each lightning strike event, its steepness value and the corresponding lightning discharge type, such as cloud-to-cloud discharge or cloud-to-ground discharge, are recorded. The K-means clustering algorithm is used to classify the steepness characteristics. The number of clusters is set to 3, corresponding to the three steepness levels of low, medium and high. The iterative calculation is performed until the cluster center is stable. According to the clustering results, the preliminary distribution characteristics of the steepness of the rising edge of the lightning current in the area are obtained. The thundercloud charge distribution model is established using the Gaussian mixture model. It is assumed that there are three main charge regions in the thundercloud, and each charge region is represented by a Gaussian distribution. The model parameters, including the mean, covariance matrix and weight of each Gaussian distribution, are estimated by the expectation maximization algorithm. The established thundercloud charge distribution model is used to correct the steepness characteristics of the rising edge, considering the influence of charge density on the steepness. The Morlet wavelet transform is used to perform time-frequency analysis on the corrected lightning current rising edge, and different scale parameters such as 1, 2, 4, 8, and 16 are selected for transformation to obtain the steepness distribution characteristics at different time scales. The principal component analysis method is used to extract the main feature vectors of the obtained multi-scale steepness distribution characteristics.Calculate the eigenvalues ​​and eigenvectors, select the principal components with a cumulative contribution rate of 95%, and construct the eigenvector matrix. Use the eigenvector matrix to reduce the dimension of the original lightning current data, map the high-dimensional data to the low-dimensional space, and obtain the lightning current characteristic data set after dimension reduction. Normalize the reduced-dimensional data set and scale each eigenvalue to the interval [0,1]. Perform time series segmentation, set the time window size to 10, the step size to 1, and generate a data format suitable for LSTM model input. Construct an LSTM network, which includes a LSTM layer with 64 units and a fully connected output layer. Use the Adam optimizer and the mean square error loss function to train the model on the training set, and the validation set is used to adjust the hyperparameters. Use the trained LSTM model to predict the steepness of the rising edge of the newly collected lightning current data, and output the final distribution characteristics of the steepness of the rising edge of the lightning current in the area. Finally, use a three-dimensional electric field meter to measure the electric field strength at different heights and horizontal positions to construct a three-dimensional model of thundercloud charge distribution. Discretize the three-dimensional model into grid points, and record the charge density value at each grid point. The correlation coefficient between the charge density at the grid point and the steepness of the rising edge of the lightning current at the corresponding position is calculated, and the relationship between the two is analyzed to obtain the quantitative correlation between the charge distribution in the thundercloud and the steepness of the rising edge of the lightning current.

[0016] S102. Determine the insulation performance and mechanical strength requirements that the distribution line insulators need to meet based on the obtained distribution characteristics of the steepness of the rising edge of the lightning current, and establish a mathematical model of the insulator performance parameters in combination with the distance between the lightning strike point and the line, the material of the insulator, and the structural parameters.

[0017] According to the steepness distribution characteristics of the rising edge of lightning current, a function of the relationship between the withstand voltage level and the steepness of the insulator is established, and the support vector regression method is used to fit the data points. The nonlinear relationship is processed by the radial basis function to obtain the insulation performance requirement function that the insulator needs to meet; according to the insulation performance requirement function, a mechanical strength demand model is constructed through the correlation between the steepness of the rising edge of lightning current, the rise time and the peak current and the mechanical impact force, and the nonlinear least squares method is used to fit the historical data to obtain the mechanical strength requirement function that the insulator needs to meet; according to the distance between the lightning strike point and the line, the material and structural parameters of the insulator, an insulator performance parameter influencing factor matrix is ​​established, and the multivariate linear regression method is used to establish the direct relationship between the influencing factors and the performance parameters, and the insulator performance parameters are constructed. and the relationship function between the insulator performance parameters and the influencing factors; using a multilayer perceptron neural network, taking the insulation performance requirement function, the mechanical strength requirement function and the relationship function between the performance parameters and the influencing factors as input, after data preprocessing, standardization and normalization, setting the number of hidden layer nodes to twice that of the input layer, using the ReLU activation function, using the Adam optimizer and the mean square error loss function, adjusting the hyperparameters through cross-validation, and training the multilayer perceptron neural network to obtain the insulator performance parameter mathematical model; adding a safety factor calculation node to the output layer, if the key performance parameter value output by the insulator performance parameter mathematical model is less than the preset performance parameter threshold, adjusting the weight of the safety factor calculation node, and retraining the insulator performance parameter mathematical model.

[0018] For example, according to the steepness distribution characteristics of the rising edge of lightning current, the relationship function between the withstand voltage level and the steepness of the insulator is established, the support vector regression method is used to fit the data points, and the nonlinear relationship is processed by the radial basis function to obtain the insulation performance requirement function that the insulator needs to meet. Through the correlation between the steepness of the rising edge of lightning current, the rise time and the peak current and the mechanical impact force, a mechanical strength demand model is constructed, and the nonlinear least squares method is used to fit the historical data. Considering the influence of the lightning current waveform characteristics on the mechanical impact force, the mechanical strength requirement function that the insulator needs to meet is obtained. According to the distance between the lightning strike point and the line, the material and structural parameters of the insulator, the insulator performance parameter influencing factor matrix is ​​established, and the multivariate linear regression method is used to establish the direct relationship between the influencing factors and the performance parameters, and the relationship function between the insulator performance parameters and the influencing factors is constructed. Using a multi-layer perceptron neural network, the insulation performance requirement function, mechanical strength requirement function and performance parameter influencing factor relationship function are used as input. After data preprocessing, standardization and normalization, the number of hidden layer nodes is set to twice the input layer. The ReLU activation function is used, the Adam optimizer and the mean square error loss function are adopted, and the hyperparameters are adjusted through cross-validation. The mathematical model of insulator performance parameters is trained to output the key performance parameter values ​​of the insulator. Safety factor calculation nodes are added to the output layer to consider the performance margin under extreme conditions to ensure the reliability of insulators under various working conditions. According to the historical lightning current data of the geographical location of the distribution line, the steepness characteristics of the rising edge are extracted, such as 1000 lightning strikes recorded in a certain area within five years, with a steepness range of 10-100kA / μs. The support vector regression method is used to establish the relationship function between the insulator withstand voltage level and steepness. The radial basis function kernel function is selected, the kernel parameter γ is set to 0.1, the penalty parameter C is 100, and the hyperparameters are optimized through grid search. The fitted function shows that when the steepness is 50kA / μs, the required withstand voltage level of the insulator is 150kV. At the same time, considering the characteristics of the lightning current waveform, such as 1-10μs rise time and 10-100kA peak current, a mechanical strength requirement model is constructed. The nonlinear least squares method is used to fit the historical data to obtain the mechanical strength requirement function, which shows that when the steepness is 50kA / μs, the rise time is 5μs, and the peak current is 50kA, the required mechanical strength of the insulator is 70kN. For the factors affecting the performance of the insulator, such as the distance between the lightning strike point and the line, the insulator material such as porcelain, glass, composite materials and structural parameters, an influencing factor matrix is ​​established. The multivariate linear regression method is used to obtain the relationship function between the performance parameters and the influencing factors, such as the breakdown voltage of the insulator = 150 + 0.05 × distance + 10 × material coefficient + 2 × number of sheds + 0.1 × stretching length. Finally, a multi-layer perceptron neural network is constructed. The number of nodes in the input layer is 10, the hidden layer is set to two layers, the number of nodes in each layer is 20 and 10 respectively, and the ReLU activation function is used.The Adam optimizer was used, with a learning rate of 0.001, a batch size of 32, and 1000 training rounds. The hyperparameters were adjusted through 5-fold cross validation, and the mean square error of the final model on the test set was 0.05. A safety factor calculation node was added to the output layer to consider 1.5 times the extreme case to ensure the performance margin of the insulator. The model output shows that under the given conditions, the optimal performance parameters of the insulator are a withstand voltage level of 180kV, a mechanical strength of 85kN, and a safety factor of 1.2.

[0019] According to the steepness distribution characteristics of the rising edge of lightning current and the distance between the lightning strike point and the line, the mechanical stress of the insulator under the mechanical impact force of lightning current is analyzed. Through theoretical calculation and finite element analysis, the mechanical strength requirements that the distribution line insulators need to meet are determined, including bending strength and compressive strength.

[0020] According to the distribution characteristics of the steepness of the rising edge of the lightning current and the distance between the lightning strike point and the line, a mathematical model of the lightning current acting on the insulator is established, and the distribution of the lightning current on the insulator is calculated by the finite difference time domain method to obtain the current density distribution function on the surface of the insulator. Using the current density distribution function, the electromagnetic force generated by the lightning current on the insulator is calculated by the Maxwell stress tensor method, and the electromagnetic force is input as a load to the insulator force analysis model. The insulator force analysis model is established, and the stress distribution inside the insulator is calculated by the finite element method. Using finite element analysis software, a three-dimensional model of the insulator is established, material properties and boundary conditions are set, and the stress distribution is used as a load input to perform static analysis and transient dynamic analysis to obtain the stress and deformation data of each part of the insulator under static and dynamic loads. According to the results of finite element analysis, the maximum stress value of the key parts of the insulator is extracted, the bending performance is judged by the vonMises stress criterion, and the compressive performance is judged by the principal stress analysis. If the maximum stress value exceeds the preset stress threshold, the safety factor is determined in combination with the reliability design method based on statistical analysis, the bending strength and compressive strength requirements that the insulator needs to meet are determined, and the design parameters of the mechanical strength requirements of the insulator are output.

[0021] For example, according to the steepness distribution characteristics of the rising edge of the lightning current and the distance between the lightning strike point and the line, a mathematical model of the lightning current acting on the insulator is established, and the distribution of the lightning current on the insulator is calculated by the finite difference time domain method. Considering the attenuation effect of the distance from the lightning strike point on the current density, the current density distribution function on the surface of the insulator is obtained. Using the current density distribution function, the electromagnetic force generated by the lightning current on the insulator is calculated by the Maxwell stress tensor method. Combined with the geometric structure and material parameters of the insulator, an insulator force analysis model is established, and the stress distribution inside the insulator is calculated by the finite element method. Finite element analysis software is used to establish a three-dimensional model of the insulator, set material properties and boundary conditions, use the calculated stress distribution as load input, perform static analysis and transient dynamic analysis, and obtain stress and deformation data of various parts of the insulator under static and dynamic loads. According to the results of finite element analysis, the maximum stress value of the key parts of the insulator is extracted, the bending performance is judged by the vonMises stress criterion, the compressive performance is judged by the principal stress analysis, and the safety factor is determined by combining the reliability design method based on statistical analysis. The bending strength and compressive strength requirements that the insulator needs to meet are determined, and the mechanical strength design parameters of the insulator are output. For the mechanical strength analysis of the insulator of the distribution line, the steepness distribution model of the rising edge of the lightning current is first established based on the historical lightning data. For example, the steepness range in a certain area is 10-100kA / μs, and the average value is 50kA / μs. Considering the attenuation effect of the distance between the lightning strike point and the line on the current density, the finite difference time domain method is used for electromagnetic field calculation. The grid size is set to 1mm, the time step is 0.1ns, and the lightning process of 30μs is simulated. The calculation results show that when the lightning strike distance is 100m, the maximum current density on the insulator surface is 500A / m². The Maxwell stress tensor method was used to convert the current density into electromagnetic force. The finite element model was established by combining the insulator geometry such as length 300mm, diameter 100mm and material parameters such as elastic modulus 70GPa and Poisson's ratio 0.3. A tetrahedral mesh was used with a cell size of 2mm and a total number of meshes of about 500,000. Static analysis and transient dynamic analysis were performed with a time step of 1μs and an analysis duration of 100μs. The static analysis results showed that the maximum vonMises stress of the insulator was 50MPa, located at the connection of the hardware; the dynamic analysis showed that the stress peak of 65MPa occurred at 10μs. The bending performance was judged by the vonMises stress criterion, and the maximum stress was less than the material yield strength of 80MPa; the principal stress analysis showed that the maximum compressive stress was 30MPa, which was less than the compressive strength of 100MPa. Monte Carlo simulation was used for reliability analysis. Considering the randomness of material strength and load, the confidence level was set to 99.99%, and the safety factor was 1.5. It was finally determined that the bending strength that the insulator needs to meet is 65MPa×1.5=97.5MPa, and the compressive strength is 30MPa×1.5=45MPa.The output mechanical strength design parameters are minimum cross-sectional area 3000mm² and minimum bending moment 1.5kN·m.

[0022] According to the distribution characteristics of the steepness of the rising edge of the lightning current, the insulator material formula is optimized to improve the solubility of the insulator under the action of lightning energy. On the premise of meeting the requirements of insulation performance and mechanical strength, the structural parameters of the insulator, including the number of sheds, shed spacing and insulator height, are optimized to improve the lightning current tolerance of the insulator and reduce the risk of damage to the insulator by lightning current.

[0023] Obtain the characteristic data of the steepness distribution of the rising edge of the lightning current, and establish a relationship model between the steepness distribution characteristics of the rising edge of the lightning current and the material properties. According to the relationship model, the optimization target is determined, including the material solubility improvement value, the minimum breakdown voltage value and the maximum stress limit value. The insulator material formula is optimized by a multi-objective genetic algorithm to obtain the optimized material component ratio. The molecular dynamics method is used to simulate the dissolution behavior of the optimized material under high-energy impact to obtain the quantitative description parameters of the material solubility improvement mechanism. The quantitative description parameters are input into the finite element analysis software to construct the three-dimensional model of the insulator. Set boundary conditions, perform static and transient analysis on the three-dimensional model, and simulate the stress distribution and dynamic response of the insulator under different structural parameters and steepness of the rising edge of the lightning current. According to the stress distribution and dynamic response data, a relationship function between the structural parameters and the lightning current tolerance is established, and the lightning current tolerance includes the minimum breakdown voltage value and the maximum stress limit value. The particle swarm optimization algorithm is adopted to optimize the structural parameters of the insulator in combination with the lightning current withstand capacity under the constraints of satisfying the minimum breakdown voltage value and the maximum stress limit, so as to obtain the optimal number of sheds, shed spacing and insulator height.

[0024] For example, according to the steepness distribution characteristics of the rising edge of lightning current, a model of the relationship between material performance and steepness is established, and a multi-objective genetic algorithm is used to optimize the insulator material formula. At the same time, the proportion of each component is adjusted to improve the solubility, insulation performance and mechanical strength of the material under the action of lightning energy, and the solubility improvement, minimum breakdown voltage and maximum stress limit are set as optimization targets. Molecular dynamics simulation is used to study the dissolution behavior of the material under high-energy impact, and a quantitative description of the solubility improvement mechanism is obtained. The optimized material parameters are input into the finite element analysis software, the three-dimensional model of the insulator is constructed, the boundary conditions are set, and static and transient analysis is performed to simulate the stress distribution and dynamic response of the insulator under the conditions of different structural parameters and steepness of the rising edge of lightning current, and the influence data of the number of sheds, the spacing between sheds and the height of the insulator on the lightning current withstand capacity are obtained. Based on the obtained influence data, the relationship function between the structural parameters and the lightning current withstand capacity is established, and the particle swarm optimization algorithm is used to optimize the structural parameters of the insulator under the constraints of the minimum breakdown voltage and the maximum stress limit, and the optimal number of sheds, the spacing between sheds and the height of the insulator are obtained. The optimized material formula and structural parameters are input into the insulator performance evaluation model, which comprehensively considers the material properties, structural parameters and lightning current characteristics. Through the Monte Carlo simulation method, multiple simulations are performed under different lightning current rising edge steepness conditions, and the lightning current tolerance and damage risk of the insulator are statistically analyzed to output the final insulator optimization design scheme. For the optimization design of distribution line insulators, a lightning current rising edge steepness distribution model is first established based on historical lightning data. For example, the steepness range in a certain area is 10-100kA / μs, and the average value is 50kA / μs. A multi-objective genetic algorithm is used to optimize the insulator material formula, setting the population size to 100, the number of iterations to 500, the crossover probability to 0.8, and the mutation probability to 0.1. The optimization goals include a 20% increase in solubility, a minimum breakdown voltage of 50kV / cm, and a maximum stress limit of 100MPa. Through molecular dynamics simulation, LAMMPS software was used to simulate the behavior of 10,000 atoms at a time scale of 10ns, and the dissolution rate of the material at a steepness of 50kA / μs was obtained to be 2nm / ns. The optimized material parameters were input into the ANSYS finite element software to construct a three-dimensional model of the insulator, with a mesh size of 1mm and a total number of units of about 1 million. Static and transient analysis were performed with a time step of 0.1μs and an analysis duration of 100μs. The simulation results show that at a steepness of 50kA / μs, the maximum stress is 80MPa, which occurs at 10μs. Based on the simulation data, the relationship function between the structural parameters and the lightning current withstand capability was established. The particle swarm optimization algorithm was used with 50 particles and 200 iterations to obtain the optimal structural parameters of 12 sheds, 40mm shed spacing, and 420mm insulator height.Finally, through Monte Carlo simulation, 10,000 simulations were performed, covering the steepness range of 10-100kA / μs. The statistical analysis results showed that the optimized insulators can remain intact in 95% of lightning strikes, and the risk of damage is reduced by 30%. The final output of the insulator optimization design includes the material formula SiO270%, Al2O320%, other additives 10% and the structural parameters corresponding to the above optimal values.

[0025] S103. By real-time monitoring of the ambient temperature and humidity along the distribution line, the dynamic change data of environmental factors are obtained, and the data are used as the input of the mathematical model of insulator performance parameters to obtain the final mathematical model of insulator performance parameters that can adapt to changes in environmental conditions.

[0026] The ambient temperature and humidity data collected by the temperature and humidity sensor network are obtained, and the temperature and humidity sensor network is installed along the distribution line; according to the obtained ambient temperature and humidity data, an autoregressive integrated moving average model is used for preprocessing and feature extraction to obtain a time series feature vector of the environmental factor; the time series feature vector is input into a long short-term memory network, and the influence of the environmental factors on the insulator performance is trained to obtain a performance impact prediction model, and the performance impact prediction model includes two LSTM layers and one fully connected layer; wherein the performance impact prediction model is a mathematical model of the insulator performance parameters: if the performance impact prediction model is completed, the influence function of the ambient temperature and temperature data changes on the electrical and mechanical properties of the insulator is obtained; the influence function is integrated into the trained performance impact prediction model to construct a final mathematical model of the insulator performance parameters.

[0027] Exemplarily, a temperature and humidity sensor network is installed along the distribution line, the sensor accuracy is set to ±0.5°C for temperature, ±3%RH for humidity, and the sampling frequency is 15 minutes / time. Wireless transmission technology is used to collect environmental temperature and humidity data in real time, which are aggregated by the data collector and transmitted to the central processing unit to form a database of dynamic changes in environmental factors. The collected temperature and humidity data are preprocessed and feature extracted using the autoregressive integrated moving average model, including denoising, outlier detection and periodicity analysis. The missing data is supplemented by linear interpolation to obtain the time series feature vector of the environmental factor. The long short-term memory network algorithm is used to train the prediction model of the impact of environmental factors on insulator performance with the time series feature vector as input. The network structure includes two LSTM layers and one fully connected layer. The root mean square error is used to evaluate the prediction accuracy. Specifically, the input of the prediction model for the impact of insulator performance is the time series feature vector, which can include time series data of multiple environmental parameters, such as temperature, humidity, lightning activity frequency, wind speed, etc. The output of the insulator performance impact prediction model is the predicted insulator performance parameters, which can be selected according to actual conditions, such as electrical breakdown voltage, mechanical strength, insulation resistance, etc. Obtain the impact function of temperature and humidity changes on the electrical and mechanical properties of the insulator. Integrate the environmental factor impact function into the trained performance impact prediction model to build a highly adaptable dynamic insulator performance parameter mathematical model. The sliding window method is used to update the model parameters every 6 hours. The insulator performance parameters are updated in real time through online calculation, and the final insulator performance parameters that adapt to changes in environmental conditions are output. The adaptability and accuracy of the model are evaluated through historical data backtesting and performance testing under actual environmental conditions. If the prediction error exceeds the preset error threshold, the model retraining mechanism is triggered. 100 temperature and humidity sensors are installed along a distribution line, covering a range of 50 kilometers. The sensor accuracy is ±0.5°C for temperature and ±3%RH for humidity, and the sampling frequency is 15 minutes / time. The ZigBee wireless network is used to transmit data in real time. After being aggregated by 10 data collectors, it is transmitted to the central processing unit through the 4G network to form a database of dynamic changes in environmental factors, accumulating 9,600 records per day. The data was preprocessed using the autoregressive integrated moving average model ARIMA, with the order p=2, d=1, q=1, for denoising and outlier detection, and data exceeding 3 times the standard deviation was marked as abnormal. Missing data was supplemented by linear interpolation, and a 96-dimensional time series feature vector was finally obtained. A long short-term memory network LSTM was constructed, which contained two LSTM layers, each with 64 neurons and a fully connected layer with 32 neurons. The input was a 96-dimensional feature vector and the output was the four key performance parameters of the insulator. 80% of the data was used for training and 20% for validation. The Adam optimizer was used, with a learning rate of 0.001, a batch size of 32, and 500 rounds of training.The root mean square error of the model on the validation set is 0.05, and the impact function of temperature and humidity changes on insulator performance is obtained. The impact function is integrated into the original model, and the model parameters are updated every 6 hours using the data from the last 7 days. By comparing the historical data of the past 3 months, the prediction accuracy of the new model is improved by 15%. In the actual environment, after 30 days of continuous monitoring, the average deviation between the model prediction value and the measured value is controlled within ±3%. If the prediction error exceeds 5%, the model retraining mechanism is triggered to ensure that the model continues to adapt to environmental changes.

[0028] S104. Taking the steepness requirement of the rising edge of lightning current and environmental adaptability as the optimization goal, taking the insulator material and structural parameters as the optimization variables, and applying the multi-objective optimization method, the optimal combination of insulator material and structural parameters that meets the actual needs of the distribution line is obtained based on the final insulator performance parameter mathematical model.

[0029] According to the actual needs of the distribution line, the specific requirements under different geographical environments and operating conditions are obtained, and the specific requirements are converted into constraints and objective functions of the optimization problem. Based on the constraints and objective functions, a mathematical description of the multi-objective optimization problem is established, and the mathematical description includes the steepness requirements of the rising edge of the lightning current and the environmental adaptability as the optimization objective function. According to the mathematical description, the insulator material parameters and structural parameters are set as optimization variables to determine the material strength, insulation performance and cost. The non-dominated sorting genetic algorithm is used to encode the insulator material parameters and structural parameters to obtain the chromosome structure. A new population is generated through the selection, crossover and mutation operations of the genetic algorithm. If the termination condition is met, the optimal solution is selected from the final non-dominated solution set. If the termination condition is not met, the genetic algorithm operation is repeated.

[0030] For example, according to the actual needs of the distribution line, the specific requirements under different geographical environments and operating conditions are analyzed and converted into constraints and objective functions of the optimization problem. According to the mathematical model of the insulator performance parameters, a mathematical description of the multi-objective optimization problem is established, and the steepness requirement of the rising edge of the lightning current and the environmental adaptability are used as the optimization objective function, where the environmental adaptability is quantified by the performance stability index under temperature and humidity changes. The insulator material parameters and structural parameters are used as optimization variables, and the material strength, insulation performance and cost are set, and the parameter range and the objective function normalization processing method are clarified. The non-dominated sorting genetic algorithm is used to initialize the population, set the crossover probability, mutation probability and number of iterations, encode the insulator material and structural parameters, and construct the chromosome structure. Through the selection, crossover and mutation operations of the genetic algorithm, a new population is generated, and the population is sorted and selected using fast non-dominated sorting and crowding distance calculation to obtain the non-dominated solution set of each generation. The iterative optimization process is repeated until the termination condition is reached. From the final non-dominated solution set, the fuzzy comprehensive evaluation method is used to quantify the decision maker's preference, set the weight of each goal, select the optimal solution, and obtain the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line. For the optimization design of insulators for a distribution line, the actual needs are analyzed first. For example, the line is located in a coastal area with frequent lightning activities, an average annual temperature change of -10°C to 40°C, and a relative humidity of 30% to 95%. The steepness requirement of the rising edge of the lightning current is set to 50kA / μs, and the environmental adaptability index is defined as the standard deviation of the insulation strength under temperature and humidity changes not exceeding 5%. The optimization variables include materials, such as ceramics, glass, composite materials, and structural parameters, including insulator height 200-400mm and the number of sheds 8-16. The constraints are set as mechanical strength ≥70kN, breakdown voltage ≥110kV / cm, and cost ≤500 yuan / piece. The non-dominated sorting genetic algorithm is used, with the population size set to 100, the crossover probability 0.8, the mutation probability 0.1, and the maximum number of iterations 1000. The material is coded with integers, such as 1-ceramic, 2-glass, 3-composite materials, and the structural parameters are coded with real numbers. The chromosome structure is [material, height, number of sheds]. New populations were generated using binary tournament selection, simulated binary crossover, and polynomial mutation operations. Standardized Euclidean distance was used to calculate crowding distance. After 987 iterations, 20 non-dominated solutions were obtained. Using the fuzzy comprehensive evaluation method, the lightning performance weight was set to 0.6 and the environmental adaptability weight was set to 0.4. The optimal solution was selected from the non-dominated solution set as a composite material with a height of 350mm and 14 shed skirts. The insulation strength of this solution is 125kV / cm at a steepness of 50kA / μs, and the performance standard deviation is 3.8% under temperature and humidity changes, which meets all constraints.

[0031] S105. Obtain and combine the insulator manufacturing database, determine the specific manufacturing process parameters of the insulator according to the optimal combination of insulator materials and structural parameters that meet the actual needs of the distribution line, and produce and put the distribution line into operation.

[0032] The process parameter ranges corresponding to different insulator materials are obtained from the insulator manufacturing database, and the process parameter ranges are used for subsequent cluster analysis; the process parameter ranges are processed by a fuzzy c-means clustering algorithm to obtain a value interval of key parameters of the manufacturing process; a response surface model is established according to the value interval of the key parameters, and the response surface model characterizes the relationship between the process parameters and the insulator performance indicators; a multi-objective genetic algorithm is optimized for the response surface model, and if there are multiple objective function conflicts, the Pareto optimal set is used for processing; the optimal specific manufacturing process parameters obtained by the multi-objective genetic algorithm optimization are input into a manufacturing execution system, and the manufacturing execution system generates production instructions and quality control standards according to the optimal specific manufacturing process parameters and puts into operation a distribution line, and produces and puts into operation a distribution line based on the production instructions and quality control standards.

[0033] Exemplarily, according to the optimal combination of insulator material and structural parameters, database query conditions are set, corresponding process parameter ranges are extracted for different materials, fuzzy c-means clustering algorithm is used to classify process parameters, and key parameters of manufacturing process and their value ranges are determined. The response surface method is used to establish a relationship model between process parameters and insulator performance indicators, an orthogonal experimental design scheme is set, the test points are determined by Box-Behnken design method, the interaction between process parameters is considered, and the least squares method is used to fit the response surface. Based on the response surface model, a multi-objective genetic algorithm is used to optimize process parameters, with a population size of 100, a crossover probability of 0.8, and a mutation probability of 0.1. Multiple performance indicators of insulators are used as objective functions, and the Pareto optimal set is used to deal with multi-objective conflicts. The optimal process parameter combination is solved with process constraints as boundary conditions. The optimized process parameters are input into the manufacturing execution system to automatically generate production instructions and quality control standards. The production process is monitored in real time through the data acquisition system. XR control charts are used for statistical process control. Product performance tests are carried out, including mechanical strength, insulation performance and weather resistance tests. Insulator products that meet the actual needs of distribution lines are output, and a distribution line commissioning preparation plan is formulated. For the optimization of the manufacturing of insulators for a distribution line, first, according to the optimal combination, such as composite materials, height 350mm, and 14 sheds, the manufacturing database is queried, and the process parameters are extracted in the range of curing temperature 130-170°C, curing time 30-60min, and pultrusion speed 0.5-2m / min. The fuzzy c-means clustering algorithm is used, with the number of clusters set to 3, the fuzzy index set to 2, and the number of iterations set to 100. The key parameters are obtained, namely, curing temperature 150°C, curing time 45min, and pultrusion speed 1m / min. The Box-Behnken design method is used to design 15 test points. Considering the interaction of parameters, the quadratic polynomial is used to fit the response surface, and the determination coefficient R² reaches 0.95. A multi-objective genetic algorithm was used with a population size of 100, a crossover probability of 0.8, a mutation probability of 0.1, and 500 iterations. The mechanical strength and insulation strength were used as the optimization targets to obtain the Pareto optimal set. The compromise solution was selected as a curing temperature of 155°C, a curing time of 50min, and a pultrusion speed of 1.2m / min. The parameters were input into the manufacturing execution system, 5 key control points were set, and the XR control chart was used for monitoring. The upper and lower control limits were ±3σ and ±2σ, respectively. The product performance test showed that the mechanical strength was 75kN, the insulation strength was 130kV / cm, and the performance decay was <5% after 1000h salt spray test. A commissioning preparation plan was formulated, including transportation protection measures, installation procedures, and live testing procedures to ensure the safe and reliable operation of insulators on distribution lines.

[0034] S106. After the distribution line is put into operation, the working status data of the insulators is collected in real time through the online monitoring device, and the data is correlated with the distribution characteristics of the steepness of the rising edge of the lightning current, and the use of the insulators is analyzed and predicted to discover potential insulator failure hazards.

[0035] The voltage, current, temperature and partial discharge data are obtained from the insulators of the distribution line, and the data are collected by the sensor network installed on the insulators. The voltage, current, temperature and partial discharge data are preprocessed, and the preprocessing includes denoising, outlier detection and data standardization to obtain the preprocessed insulator state data. According to the preprocessed insulator state data, the db4 wavelet basis function is used to perform 5-layer wavelet decomposition, extract energy and entropy, and construct the insulator state feature vector. The correlation between the insulator state feature vector and the steepness distribution characteristics of the rising edge of the lightning current is calculated, and a highly correlated feature subset is screened out. The insulator performance degradation model is constructed using a long short-term memory network, and the long short-term memory network includes two LSTM layers and one fully connected layer. The input is the highly correlated feature subset and historical performance data to predict the remaining service life and potential failure risk of the insulator. According to the predicted remaining service life and potential failure risk, an insulator failure warning mechanism is established, and differentiated warning thresholds are set for insulation breakdown and mechanical damage. If the predicted failure risk or performance degradation exceeds the preset risk threshold, a warning information is automatically generated.

[0036] For example, a sensor network is installed on the insulator of the distribution line to collect the voltage, current, temperature and partial discharge data of the insulator in real time. The sampling frequency is set to 1kHz. At the same time, a lightning current monitoring device is deployed around the line to record the current waveform of the lightning event. The collected data is transmitted to the data processing center through a wireless transmission module to form an insulator operation status and lightning current database. The collected insulator state data is preprocessed using the time series analysis method, including denoising, outlier detection and data standardization. The db4 wavelet basis function is used to perform a 5-layer wavelet decomposition, extract energy and entropy, and construct the insulator state feature vector. The mutual information analysis method is used to calculate the correlation between the insulator state feature vector and the steepness distribution characteristics of the rising edge of the lightning current, and a highly correlated feature subset is screened out. The long short-term memory network LSTM is used to construct an insulator performance degradation model. The network contains two LSTM layers, each with 64 neurons and a fully connected layer. The input is the screened feature subset and historical performance data to predict the remaining service life and potential failure risk of the insulator. Based on the prediction results, an insulator fault warning mechanism is established, and differentiated warning thresholds are set for insulation breakdown and mechanical damage. When the predicted fault risk or performance degradation exceeds the preset risk threshold, warning information is automatically generated, and corresponding maintenance or replacement suggestions are provided according to the fault type. Taking a 500kV distribution line as an example, intelligent sensors are installed on the insulator strings of each tower to collect voltage, current, temperature and partial discharge data. The sampling frequency is 1kHz, and the data is transmitted every 10 minutes. At the same time, lightning current monitoring devices are installed at both ends of the line to record the current waveform of lightning strike events with a sampling rate of 10MHz. In the data preprocessing stage, median filtering is used to remove noise, 3σ principle is used to eliminate outliers, and data standardization is achieved through maximum and minimum value normalization. The db4 wavelet basis function is used to perform 5-layer wavelet decomposition, extract energy, entropy, standard deviation, and construct state feature vectors. Through mutual information analysis, the correlation between features and lightning current steepness is calculated, and feature subsets with mutual information values ​​greater than 0.6 are selected. A two-layer LSTM network was constructed, with 64 neurons in each layer. The input was 30 days of historical data, and the output was the performance forecast for the next 7 days. The model was trained using the Adam optimizer, with a learning rate of 0.001, a batch size of 32, and 500 rounds of training. The early warning mechanism sets three thresholds: a level 1 alarm for insulation breakdown risk >30%, a level 2 alarm for mechanical strength reduction >20%, and a level 3 alarm for partial discharge intensity increase >50%. In actual operation, after a certain insulator string experienced a lightning strike with a peak value of 80kA and a steepness of 70kA / μs, the LSTM model predicted that its insulation breakdown risk after 7 days would be 35%, triggering a level 1 alarm, and the system automatically generated a replacement recommendation.

[0037] S107. If an insulator is found to have a potential fault, a corresponding replacement product is matched from the insulator manufacturing database according to the location and model of the faulty insulator, and an insulator replacement instruction is triggered to ensure safe and stable operation of the distribution line.

[0038] The precise location coordinates and model information of the faulty insulator are obtained; based on the precise location coordinates and model information, a geographic information system is used to locate the pole tower number and line section where the faulty insulator is located; the model information of the faulty insulator and the environmental parameters obtained from the on-site sensors are input into the insulator manufacturing database query system; a fuzzy matching algorithm based on cosine similarity is used to search for replacement products that meet the requirements; based on the search results, a list of candidate replacement products is outputted in descending order according to the matching degree; a decision tree algorithm is used to screen the candidate replacement products to obtain the optimal replacement product; based on the optimal replacement product, a replacement plan and a procurement recommendation are generated; a replacement product allocation instruction is sent to the nearest intelligent storage center; the nearest storage warehouse with sufficient inventory is selected; based on the replacement plan and the procurement recommendation, an insulator replacement operation plan is generated; in response to the completion of the insulator replacement, an insulator performance acceptance test is performed; an insulator performance acceptance test is performed; if the performance acceptance test result meets the preset operation requirements, it is determined that the newly installed insulator meets the operation standards.

[0039] Exemplarily, according to the fault warning information issued by the online monitoring system, the precise location coordinates and model information of the faulty insulator are extracted, the tower number and line section where the faulty insulator is located are located through the geographic information system, and the situation of multiple fault points is processed at the same time to generate a list of faulty insulators. The model information of the faulty insulator and the environmental parameters obtained from the on-site sensors, including temperature, humidity, and contamination, are input into the insulator manufacturing database query system, and a fuzzy matching algorithm based on cosine similarity is used to search for replacement products that meet the requirements, and a list of candidate replacement products is output in descending order according to the matching degree. The candidate replacement products are screened using a decision tree algorithm, and the decision nodes include product performance parameters, including mechanical strength and insulation strength, inventory status, transportation distance and cost. The optimal replacement product is selected by weighted scoring, and a replacement plan and procurement recommendations are generated. The replacement product allocation instruction is sent to the nearest intelligent storage center through the remote control system, and the nearest storage warehouse with sufficient inventory is selected. At the same time, an insulator replacement operation plan is generated, including power outage time, replacement process and safety measures, and the relevant information is pushed to the mobile operation terminal, and intuitive installation guidance is provided to on-site operators through augmented reality technology. After the replacement is completed, the insulator performance acceptance test is performed, including insulation resistance measurement and partial discharge detection, to ensure that the newly installed insulators meet the operating requirements. In a 500kV transmission line, the online monitoring system detected abnormalities in the insulator strings of towers 3 and 7. The system immediately located the fault location and determined that the coordinates of tower 3 were (116.3°E, 39.9°N) and the coordinates of tower 7 were (116.4°E, 40.0°N). The insulator model was FXBW4-500 / 160. The on-site sensor showed that the ambient temperature was 28°C, the relative humidity was 75%, and the contamination was 0.35mg / cm². The cosine similarity was used to search for replacement products in the database, and the matching threshold was set to 0.9, resulting in 5 candidate products. The decision tree algorithm set four nodes, namely mechanical strength, insulation strength, inventory, and transportation cost. After calculation, the model FXBW4-500 / 160A was selected as the optimal replacement product. The system sends a dispatch instruction to the nearest storage center A to retrieve 6 sets of insulator strings. A replacement plan is generated, with a planned power outage time of 6:00-10:00 the next day. The replacement process includes 8 steps. 3D installation instructions are pushed to on-site personnel through AR glasses, showing the installation position and torque requirements of each insulator. After the replacement is completed, insulation resistance tests and partial discharge tests are carried out. Both sets of insulator strings pass the acceptance test and power supply is restored.

[0040] The present invention provides a distribution network lightning monitoring system, which mainly includes: a lightning data analysis module, which is used to obtain historical lightning current data of the geographical location of the distribution line, analyze the correlation between the steepness of the rising edge of the lightning current and the charge distribution of the thundercloud, and obtain the distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location of the distribution line; an insulator performance modeling module, which is used to determine the insulation performance and mechanical strength requirements that the distribution line insulator needs to meet according to the obtained distribution characteristics of the steepness of the rising edge of the lightning current, and establish an insulator performance parameter mathematical model in combination with the distance between the lightning strike point and the line, the material of the insulator, and the structural parameters; an environmental adaptability optimization module, which is used to obtain dynamic change data of environmental factors through real-time monitoring of the ambient temperature and humidity along the distribution line, and use it as the input of the insulator performance parameter mathematical model to obtain the final insulator performance parameter mathematical model that can adapt to changes in environmental conditions;

[0041] The multi-objective optimization calculation module is used to take the steepness of the rising edge of lightning current and environmental adaptability as the optimization objectives, and the insulator material and structural parameters as the optimization variables. The multi-objective optimization method is used to obtain the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line based on the final insulator performance parameter mathematical model; the insulator parameter optimization module obtains and combines the insulator manufacturing database, determines the specific manufacturing process parameters of the insulator according to the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line, and produces and puts the distribution line into operation; the online monitoring and maintenance module is used to collect the working status data of the insulator in real time through the online monitoring device after the distribution line is put into operation, and correlates it with the distribution characteristics of the steepness of the rising edge of lightning current, analyzes and predicts the use of insulators, and discovers potential insulator failure hazards;

[0042] The fault diagnosis and maintenance instruction module is used to match the corresponding replacement product from the insulator manufacturing database according to the location and model of the faulty insulator if a potential fault is found in the insulator, and trigger the instruction to replace the insulator to ensure the safe and stable operation of the distribution line.

[0043] It should be noted that the above examples are only some specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments, and there are many variations. All variations that can be directly derived or associated with the content disclosed by a person skilled in the art should be considered as the protection scope of the present invention.

Claims

1. A distribution network lightning monitoring method, characterized in that: The method comprises: The historical lightning current data of the geographical location of the distribution line is obtained, and the correlation between the steepness of the rising edge of the lightning current and the charge distribution of the thundercloud is analyzed to obtain the distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location of the distribution line; according to the obtained distribution characteristics of the steepness of the rising edge of the lightning current, the insulation performance and mechanical strength requirements that the distribution line insulators need to meet are determined, and the mathematical model of the insulator performance parameters is established in combination with the distance between the lightning strike point and the line, the material and structural parameters of the insulator; through real-time monitoring of the ambient temperature and humidity along the distribution line, the dynamic change data of environmental factors are obtained, and it is used as the input of the mathematical model of the insulator performance parameters to obtain the final mathematical model of the insulator performance parameters that can adapt to changes in environmental conditions; with the requirements of the steepness of the rising edge of the lightning current and environmental adaptability as the optimization goal, the insulator material and structural parameters are used as the optimization variables The optimal combination of insulator materials and structural parameters that meet the actual needs of distribution lines is obtained by using a multi-objective optimization method based on the mathematical model of the final insulator performance parameters; the insulator manufacturing database is obtained and combined, and the specific manufacturing process parameters of the insulator are determined according to the optimal combination of insulator materials and structural parameters that meet the actual needs of distribution lines, and the distribution lines are produced and put into operation; after the distribution lines are put into operation, the working status data of the insulators are collected in real time through online monitoring devices, and the data is correlated with the distribution characteristics of the steepness of the rising edge of the lightning current, and the use of insulators is analyzed and predicted to discover potential insulator failure hazards; if a failure hazard is found in the insulator, the corresponding replacement product is matched from the insulator manufacturing database according to the location and model of the faulty insulator, and the instruction to replace the insulator is triggered to ensure the safe and stable operation of the distribution lines.

2. The method according to claim 1, characterized in that: The method of obtaining historical lightning current data of the geographical location of the distribution line, analyzing the correlation between the steepness of the rising edge of the lightning current and the charge distribution of the thundercloud, and obtaining the distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location of the distribution line includes: Acquire historical lightning current data of the geographical location of the distribution line, and extract rising edge steepness characteristics from the historical lightning current data; The K-means clustering algorithm is used to classify the rising edge steepness characteristics to obtain the preliminary distribution characteristics of the lightning current rising edge steepness at the geographical location; A thundercloud charge distribution model is established using a Gaussian mixture model according to thundercloud charge distribution data, and the thundercloud charge distribution model is used to correct the preliminary distribution characteristics of the rising edge steepness; Morlet wavelet transform is used to perform time-frequency analysis on the preliminary distribution characteristics of the corrected rising edge steepness to obtain the steepness distribution characteristics at different time scales. A characteristic vector matrix is ​​constructed based on the steepness distribution characteristics at different time scales, and the characteristic vector matrix is ​​used to perform dimensionality reduction processing on the historical lightning current data to obtain a lightning current characteristic data set after dimensionality reduction; Based on the dimensionality-reduced lightning current characteristic data set, a long short-term memory network is constructed as a lightning current rising edge steepness prediction model. The long short-term memory network is used to predict the rising edge steepness of the newly collected lightning current data, and the distribution characteristics of the lightning current rising edge steepness at the geographical location are output.

3. The method according to claim 1, characterized in that The insulation performance and mechanical strength requirements that the distribution line insulators need to meet are determined based on the obtained distribution characteristics of the steepness of the rising edge of the lightning current, and a mathematical model of the insulator performance parameters is established in combination with the distance between the lightning strike point and the line, the material of the insulator, and the structural parameters, including: According to the steepness distribution characteristics of the rising edge of lightning current, a function of the relationship between the withstand voltage level and the steepness of the insulator is established, the support vector regression method is used to fit the data points, and the radial basis function is used to process the nonlinear relationship to obtain the insulation performance requirement function that the insulator needs to meet; According to the insulation performance requirement function, a mechanical strength requirement model is constructed by the correlation between the steepness of the rising edge of the lightning current, the rise time and the peak current and the mechanical impact force, and the nonlinear least squares method is used to fit the historical data to obtain the mechanical strength requirement function that the insulator needs to meet; According to the distance between the lightning strike point and the line, the material and structural parameters of the insulator, the insulator performance parameter influencing factor matrix is ​​established, the direct relationship between the influencing factors and the performance parameters is established by using the multivariate linear regression method, and the relationship function between the insulator performance parameters and the influencing factors is constructed; Using a multi-layer perceptron neural network, taking the insulation performance requirement function, the mechanical strength requirement function and the relationship function between the performance parameter and the influencing factor as input, after data preprocessing, standardization and normalization, setting the number of hidden layer nodes to twice that of the input layer, using a ReLU activation function, using an Adam optimizer and a mean square error loss function, adjusting hyperparameters through cross-validation, and training a multi-layer perceptron neural network to obtain the insulator performance parameter mathematical model; A safety factor calculation node is added to the output layer. If the key performance parameter value output by the insulator performance parameter mathematical model is less than a preset performance parameter threshold, the weight of the safety factor calculation node is adjusted and the insulator performance parameter mathematical model is retrained.

4. The method according to claim 3, characterized in that The invention optimizes the insulator material formula according to the distribution characteristics of the steepness of the rising edge of the lightning current, improves the solubility of the insulator under the action of lightning energy; and optimizes the structural parameters of the insulator, including the number of sheds, the spacing between sheds and the height of the insulator, to improve the lightning current tolerance of the insulator and reduce the risk of damage to the insulator by lightning current, under the premise of meeting the requirements of insulation performance and mechanical strength. Acquire the steepness distribution characteristic data of the rising edge of lightning current, and establish a relationship model between the steepness distribution characteristic of the rising edge of lightning current and material properties; Determining optimization targets according to the relationship model, including material solubility improvement value, minimum breakdown voltage value and maximum stress limit value; The multi-objective genetic algorithm is used to optimize the insulator material formula and obtain the optimized material component ratio; Molecular dynamics methods are used to simulate the dissolution behavior of optimized materials under high-energy impact to obtain quantitative descriptive parameters of the material solubility improvement mechanism; Inputting the quantitative description parameters into finite element analysis software to construct a three-dimensional model of the insulator; Setting boundary conditions, performing static and transient analysis on the three-dimensional model, and simulating stress distribution and dynamic response of insulators subjected to lightning current impact under conditions of different structural parameters and steepness of lightning current rising edge; According to the stress distribution and dynamic response data, a relationship function between structural parameters and lightning current withstand capability is established, wherein the lightning current withstand capability includes a minimum breakdown voltage value and a maximum stress limit value; The particle swarm optimization algorithm is adopted to optimize the structural parameters of the insulator in combination with the lightning current withstand capacity under the constraints of satisfying the minimum breakdown voltage value and the maximum stress limit, so as to obtain the optimal number of sheds, shed spacing and insulator height.

5. The method according to claim 1, characterized in that The real-time monitoring of the ambient temperature and humidity along the distribution line is used to obtain the dynamic change data of environmental factors, and the data is used as the input of the mathematical model of the insulator performance parameters to obtain the final mathematical model of the insulator performance parameters that can adapt to the changes in environmental conditions, including: Acquiring ambient temperature and humidity data collected by a temperature and humidity sensor network installed along a power distribution line; According to the acquired ambient temperature and humidity data, the autoregressive integrated moving average model is used for preprocessing and feature extraction to obtain the time series feature vector of the environmental factors; The time series feature vector is input into a long short-term memory network, and the influence of environmental factors on the performance of the insulator is trained to obtain a performance impact prediction model, wherein the performance impact prediction model comprises two LSTM layers and one fully connected layer; wherein the performance impact prediction model is a mathematical model of the insulator performance parameters: if the performance impact prediction model is completed, the influence function of the ambient temperature and temperature data changes on the electrical and mechanical properties of the insulator is obtained; The influence function is integrated into the trained performance impact prediction model to construct the final insulator performance parameter mathematical model.

6. The method according to claim 1, characterized in that The optimization goal is to meet the steepness requirements of the rising edge of lightning current and environmental adaptability, and the insulator material and structural parameters are used as optimization variables. The multi-objective optimization method is used to obtain the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line based on the final insulator performance parameter mathematical model, including: Obtain specific requirements under different geographical environments and operating conditions based on actual needs of distribution lines, and convert the specific requirements into constraints and objective functions of the optimization problem; Establishing a mathematical description of a multi-objective optimization problem based on the constraint conditions and the objective function, wherein the mathematical description includes the steepness requirement of the rising edge of the lightning current and the environmental adaptability as the optimization objective function; According to the mathematical description, material parameters and structural parameters of the insulator are set as optimization variables to determine material strength, insulation performance and cost; A non-dominated sorting genetic algorithm is used to encode the material parameters and structural parameters of the insulator to obtain a chromosome structure; A new population is generated through the selection, crossover and mutation operations of the genetic algorithm. If the termination condition is met, the optimal solution is selected from the final non-dominated solution set. If the termination condition is not met, the operation of the genetic algorithm is repeated.

7. The method according to claim 1, characterized in that The obtaining and combining of the insulator manufacturing database, determining the specific manufacturing process parameters of the insulator according to the optimal combination of insulator materials and structural parameters that meet the actual needs of the distribution line, and producing and putting the distribution line into operation, includes: Obtaining process parameter ranges corresponding to different insulator materials from an insulator manufacturing database, wherein the process parameter ranges are used for subsequent cluster analysis; The process parameter range is processed by using a fuzzy c-means clustering algorithm to obtain a value interval of a key parameter of the manufacturing process; Establishing a response surface model according to the key parameter value range, wherein the response surface model characterizes the relationship between the process parameters and the insulator performance index; A multi-objective genetic algorithm is used to optimize the response surface model. If there are multiple objective functions conflicting, the Pareto optimal set is used for processing. The optimal specific manufacturing process parameters obtained through multi-objective genetic algorithm optimization are input into the manufacturing execution system, and the manufacturing execution system generates production instructions and quality control standards according to the optimal specific manufacturing process parameters, and produces and puts into operation the distribution lines based on the production instructions and quality control standards.

8. The method according to claim 1, characterized in that After the distribution line is put into operation, the working status data of the insulator is collected in real time through the online monitoring device, and the data is correlated with the distribution characteristics of the steepness of the rising edge of the lightning current to analyze and predict the use of the insulator to discover potential insulator failure hazards, including: Obtaining voltage, current, temperature and partial discharge data from distribution line insulators, wherein the voltage, current, temperature and partial discharge data are collected by a sensor network installed on the insulators; Preprocessing the voltage, current, temperature and partial discharge data, wherein the preprocessing includes denoising, outlier detection and data standardization to obtain preprocessed insulator state data; According to the preprocessed insulator state data, db4 wavelet basis function is used to perform 5-layer wavelet decomposition, extract energy and entropy, and construct an insulator state feature vector; Calculating the correlation between the insulator state feature vector and the steepness distribution feature of the lightning current rising edge, and screening out a highly correlated feature subset; A long short-term memory network is used to construct an insulator performance degradation model, wherein the long short-term memory network includes two LSTM layers and one fully connected layer, and the input is the highly correlated feature subset and historical performance data to predict the remaining service life and potential failure risk of the insulator; Based on the predicted remaining service life and potential failure risk, an insulator failure warning mechanism is established, and differentiated warning thresholds are set for insulation breakdown and mechanical damage. If the predicted failure risk or performance degradation exceeds the preset risk threshold, a warning message is automatically generated.

9. The method according to claim 1, characterized in that: If a potential fault is found in the insulator, a corresponding replacement product is matched from the insulator manufacturing database according to the location and model of the faulty insulator, and an insulator replacement instruction is triggered to ensure the safe and stable operation of the distribution line, including: Obtain the precise location coordinates and model information of the faulty insulator; According to the precise location coordinates and model information, a geographic information system is used to locate the tower number and line section where the faulty insulator is located; Inputting the model information of the faulty insulator and the environmental parameters obtained from the on-site sensor into the insulator manufacturing database query system; Use fuzzy matching algorithm based on cosine similarity to search for replacement products that meet the requirements; According to the search results, output a list of candidate replacement products in descending order of matching degree; Using a decision tree algorithm to screen the candidate replacement products to obtain the optimal replacement product; Generate replacement plans and purchase recommendations based on the optimal replacement products; Send replacement product allocation instructions to the smart storage center; Choose a warehouse that is closest and well-stocked; Generate an insulator replacement work plan based on the replacement options and procurement recommendations; In response to the insulator replacement being completed, performing an insulator performance acceptance test; If the performance acceptance test results meet the preset operating requirements, it is determined that the newly installed insulator meets the operating standards.

10. Distribution network lightning monitoring system, characterized in that: The system comprises: a lightning data analysis module, which is used to obtain historical lightning current data of the geographical location of the distribution line, analyze the correlation between the steepness of the rising edge of the lightning current and the charge distribution of the thundercloud, and obtain the distribution characteristics of the steepness of the rising edge of the lightning current at the geographical location of the distribution line; an insulator performance modeling module, which is used to determine the insulation performance and mechanical strength requirements that the distribution line insulator needs to meet according to the obtained distribution characteristics of the steepness of the rising edge of the lightning current, and establish an insulator performance parameter mathematical model in combination with the distance between the lightning strike point and the line, the material of the insulator, and the structural parameters; an environmental adaptability optimization module, which is used to obtain dynamic change data of environmental factors through real-time monitoring of the ambient temperature and humidity along the distribution line, and use it as the input of the insulator performance parameter mathematical model to obtain the final insulator performance parameter mathematical model that can adapt to changes in environmental conditions; The multi-objective optimization calculation module is used to take the steepness of the rising edge of lightning current and environmental adaptability as the optimization objectives, and the insulator material and structural parameters as the optimization variables. The multi-objective optimization method is used to obtain the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line based on the final insulator performance parameter mathematical model; the insulator parameter optimization module obtains and combines the insulator manufacturing database, determines the specific manufacturing process parameters of the insulator according to the optimal combination of insulator material and structural parameters that meet the actual needs of the distribution line, and produces and puts the distribution line into operation; the online monitoring and maintenance module is used to collect the working status data of the insulator in real time through the online monitoring device after the distribution line is put into operation, and correlates it with the distribution characteristics of the steepness of the rising edge of lightning current, analyzes and predicts the use of insulators, and discovers potential insulator failure hazards; The fault diagnosis and maintenance instruction module is used to match the corresponding replacement product from the insulator manufacturing database according to the location and model of the faulty insulator if a potential fault is found in the insulator, and trigger the instruction to replace the insulator to ensure the safe and stable operation of the distribution line.

Images