500kV overhead line windage yaw lightning stroke flashover risk data driving evaluation method and device

Abstract

The invention provides a 500kV overhead line windage yaw lightning stroke flashover risk data driving evaluation method and device. The method comprises the following steps: scanning a power transmission line and a pole tower by adopting scanning equipment to obtain point cloud data of the power transmission line and the pole tower, and constructing a three-dimensional live-action model of the power transmission line and the pole tower according to the point cloud data; determining a wind speed spectrum parameter according to the wind speed spectrum of the non-uniform wind field, and converting the wind speed spectrum parameter into a wind pressure time history; importing the three-dimensional real scene model and the wind pressure time history into finite element simulation software, and carrying out finite element simulation analysis to obtain a dynamic gap distance between the power transmission line and the tower; constructing a neural network prediction model according to the improved genetic algorithm, and training the neural network prediction model to obtain a risk assessment model; and performing risk assessment on the overhead transmission line according to the risk assessment model and the dynamic gap distance between the transmission line and the tower.

Description

Technical Field

[0001] The present invention relates to the field of power transmission, and in particular to a data-driven assessment method and device for windage lightning flashover risk of a 500kV overhead line. Background Art

[0002] With the development of power systems and the expansion of power grids, overhead transmission lines of 500kV and above are widely used around the world, providing important support for long-distance power transmission. However, these high-voltage overhead lines often face the risk of wind deflection flashover under complex environmental conditions, especially under strong winds, posing a potential threat to the safe and reliable operation of the power grid.

[0003] The assessment of the risk of lightning flashover of overhead lines under wind loads requires the understanding of the development trend of the tower head air gap in the wind field and the dynamic wind deviation distance of the insulator string and the conductor. Traditionally, the assessment of wind deviation flashover risk of overhead lines mainly relies on laboratory tests, theoretical analysis and statistical data. However, these methods have a gap between the simulated environment and the actual scene, and it is difficult to accurately reflect the impact of the actual wind field on power facilities.

[0004] At present, for wind-induced flashover of transmission lines, people mainly focus on how to calculate the wind deflection angle of the insulator string when the transmission line is wind-deflected. The gap between the conductor and the tower is obtained by the wind deflection angle, and compared with the minimum air spacing in the design specification to determine whether wind-induced flashover occurs. However, this method is difficult to adapt to all tower types, that is, the minimum air spacing of different tower types may be different for the same wind deflection angle. In addition, the calculation of the wind deflection angle usually considers the wind deflection characteristics of the conductor under the action of steady-state wind, but the wind in nature is a random process, and the wind deflection response of the conductor and the insulator string to non-uniform wind fields (such as gusts, pulsating wind, etc.) is quite different. Therefore, the method in the relevant technology is not accurate enough when evaluating the risk of lightning wind-induced flashover of overhead transmission lines. How to provide a more accurate method to evaluate the risk of lightning wind-induced flashover of overhead transmission lines has become a top priority. Summary of the invention

[0005] The present invention provides a data-driven assessment method and device for the risk of wind-induced lightning flashover of a 500kV overhead line, which can improve the prediction accuracy of the probability of air gap breakdown of overhead transmission lines.

[0006] On the one hand, a data-driven assessment method for wind-induced lightning flashover risk of 500kV overhead lines is provided, including: Use scanning equipment to scan transmission lines and towers to obtain point cloud data of transmission lines and towers. Construct 3D real-life models of transmission lines and towers based on point cloud data; Determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and convert the wind speed spectrum parameters into wind pressure time history; Import the 3D real-scene model and wind pressure time history into the finite element simulation software, perform finite element simulation analysis, and obtain the dynamic clearance distance between the transmission line and the tower; A neural network prediction model is constructed based on an improved genetic algorithm, and the neural network prediction model is trained to obtain a risk assessment model; Risk assessment of overhead transmission lines is carried out based on the risk assessment model and the dynamic clearance distance between transmission lines and towers.

[0007] Optionally, determining wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and converting the wind speed spectrum parameters into a wind pressure time history includes: Determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, the wind speed spectrum parameters include average wind speed, frequency range and shape parameters of the wind speed spectrum; Generate frequency components, calculate the wind speed amplitude corresponding to each frequency component, apply the harmonic superposition method to each frequency component, and generate a sine wave of the corresponding frequency; superimpose the waveforms of all frequency components together to obtain the wind speed time history; Convert the generated wind speed time history into wind pressure time history.

[0008] Optionally, the three-dimensional real-scene model and wind pressure time history are imported into finite element simulation software to perform finite element simulation analysis to obtain the dynamic clearance distance between the transmission line and the tower, including: Import the 3D real-scene model into the finite element simulation software; Set the deadweight load of insulator strings and conductors in the transmission line, and use the interpolation function to add the wind pressure time to the boundary load; Perform mesh generation and set calculation step size in finite element simulation software; Finite element simulation is performed to obtain the end displacement and wind deflection angle of the insulator string; According to the end displacement of the insulator string and the wind deflection angle, the dynamic gap distance between the conductor, the insulator string and the tower in a non-uniform wind field is obtained; The minimum gap distance between the conductor, insulator string and tower in a non-uniform wind field is obtained according to the dynamic gap distance.

[0009] Optionally, a neural network prediction model is constructed according to an improved genetic algorithm, and the neural network prediction model is trained to obtain a risk assessment model, including: Step 1: Construct a neural network and determine the number of neurons in the input layer, hidden layer, and output layer; Step 2: Determine the activation function of the hidden layer and the output layer. The activation function of the hidden layer is Function, the output layer activation function is function; Step 3: Initialize the weights and bias parameters of the input layer, hidden layer, and output layer, and encode the weights and bias parameters to obtain an initial population, wherein each set of weights and bias parameters is encoded to generate a set of chromosomes; Step 4: Based on the experimental data of the relationship between 50% lightning impulse breakdown voltage and air gap distance, a database of the risks of back-strike and shielding flashover of transmission lines under lightning strike is constructed, and the data in the database is divided into a training set and a test set; Step 5: After decoding the chromosome, the weight and bias parameters are obtained and assigned to the neural network, which is trained using the training samples, tested using the test set, and tested using the test set; Step 6: Calculate the error of the prediction probability of the neural network corresponding to each group of chromosomes as the fitness value of the chromosome; Step 7: Perform a selection operation to select individuals with higher fitness according to the fitness value of the chromosome; Step 8: Perform crossover operation on chromosomes; Step 9, performing mutation operation on chromosomes; Step 10, repeat steps 5 to 8 until the number of iterations meets the stopping condition, and the chromosome with the best fitness is the optimized neural network weight and bias; Step 11, generating a risk assessment model based on the weights and biases of the optimized neural network.

[0010] Optionally, a selection operation is performed to select individuals with higher fitness according to the fitness value of the chromosome, including: In the iteration to Generation, before The chromosome with the highest fitness value is , There is no fitness value higher than The chromosomes will Join generation, and remove the The chromosome with the lowest fitness value in the generation.

[0011] Optionally, a crossover operation is performed on the chromosome, including: Using formula Get the crossover probability ,in represents the upper bound of the crossover probability, represents the lower bound of the crossover probability, Indicates the maximum fitness of the parent chromosome with the larger fitness in the crossover operation. represents the maximum fitness value in the population, represents the average value of the population fitness, is the parameter; According to the crossover probability Get the crossover point position in the chromosome , according to the intersection position Perform a crossover operation on the parent chromosome to obtain the daughter chromosome.

[0012] Optionally, performing mutation operations on the chromosome includes: Using formula Get mutation probability ,in represents the upper bound of the mutation probability, represents the lower bound of the mutation probability, Represents the fitness value of the variant individual; The probability of gene mutation in the offspring chromosome is determined by the mutation probability, and the gene in the offspring chromosome is mutated.

[0013] On the other hand, a 500 kV overhead line windage lightning flashover risk data-driven assessment device is provided, the device comprising: A three-dimensional real scene model generation module is used to scan the transmission line and the tower using a scanning device to obtain point cloud data of the transmission line and the tower, and to construct a three-dimensional real scene model of the transmission line and the tower based on the point cloud data; The simulation module determines the wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and converts the wind speed spectrum parameters into the wind pressure time history; imports the three-dimensional real scene model and the wind pressure time history into the finite element simulation software, performs finite element simulation analysis, and obtains the dynamic clearance distance between the transmission line and the tower; The model training module constructs a neural network prediction model based on the improved genetic algorithm, and uses experimental data to train the neural network prediction model to obtain a risk assessment model; based on the risk assessment model and the dynamic gap distance between the transmission line and the tower, the risk assessment of the overhead transmission line is carried out.

[0014] On the other hand, an electronic device is provided, comprising the 500 kV overhead line windage lightning flashover risk data driven assessment device as described above.

[0015] On the other hand, a computer-readable storage medium is provided, wherein at least one program code is stored in the computer-readable storage medium, and the program code is executed by a processor to implement the data-driven assessment method for wind-induced lightning flashover risk of a 500kV overhead line as described in any one of the above items.

[0016] The technical solution provided by the embodiments of the present disclosure has the following beneficial effects: The present invention solves the problem of poor adaptability of traditional wind deflection angle methods to different tower types, and takes into account the dynamic wind deflection response of conductors and insulator strings under non-uniform wind fields. Combining the real-life three-dimensional model and non-uniform wind field simulation technology, it can more realistically and accurately reflect the natural wind field distribution and the risk of lightning wind deflection flashover. The use of an improved genetic optimization algorithm can find the optimal weight and bias of the neural network in a complex nonlinear search space, thereby achieving an accurate prediction of the probability of air gap breakdown, and taking corresponding measures based on the evaluation results to reduce the probability of lightning wind deflection flashover accidents, providing a decision-making basis for the maintenance and safe operation of overhead transmission lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A method flow chart of a data-driven assessment method for windage lightning flashover risk of a 500kV overhead line provided in an embodiment of the present disclosure; Figure 2 A flow chart of a training risk assessment model provided in an embodiment of the present disclosure; Figure 3 A structural block diagram of a 500kV overhead line windage lightning flashover risk data driven assessment device provided in an embodiment of the present disclosure; Figure 4 A structural block diagram of an electronic device provided in an embodiment of the present disclosure.

[0019] The reference numerals are as follows: 21: 3D real scene model generation module; 22: simulation module; 23: model training module; 31: processor; 32: memory. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Figure 1A method flow chart of a data-driven assessment method for windage lightning flashover risk of 500kV overhead lines provided in an embodiment of the present disclosure. Figure 1 The method steps include: S11. Scan the transmission lines and pole towers using scanning equipment to obtain point cloud data of the transmission lines and pole towers, and construct a three-dimensional real-scene model of the transmission lines and pole towers based on the point cloud data.

[0022] In one example, step S11 includes: The first step is to obtain the LiDAR point cloud data, including the point cloud information data of conductors, towers and insulator strings.

[0023] Step 2: Process point cloud data. Use point cloud data processing software to extract the point cloud data from laser scanning and generate a 3D real scene model.

[0024] S12. Determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and convert the wind speed spectrum parameters into a wind pressure time series.

[0025] In one example, step S12 includes: The first step is to determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field. The wind speed spectrum parameters include average wind speed, frequency range and shape parameters of the wind speed spectrum.

[0026] The second step is to generate frequency components, calculate the wind speed amplitude corresponding to each frequency component, apply the harmonic superposition method to each frequency component, and generate a sine wave of the corresponding frequency; superimpose the waveforms of all frequency components together to obtain the wind speed time series.

[0027] The third step is to convert the generated wind speed time series into wind pressure time series.

[0028] Among them, wind speed hour refers to the hourly average value of wind speed, and wind pressure hour refers to the hourly average value of wind pressure.

[0029] In the disclosed embodiment, the frequency range is determined by the wind speed spectrum, the frequency components are obtained according to the frequency range, and the wind speed time is obtained more accurately by superimposing the waveforms of the frequency components. By converting the wind speed time into the wind pressure time, the wind pressure time can be used to simulate a non-uniform wind field in the simulation software during subsequent simulations.

[0030] S13. Import the three-dimensional real-scene model and wind pressure time history into finite element simulation software, perform finite element simulation analysis, and obtain the dynamic gap distance between the transmission line and the tower.

[0031] In one example, step S13 includes: The first step is to import the three-dimensional real-scene model into the finite element simulation software.

[0032] The second step is to set the deadweight load of the insulator strings and conductors in the transmission line, and use the interpolation function to add the wind pressure time to the boundary load.

[0033] The third step is to perform meshing and set the calculation step size in the finite element simulation software.

[0034] The fourth step is to perform finite element simulation calculations to obtain the end displacement and wind deflection angle of the insulator string.

[0035] Step 5: According to the end displacement of the insulator string and the wind deflection angle, the dynamic gap distance between the conductor, insulator string and tower in the non-uniform wind field is obtained.

[0036] In the disclosed embodiment, under the action of a non-uniform wind field, the gap distance between the conductor, the insulator string and the pole tower will change, and thus the minimum gap distance between the conductor, the insulator string and the pole tower will also change. By obtaining the dynamic gap distance between the conductor, the insulator string and the pole tower in the non-uniform wind field, the minimum minimum gap distance is obtained based on the dynamic gap distance, which is beneficial to improving the accuracy of subsequent risk assessment.

[0037] Step 6: Obtain the minimum gap distance between the conductor, insulator string and tower in a non-uniform wind field based on the dynamic gap distance.

[0038] S14. Construct a neural network prediction model based on the improved genetic algorithm, and train the neural network prediction model to obtain a risk assessment model.

[0039] In one example, step S14 includes: Step 1: Build a neural network and determine the number of neurons in the input layer, hidden layer, and output layer.

[0040] Among them, when a counter-strike occurs, the number of input neurons is 3, which are the lightning current, wind speed and grounding resistance respectively; when a shielding failure occurs, the number of input neurons is 2, which are the lightning current and wind speed respectively. The hidden layers are two layers, each containing 6 neurons. The output layer is a neuron, which is used to output the probability of a counter-strike lightning flashover.

[0041] Step 2: Determine the activation function of the hidden layer and the output layer. The activation function of the hidden layer is Function, the output layer activation function is function.

[0042] in, The function is: .

[0043] Activation Function Defines a neuron in linear transformation The nonlinear output result after that. For the input vector from the previous layer of neural network entering the neuron ,use The neuron of the function will output To the next layer of neurons, enabling neural networks to learn complex patterns and nonlinear features.

[0044] The function is: .

[0045] The function is used for binary classification problems and compresses the input value to the range of 0 to 1 to indicate the probability of belonging to a certain category.

[0046] Step 3: Initialize the weights and bias parameters of the input layer, hidden layer, and output layer, and encode the weights and bias parameters to obtain an initial population, wherein each set of weights and bias parameters is encoded to generate a set of chromosomes.

[0047] Among them, chromosome .

[0048] in, The neural network Chromosomes consisting of layer weights and biases; Represents the neural network The weights of each layer; Represents the neural network The bias of each layer.

[0049] Step 4: Based on the experimental data on the relationship between 50% lightning impulse breakdown voltage and air gap distance, a database of the risks of back-strike and shielding flashover of transmission lines under lightning strikes is constructed, and the data in the database is divided into a training set and a test set.

[0050] Exemplarily, the data in the database is divided into a training set and a test set according to a ratio of 7:3.

[0051] Step 5: After decoding the chromosome, the weight and bias parameters are obtained and assigned to the neural network. The training samples are used for training, the test set is used for testing, and the test set is used for testing.

[0052] Step 6: Calculate the error of the prediction probability of the neural network corresponding to each group of chromosomes as the fitness value of the chromosome.

[0053] The mean square error is used as the fitness function:

[0054] in, It is The true predicted probability of samples is The neural network is based on the genes on the chromosomes and genes The predicted probability, is the number of samples, and the fitness The smaller the function value, the better the weights and biases.

[0055] Step 7: Perform a selection operation to select individuals with higher fitness based on the fitness value of the chromosome.

[0056] In one example, step 7 includes: In the iteration to Generation, before The chromosome with the highest fitness value is , There is no fitness value higher than The chromosomes will Join generation, and remove the The chromosome with the lowest fitness value in the generation.

[0057] In the disclosed embodiment, the selection operation is performed in the above manner, which can ensure that the fitness value of the offspring chromosome generated after the subsequent crossover mutation is better.

[0058] Step 8: Perform crossover operation on chromosomes.

[0059] In one example, step 8 includes: Step 1: Use the formula Get the crossover probability ,in represents the upper bound of the crossover probability, represents the lower bound of the crossover probability, Indicates the maximum fitness of the parent chromosome with the larger fitness in the crossover operation. represents the maximum fitness value in the population, represents the average value of the population fitness, as a parameter.

[0060] in, is 9.903438.

[0061] Step 2: According to the crossover probability Get the crossover point position in the chromosome , according to the intersection position Perform a crossover operation on the parent chromosome to obtain the daughter chromosome.

[0062] Determine the crossover point location based on the crossover probability and the number of genes in the chromosome For example, if the crossover probability is 0.5 and the number of genes in a chromosome is 4, then by multiplying the crossover probability by the number of chromosomes, we can get the crossover point position between the second gene and the third gene.

[0063] For example, by Find the intersection point , then the offspring chromosome code is

[0064]

[0065] For example, two chromosomes and The genes are: ,

[0066] Crossover probability If is 0.5, after the crossover operation is completed, half of the genes of the two chromosomes will be exchanged to generate two sub-chromosomes, and their corresponding genes are: ,

[0067] In the disclosed embodiment, the crossover probability calculated by the above formula is more consistent with the population, which is beneficial to improving the fitness value of the chromosome.

[0068] Step 9: perform mutation operation on chromosomes.

[0069] In one example, step 9 includes: Step 1: Use the formula Get mutation probability ,in represents the upper bound of the mutation probability, represents the lower bound of the mutation probability, Represents the fitness value of the mutated individual.

[0070] The second step is to determine the probability of gene mutation in the offspring chromosomes based on the mutation probability, and perform mutation operations on the genes in the offspring chromosomes.

[0071] Depend on Determine what proportion of genes in a chromosome will mutate. If a gene is selected for mutation, its value will be randomly changed.

[0072] A chromosome The following genes are present:

[0073] Mutation probability When acting on a gene on the chromosome, , then the gene Will be randomly changed:

[0074] in is a random perturbation value, subject to . In the disclosed embodiment, the mutation operation is performed through the above formula, and the obtained mutation probability is more consistent with the population, which is beneficial to improving the fitness value of subsequent chromosomes.

[0075] Step 10, repeat steps 5 to 8 until the number of iterations meets the stopping condition. The chromosome with the best fitness is the optimized neural network weight and bias.

[0076] For example, the number of iterations set is , when the number of iterations is satisfied When , stop the iteration.

[0077] Step 11, generating a risk assessment model based on the weights and biases of the optimized neural network.

[0078] S15. Perform risk assessment on overhead transmission lines based on the risk assessment model and the dynamic clearance distance between the transmission lines and the towers.

[0079] In step S15, the smallest minimum gap distance is first obtained from the dynamic gap distances, and the risk assessment model performs risk assessment on the overhead transmission line according to the smallest minimum gap distance.

[0080] The present invention solves the problem of poor adaptability of traditional wind deflection angle methods to different tower types, and takes into account the dynamic wind deflection response of conductors and insulator strings under non-uniform wind fields. Combining the real-life three-dimensional model and non-uniform wind field simulation technology, it can more realistically and accurately reflect the natural wind field distribution and the risk of lightning wind deflection flashover. The use of an improved genetic optimization algorithm can find the optimal weight and bias of the neural network in a complex nonlinear search space, thereby achieving an accurate prediction of the probability of air gap breakdown, and taking corresponding measures based on the evaluation results to reduce the probability of lightning wind deflection flashover accidents, providing a decision-making basis for the maintenance and safe operation of overhead transmission lines.

[0081] Figure 2 A flowchart of a training risk assessment model provided in an embodiment of the present disclosure. Figure 2 The operation flow in step S14 is shown in FIG.

[0082] Figure 3 This is a structural block diagram of a 500kV overhead line windage lightning flashover risk data driven assessment device provided by an embodiment of the present disclosure. Figure 3, the device comprises: A three-dimensional real scene model generation module 21 is used to scan the transmission line and the tower using a scanning device to obtain point cloud data of the transmission line and the tower, and to construct a three-dimensional real scene model of the transmission line and the tower according to the point cloud data; The simulation module 22 determines wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and converts the wind speed spectrum parameters into wind pressure time history; imports the three-dimensional real scene model and the wind pressure time history into the finite element simulation software, performs finite element simulation analysis, and obtains the dynamic gap distance between the transmission line and the tower; The model training module 23 constructs a neural network prediction model based on the improved genetic algorithm, and uses experimental data to train the neural network prediction model to obtain a risk assessment model; and conducts risk assessment on overhead transmission lines based on the risk assessment model and the dynamic clearance distance between the transmission line and the tower.

[0083] Optionally, the simulation module 22 is used to: Determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, the wind speed spectrum parameters include average wind speed, frequency range and shape parameters of the wind speed spectrum; Generate frequency components, calculate the wind speed amplitude corresponding to each frequency component, apply the harmonic superposition method to each frequency component, and generate a sine wave of the corresponding frequency; superimpose the waveforms of all frequency components together to obtain the wind speed time history; Convert the generated wind speed time history into wind pressure time history.

[0084] Optionally, the simulation module 22 is used to: Import the 3D real-scene model into the finite element simulation software; Set the deadweight load of insulator strings and conductors in the transmission line, and use the interpolation function to add the wind pressure time to the boundary load; Perform mesh generation and set calculation step size in finite element simulation software; Finite element simulation is performed to obtain the end displacement and wind deflection angle of the insulator string; According to the end displacement of the insulator string and the wind deflection angle, the dynamic gap distance between the conductor, the insulator string and the tower in a non-uniform wind field is obtained; The minimum gap distance between the conductor, insulator string and tower in a non-uniform wind field is obtained according to the dynamic gap distance.

[0085] Optionally, the model training module is used to: Step 1: Construct a neural network and determine the number of neurons in the input layer, hidden layer, and output layer; Step 2: Determine the activation function of the hidden layer and the output layer. The activation function of the hidden layer is Function, the output layer activation function is function; Step 3: Initialize the weights and bias parameters of the input layer, hidden layer, and output layer, and encode the weights and bias parameters to obtain an initial population, wherein each set of weights and bias parameters is encoded to generate a set of chromosomes; Step 4: Based on the experimental data of the relationship between 50% lightning impulse breakdown voltage and air gap distance, a database of the risks of back-strike and shielding flashover of transmission lines under lightning strike is constructed, and the data in the database is divided into a training set and a test set; Step 5: After decoding the chromosome, the weight and bias parameters are obtained and assigned to the neural network, which is trained using the training samples, tested using the test set, and tested using the test set; Step 6: Calculate the error of the prediction probability of the neural network corresponding to each group of chromosomes as the fitness value of the chromosome; Step 7: Perform a selection operation to select individuals with higher fitness according to the fitness value of the chromosome; Step 8: Perform crossover operation on chromosomes; Step 9, performing mutation operation on chromosomes; Step 10, repeat steps 5 to 8 until the number of iterations meets the stopping condition, and the chromosome with the best fitness is the optimized neural network weight and bias; Step 11, generating a risk assessment model based on the weights and biases of the optimized neural network.

[0086] Optionally, the model training module 23 is used to: In the iteration to Generation, before The chromosome with the highest fitness value is , There is no fitness value higher than The chromosomes will Join generation, and remove the The chromosome with the lowest fitness value in the generation.

[0087] Optionally, the model training module 23 is used to: Using formula Get the crossover probability ,in represents the upper bound of the crossover probability, represents the lower bound of the crossover probability, Indicates the maximum fitness of the parent chromosome with the larger fitness in the crossover operation. represents the maximum fitness value in the population, represents the average value of the population fitness, is the parameter; According to the crossover probability Get the crossover point position in the chromosome , according to the intersection position Perform a crossover operation on the parent chromosome to obtain the daughter chromosome.

[0088] Optionally, the model training module 23 is used to: Using formula Get mutation probability ,in represents the upper bound of the mutation probability, represents the lower bound of the mutation probability, Represents the fitness value of the variant individual; The probability of gene mutation in the offspring chromosome is determined by the mutation probability, and the gene in the offspring chromosome is mutated.

[0089] Figure 4 This is a structural block diagram of an electronic device provided by an embodiment of the present disclosure. Figure 4 The electronic device includes a 500kV overhead line windage lightning flashover risk data driven assessment device. Generally, the electronic device includes: a processor 31 and a memory 32.

[0090] The processor 31 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 31 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 31 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state.

[0091] The memory 32 may include one or more computer-readable storage media, which may be non-transitory. The memory 32 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 32 is used to store at least one instruction, which is used to be executed by the processor 31 to implement the method for data transmission between databases performed by an electronic device provided in the method embodiment of the present application.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data-driven assessment method for wind-induced lightning flashover risk of 500kV overhead lines, characterized in that: include: Use scanning equipment to scan transmission lines and towers to obtain point cloud data of transmission lines and towers. Construct 3D real-life models of transmission lines and towers based on point cloud data; Determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and convert the wind speed spectrum parameters into wind pressure time history; Import the 3D real-scene model and wind pressure time history into the finite element simulation software, perform finite element simulation analysis, and obtain the dynamic clearance distance between the transmission line and the tower; A neural network prediction model is constructed based on an improved genetic algorithm, and the neural network prediction model is trained to obtain a risk assessment model; Risk assessment of overhead transmission lines is carried out based on the risk assessment model and the dynamic clearance distance between transmission lines and towers.

2. The 500kV overhead line windage lightning flashover risk data-driven assessment method according to claim 1 is characterized in that: The wind speed spectrum parameters are determined according to the wind speed spectrum of the non-uniform wind field, and the wind speed spectrum parameters are converted into wind pressure time history, including: Determine wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, the wind speed spectrum parameters include average wind speed, frequency range and shape parameters of the wind speed spectrum; Generate frequency components, calculate the wind speed amplitude corresponding to each frequency component, apply the harmonic superposition method to each frequency component, and generate a sine wave of the corresponding frequency; superimpose the waveforms of all frequency components together to obtain the wind speed time history; Convert the generated wind speed time history into wind pressure time history.

3. The 500kV overhead line windage lightning flashover risk data-driven assessment method according to claim 1 is characterized in that: Import the 3D real-life model and wind pressure time history into the finite element simulation software for finite element simulation analysis to obtain the dynamic clearance distance between the transmission line and the tower, including: Import the 3D real-scene model into the finite element simulation software; Set the deadweight load of insulator strings and conductors in the transmission line, and use the interpolation function to add the wind pressure time to the boundary load; Perform mesh generation and set calculation step size in finite element simulation software; Finite element simulation is performed to obtain the end displacement and wind deflection angle of the insulator string; According to the end displacement of the insulator string and the wind deflection angle, the dynamic gap distance between the conductor, the insulator string and the tower in a non-uniform wind field is obtained; The minimum gap distance between the conductor, insulator string and tower in a non-uniform wind field is obtained according to the dynamic gap distance.

4. The 500kV overhead line windage lightning flashover risk data-driven assessment method according to claim 1 is characterized in that: A neural network prediction model is constructed based on the improved genetic algorithm, and the neural network prediction model is trained to obtain a risk assessment model, including: Step 1: Construct a neural network and determine the number of neurons in the input layer, hidden layer, and output layer; Step 2: Determine the activation function of the hidden layer and the output layer. The activation function of the hidden layer is Function, the output layer activation function is function; Step 3: Initialize the weights and bias parameters of the input layer, hidden layer, and output layer, and encode the weights and bias parameters to obtain an initial population, wherein each set of weights and bias parameters is encoded to generate a set of chromosomes; Step 4: Based on the experimental data of the relationship between 50% lightning impulse breakdown voltage and air gap distance, a database of the risks of back-strike and shielding flashover of transmission lines under lightning strike is constructed, and the data in the database is divided into a training set and a test set; Step 5: After decoding the chromosome, the weight and bias parameters are obtained and assigned to the neural network, which is trained using the training samples, tested using the test set, and tested using the test set; Step 6: Calculate the error of the prediction probability of the neural network corresponding to each group of chromosomes as the fitness value of the chromosome; Step 7: Perform a selection operation to select individuals with higher fitness according to the fitness value of the chromosome; Step 8: Perform crossover operation on chromosomes; Step 9, performing mutation operation on chromosomes; Step 10, repeat steps 5 to 8 until the number of iterations meets the stopping condition, and the chromosome with the best fitness is the optimized neural network weight and bias; Step 11, generating a risk assessment model based on the weights and biases of the optimized neural network.

5. The 500kV overhead line windage lightning flashover risk data-driven assessment method according to claim 4 is characterized in that: Perform selection operations to select individuals with higher fitness according to the fitness values ​​of chromosomes, including: In the iteration to Generation, before The chromosome with the highest fitness value is , There is no fitness value higher than The chromosomes will Join generation, and remove the The chromosome with the lowest fitness value in the generation.

6. The 500kV overhead line windage lightning flashover risk data-driven assessment method according to claim 4 is characterized in that: Perform crossover operations on chromosomes, including: Using formula Get the crossover probability ,in represents the upper bound of the crossover probability, represents the lower bound of the crossover probability, Indicates the maximum fitness of the parent chromosome with the larger fitness in the crossover operation. represents the maximum fitness value in the population, represents the average value of the population fitness, is the parameter; According to the crossover probability Get the crossover point position in the chromosome , according to the intersection position Perform a crossover operation on the parent chromosome to obtain the daughter chromosome.

7. The 500kV overhead line windage lightning flashover risk data-driven assessment method according to claim 6 is characterized in that: The mutation operations on chromosomes include: Using formula Get mutation probability ,in represents the upper bound of the mutation probability, represents the lower bound of the mutation probability, Represents the fitness value of the variant individual; The probability of gene mutation in the offspring chromosome is determined by the mutation probability, and the gene in the offspring chromosome is mutated.

8. A 500kV overhead line windage lightning flashover risk data driven assessment device, characterized in that: The device includes: A three-dimensional real scene model generation module is used to scan the transmission line and the tower using a scanning device to obtain point cloud data of the transmission line and the tower, and to construct a three-dimensional real scene model of the transmission line and the tower based on the point cloud data; The simulation module determines the wind speed spectrum parameters according to the wind speed spectrum of the non-uniform wind field, and converts the wind speed spectrum parameters into the wind pressure time history; imports the three-dimensional real scene model and the wind pressure time history into the finite element simulation software, performs finite element simulation analysis, and obtains the dynamic clearance distance between the transmission line and the tower; The model training module constructs a neural network prediction model based on the improved genetic algorithm, and uses experimental data to train the neural network prediction model to obtain a risk assessment model; based on the risk assessment model and the dynamic gap distance between the transmission line and the tower, the risk assessment of the overhead transmission line is carried out.

9. An electronic device, characterized in that: The electronic equipment includes the 500kV overhead line windage lightning flashover risk data driven assessment device as described in claim 8.

10. A computer-readable storage medium, characterized in that: At least one program code is stored in the computer-readable storage medium, and the program code is executed by a processor to implement the data-driven assessment method for wind-induced lightning flashover risk of a 500 kV overhead line as described in any one of claims 1 to 7.

Images