Method and system for evaluating structural reliability of power transmission tower

Abstract

The invention discloses a power transmission tower structure reliability assessment method and system, and the method comprises the steps: building a reliability database of a power transmission tower, training a deep neural network model based on variable structure parameters in the database and corresponding reliability correction coefficients, inputting the structure parameters of a to-be-assessed power transmission tower into the trained deep neural network model, and obtaining the reliability of the to-be-assessed power transmission tower. The reliability correction coefficient is output, the reliability index of the to-be-evaluated power transmission tower is calculated according to the reliability correction coefficient and the reliability index corresponding to the standard structure parameter, and when the structural reliability of the power transmission tower is evaluated each time subsequently, complex parameter calculation is not needed, and only the structural parameter of the to-be-evaluated power transmission tower needs to be input into the trained deep neural network model, so that the reliability of the to-be-evaluated power transmission tower is evaluated. Therefore, the reliability index can be calculated by using the reliability correction coefficient and the reliability index corresponding to the standard structure parameter, and the reliability evaluation of the power transmission tower structure can be accurately and quickly realized, so that the accuracy of the reliability evaluation of the power transmission tower structure is improved, and the evaluation complexity is reduced.

Description

Technical Field

[0001] This invention relates to the field of transmission tower structure technology, and in particular to a method and system for assessing the structural reliability of transmission towers. Background Technology

[0002] As the "backbone" of power transmission, the reliability of transmission towers directly determines the safe and stable operation of the power grid. However, traditional methods for assessing the structural reliability of transmission towers face numerous challenges. They require comprehensive consideration of structural mechanics, material properties, and complex environmental loads such as wind and ice. These parameters not only exhibit uncertainties but are also interconnected, making the assessment process extremely cumbersome. This not only consumes significant time and computational resources but also makes it difficult to accurately assess the probability of transmission tower structural failure, failing to meet the demands of modern power engineering for efficient and precise calculations.

[0003] Currently, there are various methods for assessing the reliability of transmission tower structures, such as the first-order second-moment method, the improved first-order second-moment method, Monte Carlo (MC) simulation, the checkpoint (JC) method, and the response surface methodology. Among these, the first-order second-moment method is widely used. It requires determining the structural function value Z, i.e., Z = RS, where R and S are the structural resistance and effects, both functions of multiple random variables. However, due to the large number of members in the truss structure of transmission towers and their complex nonlinear characteristics, the resistance R and effects S are complex functions with multiple parameters, making them difficult to express with simple explicit formulas. Therefore, the reliability assessment of statically indeterminate structures is usually based on the reliability assessment of the members. The minimum reliability of the members can be used as the reliability index of the structural system, or the reliability of the structural system can be obtained by combining failure modes and using series and parallel relationships. However, the former approach is relatively coarse and has low accuracy, while the latter involves the complex failure mode discrimination of statically indeterminate structures, which is highly complex. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for assessing the structural reliability of transmission towers, which can improve the accuracy of the assessment of the structural reliability of transmission towers and reduce the assessment complexity.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for assessing the structural reliability of transmission towers, comprising the following steps: A reliability database for transmission towers is established, which includes standard structural parameters of transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. The deep neural network model is trained based on the changing structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model. Obtain the structural parameters of the transmission tower to be evaluated; The structural parameters are input into the trained deep neural network model, and the reliability correction coefficient of the transmission tower to be evaluated is output. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the transmission tower to be evaluated and the reliability index corresponding to the standard structural parameters.

[0006] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A structural reliability assessment system for transmission towers includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it performs the following steps: A reliability database for transmission towers is established, which includes standard structural parameters of transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. The deep neural network model is trained based on the changing structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model. Obtain the structural parameters of the transmission tower to be evaluated; The structural parameters are input into the trained deep neural network model, and the reliability correction coefficient of the transmission tower to be evaluated is output. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the transmission tower to be evaluated and the reliability index corresponding to the standard structural parameters.

[0007] The beneficial effects of this invention are as follows: A reliability database for transmission towers is established. A deep neural network model is trained based on the varying structural parameters and their corresponding reliability correction coefficients in the database, resulting in a trained deep neural network model. The structural parameters of the transmission tower to be evaluated are input into the trained deep neural network model, which outputs the reliability correction coefficients. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficients and the reliability indices corresponding to the standard structural parameters. By using a pre-built reliability database to train the deep neural network model, subsequent assessments of the transmission tower's structural reliability do not require complex parameter calculations. Simply inputting the structural parameters of the transmission tower to be evaluated into the trained deep neural network model yields the reliability correction coefficients. The reliability index of the transmission tower to be evaluated can then be calculated using the reliability correction coefficients and the reliability indices corresponding to the standard structural parameters. By combining the reliability database with the deep neural network model, the reliability assessment of transmission tower structures can be achieved accurately and quickly, thereby improving the accuracy of transmission tower structural reliability assessment and reducing the assessment complexity. Attached Figure Description

[0008] Figure 1This is a flowchart illustrating a method for assessing the structural reliability of a transmission tower according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a structural reliability assessment system for a transmission tower according to an embodiment of the present invention; Figure 3 This is a schematic diagram of a deep neural network model structure in a method for assessing the structural reliability of a transmission tower according to an embodiment of the present invention. Detailed Implementation

[0009] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0010] Before detailing the embodiments of this application, some related concepts will first be explained: Deep neural networks: an artificial neural network model with multiple hidden layers that extracts and abstracts data features layer by layer to achieve complex task processing; Prior knowledge in mechanics: refers to the universal laws and principles obtained in the field of mechanics through rational reasoning or experimental summary; Monte Carlo simulation: a numerical computation method based on probability and statistics that simulates the behavior of complex systems through random sampling and statistical analysis. It is named after the casino in Monaco because its core idea is to simulate actual physical processes through probabilistic models. Finite element model: A model established using the finite element analysis method is a combination of elements that are connected only at the nodes, transmit forces only through the nodes, and are constrained only at the nodes.

[0011] In existing technologies, the first-order second-moment method is widely used. This requires determining the structural function value Z, i.e., Z = RS, where R and S represent the structural resistance and effects, both functions of multiple random variables. However, due to the large number of members in the truss structure of transmission towers and its complex nonlinear characteristics, the resistance R and effects S are complex functions with multiple parameters, making them difficult to express with simple explicit formulas. Therefore, the reliability assessment of statically indeterminate structures is usually based on the reliability assessment of the members. The minimum member reliability can be used as the structural system reliability index, or the reliability of the structural system can be obtained by combining failure modes and using series-parallel relationships. However, the former approach is relatively coarse and inaccurate, while the latter involves the complex failure mode discrimination of statically indeterminate structures, resulting in high complexity.

[0012] To at least solve the above problems, please refer to Figure 1 This invention provides a method for assessing the structural reliability of transmission towers, comprising the following steps: A reliability database for transmission towers is established, which includes standard structural parameters of transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. The deep neural network model is trained based on the changing structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model. Obtain the structural parameters of the transmission tower to be evaluated; The structural parameters are input into the trained deep neural network model, and the reliability correction coefficient of the transmission tower to be evaluated is output. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the transmission tower to be evaluated and the reliability index corresponding to the standard structural parameters.

[0013] As can be seen from the above description, the beneficial effects of this invention are as follows: A reliability database for transmission towers is established; a deep neural network model is trained based on the varying structural parameters and their corresponding reliability correction coefficients in the database to obtain a trained deep neural network model; the structural parameters of the transmission tower to be evaluated are input into the trained deep neural network model, which outputs the reliability correction coefficients; and the reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficients and the reliability indexes corresponding to the standard structural parameters. By using the pre-built reliability database to train the deep neural network model, subsequent assessments of the transmission tower's structural reliability do not require complex parameter calculations. Simply inputting the structural parameters of the transmission tower to be evaluated into the trained deep neural network model yields the reliability correction coefficients. The reliability index of the transmission tower to be evaluated can then be calculated using the reliability correction coefficients and the reliability indexes corresponding to the standard structural parameters. By combining the reliability database with the deep neural network model, the reliability assessment of transmission tower structures can be achieved accurately and quickly, thereby improving the accuracy of transmission tower structural reliability assessment and reducing the assessment complexity.

[0014] Furthermore, the deep neural network model is trained based on the changed structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model, which includes: The deep neural network model includes an input layer, hidden layers, a feature separation layer, a grouped feature attention layer, and an output layer; The changed structural parameters are input into the input layer as input layer nodes; Based on prior mechanical knowledge, the input layer nodes are grouped according to the correlation between the input layer nodes and the reliability correction coefficients corresponding to the changing structural parameters, resulting in grouped input layer nodes. The grouped input layer nodes are sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer for training to obtain the trained deep neural network model.

[0015] As described above, the deep neural network model includes an input layer, a hidden layer, a feature separation layer, a grouped feature attention layer, and an output layer. Based on prior knowledge of mechanics, the input layer nodes are grouped according to the correlation between the reliability correction coefficients corresponding to the input layer nodes and the changing structural parameters. The grouped input layer nodes are then sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer for training. By combining prior knowledge of mechanics and introducing an attention mechanism and a grouped network structure, irrelevant connections are cut off to the maximum extent, which can effectively reduce the sample requirement and avoid the overfitting phenomenon caused by the small number of samples in the existing fully connected deep neural network structure. This improves the model training efficiency and ensures that the trained deep neural network model can output more accurate reliability correction coefficients.

[0016] Furthermore, establishing a reliability database for transmission towers includes: Obtain the standard structural parameters of the transmission tower; Multiple sets of standard structural parameter samples were generated using Monte Carlo simulation based on the aforementioned standard structural parameters. A finite element model of the first transmission tower was established based on the aforementioned multiple sets of standard structural parameter samples. The reliability index corresponding to the standard structural parameters is calculated based on the finite element model of the first transmission tower. Generate varying structural parameters based on the standard structural parameters; Based on the aforementioned changing structural parameters, Monte Carlo simulation is used to generate multiple sets of samples of changing structural parameters; A finite element model of the second transmission tower was established based on the aforementioned multiple sets of varying structural parameter samples. The reliability index corresponding to the changing structural parameters is calculated based on the finite element model of the second transmission tower; Calculate the reliability correction coefficient corresponding to the changed structural parameters based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters; A reliability database is generated based on the standard structural parameters, the reliability index corresponding to the standard structural parameters, the variable structural parameters, and the reliability index and reliability correction coefficient corresponding to the variable structural parameters.

[0017] As described above, Monte Carlo simulation performs well in solving reliability calculation problems of nonlinear and large complex structures. It is easy to implement. When establishing a reliability database for transmission towers, Monte Carlo simulation combined with finite element analysis can be used to calculate the reliability index corresponding to the standard structural parameters, the reliability index corresponding to the variable structural parameters, and the reliability correction coefficient corresponding to the variable structural parameters. This can accurately quantify the impact of structural parameters on reliability indexes.

[0018] Furthermore, the reliability index corresponding to the standard structural parameters calculated based on the finite element model of the first transmission tower includes: Based on the finite element model of the first transmission tower, the structural response under each standard structural parameter sample was calculated to obtain multiple sets of structural responses; Calculate the first functional function value under the serviceability limit state and the second functional function value under the bearing capacity limit state based on the multiple sets of structural responses; Calculate the first mean and first variance of the first function value; The first reliability index of the transmission tower structure under normal service limit state under standard structural parameters is calculated based on the first mean and the first variance. Calculate the second mean and second variance of the second function value; The second reliability index of the ultimate bearing capacity of the transmission tower structure under standard structural parameters is calculated based on the second mean and the second variance.

[0019] As described above, calculating the first reliability index of the serviceability limit state of the transmission tower structure under standard structural parameters and the second reliability index of the ultimate limit state of the transmission tower structure under standard structural parameters can evaluate the reliability of the transmission tower structure from different perspectives (serviceability limit state and ultimate limit state of the ultimate limit state), thus improving the comprehensiveness and accuracy of the reliability assessment of the transmission tower structure.

[0020] Further, based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the varied structural parameters, a reliability correction coefficient corresponding to the varied structural parameters is calculated, specifically as follows: ; ; In the formula, The first reliability correction factor represents the serviceability limit state of the transmission tower structure under varying structural parameters. The first reliability index representing the serviceability limit state of a transmission tower structure under varying structural parameters. The first reliability index representing the serviceability limit state of a transmission tower structure under standard structural parameters. m This indicates the number of samples with varying structural parameters. The second reliability correction factor represents the ultimate limit state of the load-bearing capacity of the transmission tower structure under varying structural parameters. The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under varying structural parameters. The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under standard structural parameter conditions.

[0021] As described above, the reliability correction coefficient corresponding to the changed structural parameters is calculated based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters. The reliability correction coefficient corresponding to the changed structural parameters obtained by calculation can facilitate the training of subsequent deep neural network models.

[0022] Further, the reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the tower and the reliability index corresponding to the standard structural parameters, specifically as follows: ; ; In the formula, The first reliability index represents the structural serviceability limit state of the transmission tower to be evaluated. The second reliability index represents the ultimate reliability state of the structural bearing capacity of the transmission tower to be evaluated. This represents the first reliability correction factor for the serviceability limit state of the transmission tower under normal structural conditions. The second reliability correction factor represents the ultimate limit state of the structural bearing capacity of the transmission tower to be evaluated.

[0023] As described above, when calculating the reliability index of a transmission tower, it is only necessary to multiply the reliability correction coefficient output by the model by the reliability index corresponding to the standard structural parameters to obtain the reliability index of the serviceability limit state and the ultimate limit state of the structural bearing capacity of the transmission tower to be evaluated. The evaluation process is more efficient and accurate.

[0024] Please refer to Figure 2 Another embodiment of the present invention provides a structural reliability assessment system for transmission towers, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it performs the following steps: A reliability database for transmission towers is established, which includes standard structural parameters of transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. The deep neural network model is trained based on the changing structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model. Obtain the structural parameters of the transmission tower to be evaluated; The structural parameters are input into the trained deep neural network model, and the reliability correction coefficient of the transmission tower to be evaluated is output. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the transmission tower to be evaluated and the reliability index corresponding to the standard structural parameters.

[0025] As can be seen from the above description, the beneficial effects of this invention are as follows: A reliability database for transmission towers is established; a deep neural network model is trained based on the varying structural parameters and their corresponding reliability correction coefficients in the database to obtain a trained deep neural network model; the structural parameters of the transmission tower to be evaluated are input into the trained deep neural network model, which outputs the reliability correction coefficients; and the reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficients and the reliability indexes corresponding to the standard structural parameters. By using the pre-built reliability database to train the deep neural network model, subsequent assessments of the transmission tower's structural reliability do not require complex parameter calculations. Simply inputting the structural parameters of the transmission tower to be evaluated into the trained deep neural network model yields the reliability correction coefficients. The reliability index of the transmission tower to be evaluated can then be calculated using the reliability correction coefficients and the reliability indexes corresponding to the standard structural parameters. By combining the reliability database with the deep neural network model, the reliability assessment of transmission tower structures can be achieved accurately and quickly, thereby improving the accuracy of transmission tower structural reliability assessment and reducing the assessment complexity.

[0026] Furthermore, the deep neural network model is trained based on the changed structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model, which includes: The deep neural network model includes an input layer, hidden layers, a feature separation layer, a grouped feature attention layer, and an output layer; The changed structural parameters are input into the input layer as input layer nodes; Based on prior mechanical knowledge, the input layer nodes are grouped according to the correlation between the input layer nodes and the reliability correction coefficients corresponding to the changing structural parameters, resulting in grouped input layer nodes. The grouped input layer nodes are sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer for training to obtain the trained deep neural network model.

[0027] As described above, the deep neural network model includes an input layer, a hidden layer, a feature separation layer, a grouped feature attention layer, and an output layer. Based on prior knowledge of mechanics, the input layer nodes are grouped according to the correlation between the reliability correction coefficients corresponding to the input layer nodes and the changing structural parameters. The grouped input layer nodes are then sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer for training. By combining prior knowledge of mechanics and introducing an attention mechanism and a grouped network structure, irrelevant connections are cut off to the maximum extent, which can effectively reduce the sample requirement and avoid the overfitting phenomenon caused by the small number of samples in the existing fully connected deep neural network structure. This improves the model training efficiency and ensures that the trained deep neural network model can output more accurate reliability correction coefficients.

[0028] Furthermore, establishing a reliability database for transmission towers includes: Obtain the standard structural parameters of the transmission tower; Multiple sets of standard structural parameter samples were generated using Monte Carlo simulation based on the aforementioned standard structural parameters. A finite element model of the first transmission tower was established based on the aforementioned multiple sets of standard structural parameter samples. The reliability index corresponding to the standard structural parameters is calculated based on the finite element model of the first transmission tower. Generate varying structural parameters based on the standard structural parameters; Based on the aforementioned changing structural parameters, Monte Carlo simulation is used to generate multiple sets of samples of changing structural parameters; A finite element model of the second transmission tower was established based on the aforementioned multiple sets of varying structural parameter samples. The reliability index corresponding to the changing structural parameters is calculated based on the finite element model of the second transmission tower; Calculate the reliability correction coefficient corresponding to the changed structural parameters based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters; A reliability database is generated based on the standard structural parameters, the reliability index corresponding to the standard structural parameters, the variable structural parameters, and the reliability index and reliability correction coefficient corresponding to the variable structural parameters.

[0029] As described above, Monte Carlo simulation performs well in solving reliability calculation problems of nonlinear and large complex structures. It is easy to implement. When establishing a reliability database for transmission towers, Monte Carlo simulation combined with finite element analysis can be used to calculate the reliability index corresponding to the standard structural parameters, the reliability index corresponding to the variable structural parameters, and the reliability correction coefficient corresponding to the variable structural parameters. This can accurately quantify the impact of structural parameters on reliability indexes.

[0030] Furthermore, the reliability index corresponding to the standard structural parameters calculated based on the finite element model of the first transmission tower includes: Based on the finite element model of the first transmission tower, the structural response under each standard structural parameter sample was calculated to obtain multiple sets of structural responses; Calculate the first functional function value under the serviceability limit state and the second functional function value under the bearing capacity limit state based on the multiple sets of structural responses; Calculate the first mean and first variance of the first function value; The first reliability index of the transmission tower structure under normal service limit state under standard structural parameters is calculated based on the first mean and the first variance. Calculate the second mean and second variance of the second function value; The second reliability index of the ultimate bearing capacity of the transmission tower structure under standard structural parameters is calculated based on the second mean and the second variance.

[0031] As described above, calculating the first reliability index of the serviceability limit state of the transmission tower structure under standard structural parameters and the second reliability index of the ultimate limit state of the transmission tower structure under standard structural parameters can evaluate the reliability of the transmission tower structure from different perspectives (serviceability limit state and ultimate limit state of the ultimate limit state), thus improving the comprehensiveness and accuracy of the reliability assessment of the transmission tower structure.

[0032] The structural reliability assessment method and system for transmission towers described above are applicable to structural reliability assessment scenarios for transmission towers. The following detailed embodiments illustrate these methods: Please refer to Figure 1 One embodiment of the present invention is as follows: A method for assessing the structural reliability of transmission towers, comprising the following steps: S1. Establish a reliability database for transmission towers. This database includes standard structural parameters of the transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. Specifically, it includes S11-S110: The structural reliability of transmission towers is related to both structural resistance and load effects. Structural resistance is related to the tower's structural form and material parameters. On the other hand, structural effects are related to structural loads. The magnitude and form of loads are diverse, and can be related to the transmission tower, power lines, suspension points, icing, wind loads, etc. The load effects caused by different loads exhibit certain patterns. Therefore, structural parameters, including the horizontal span utilization rate K, are crucial. h Vertical gap utilization rate K v Wind load return period T (unit: years), self-weight load variation coefficient at hanging point K gMaterial strength variation coefficient C y Material elastic modulus variation coefficient C E The coefficient of variation of constant load C g and live load variation coefficient C q The standard structural parameters are those with standard values, while the variable structural parameters are those that vary randomly and are not standard values.

[0033] S11. Obtain the standard structural parameters of the transmission tower.

[0034] S12. Based on the standard structural parameters, use Monte Carlo simulation to generate multiple sets of standard structural parameter samples.

[0035] Specifically, standard structural parameters are input into the finite element model, and materials and loads are considered as random distribution functions. Monte Carlo simulation sampling is then used to obtain... n Sample of standard structural parameters for a group.

[0036] S13. Establish the first transmission tower finite element model based on the multiple sets of standard structural parameter samples.

[0037] S14. Calculate the reliability index corresponding to the standard structural parameters based on the finite element model of the first transmission tower, specifically including S141-S146: S141. Based on the finite element model of the first transmission tower, calculate the structural response under each standard structural parameter sample to obtain multiple sets of structural responses.

[0038] S142. Calculate the first functional function value under the serviceability limit state and the second functional function value under the load-bearing capacity limit state based on the multiple sets of structural responses.

[0039] Wherein, the value of the first function is: Z 1,k =R 1,k -S 1,k ; In the formula, Z 1,k R represents the first function value under normal serviceability limit state conditions, given the standard structural parameters of the transmission tower. 1,k S represents the allowable displacement at the top of the transmission tower under standard structural parameters. 1,k This indicates the maximum displacement of the top of the transmission tower under the standard structural parameters.

[0040] The value of the second function is: Z 2,k =R 2,k -S 2,k ; In the formula, Z 2,kR represents the value of the second function under the ultimate limit state of bearing capacity under the standard structural parameters of the transmission tower. 2,k S represents the ultimate strength stress of the material under standard structural parameters of the transmission tower. 2,k This represents the maximum stress value of components in a transmission tower under standard structural parameters.

[0041] S143. Calculate the first mean and the first variance of the first functional value.

[0042] S144. Calculate the first reliability index of the transmission tower structure under normal serviceability limit state under standard structural parameters based on the first mean and the first variance. Specifically: ; In the formula, The first reliability index representing the serviceability limit state of a transmission tower structure under standard structural parameters. This represents the first mean. This represents the first variance.

[0043] S145. Calculate the second mean and second variance of the second function value.

[0044] S146. Calculate the second reliability index of the ultimate limit state of the transmission tower structure under standard structural parameters based on the second mean and the second variance. Specifically: ; In the formula, The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under standard structural parameter conditions. This represents the second mean. This represents the second variance.

[0045] S15. Generate variable structural parameters based on the standard structural parameters.

[0046] For example, the standard values ​​for the horizontal and vertical span utilization rates of transmission towers are 0.80, and based on this uniform variation, the range is 0.65-1.10. The return period of wind load is a discrete variable with a standard value of 50 years, and based on this, variation values ​​such as 30 years, 50 years, and 100 years are generated. The standard value for the self-weight load variation coefficient of the mounting point is 1.0, and based on this variation, the range is 0.5-1.5. The standard value for the coefficient of variation of material strength is 0.05, and based on this variation, the range is 0.05-0.20. The standard value for the coefficient of variation of material elastic modulus is 0.15, and based on this variation, the range is 0.05-0.20. The standard value for the coefficient of variation of dead load is 0.15, and based on this variation, the range is 0.05-0.20. The standard value for the coefficient of variation of live load is 0.15, and based on this variation, the range is 0.05-0.20.

[0047] S16. Based on the changed structural parameters, use Monte Carlo simulation to generate multiple sets of changed structural parameter samples.

[0048] S17. Establish a finite element model of the second transmission tower based on the multiple sets of varying structural parameter samples.

[0049] S18. Calculate the reliability index corresponding to the changing structural parameters based on the finite element model of the second transmission tower.

[0050] The reliability indices corresponding to the changing structural parameters include a first reliability index for the normal service limit state of the transmission tower structure under changing structural parameters and a second reliability index for the ultimate load-bearing capacity state of the transmission tower structure under changing structural parameters.

[0051] The calculation method for the reliability index corresponding to the changed structural parameters is the same as that for the reliability index corresponding to the standard structural parameters, and will not be repeated here.

[0052] S19. Calculate the reliability correction coefficient corresponding to the changed structural parameters based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters, specifically as follows: ; ; In the formula, The first reliability correction factor represents the serviceability limit state of the transmission tower structure under varying structural parameters. The first reliability index representing the serviceability limit state of a transmission tower structure under varying structural parameters. The first reliability index representing the serviceability limit state of a transmission tower structure under standard structural parameters. m This indicates the number of samples with varying structural parameters. The second reliability correction factor represents the ultimate limit state of the load-bearing capacity of the transmission tower structure under varying structural parameters. The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under varying structural parameters. The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under standard structural parameter conditions.

[0053] S110. Generate a reliability database based on the standard structural parameters, the reliability index corresponding to the standard structural parameters, the variable structural parameters, and the reliability index and reliability correction coefficient corresponding to the variable structural parameters.

[0054] As shown in Table 1, Table 1 presents a reliability database.

[0055] Table 1 Reliability Database

[0056] S2. The deep neural network model is trained based on the changed structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model, specifically including S21-S23: The deep neural network model includes an input layer, a hidden layer, a feature separation layer, a grouped feature attention layer, and an output layer.

[0057] S21. Input the changed structural parameters into the input layer as input layer nodes.

[0058] Specifically, considering that the return period T of wind load is a discrete variable, it is processed using one-hot encoding before input, while the other parameters are continuous variables and can be input directly.

[0059] S22. Based on prior mechanical knowledge, the input layer nodes are grouped according to the correlation between the input layer nodes and the reliability correction coefficients corresponding to the changing structural parameters to obtain the grouped input layer nodes.

[0060] For example, consider structural parameters T and C E and C q The horizontal displacement of the transmission tower structure has a significant impact, which in turn affects the reliability correction factor for the normal serviceability limit state. Strong correlation, structural parameter K g K h K v C y C g The effect on the ultimate limit state of structural bearing capacity is more significant. In addition, the four coefficients of variation C y C E C g C q The impact on the reliability index is also significant. Therefore, the input layer nodes are divided into four groups, and then sequentially pass through the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer to establish a structure as follows: Figure 3 The diagram shows a grouped deep neural network structure.

[0061] S23. The grouped input layer nodes are sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer and the output layer for training to obtain the trained deep neural network model.

[0062] Specifically, the deep neural network model structure is trained using a reliability database formed by the changing structural parameters and the reliability correction coefficients corresponding to the changing structural parameters, to obtain the trained deep neural network model.

[0063] Among them, such as Figure 3 As shown, the hidden layer contains 16 nodes, which connect the grouped input layer nodes accordingly. The feature separation layer contains reliability correction coefficients for four normal use limit states. Reliability correction coefficients for dedicated nodes, 4 shared characteristic nodes, and 4 ultimate capacity states. Dedicated nodes. The grouped feature attention layer contains 3 nodes, namely... Exclusive weights, shared feature weights, Dedicated weights. The grouped feature attention layer and the output layer (node) , Fully connected, considering the different effects of each set of parameters on the output node, sets strong and weak connections and assigns different initial weights. This utilizes prior knowledge to strengthen the role of strong dependencies, assigning different focuses to different input nodes, further ensuring the effectiveness of the deep neural network. In specific implementation, the initial value of the weight of the strong dependency path can be set to 0.5-1.0, and the initial value of the weight of the weak dependency path can be set to 0-0.1.

[0064] In practical applications, each type of transmission tower can be processed according to S1-S2. This method is applicable regardless of which type of transmission tower needs to be assessed for structural reliability.

[0065] S3. Obtain the structural parameters of the transmission tower to be evaluated.

[0066] S4. Input the structural parameters into the trained deep neural network model and output the reliability correction coefficient of the transmission tower to be evaluated.

[0067] S5. Calculate the reliability index of the transmission tower to be evaluated based on the reliability correction coefficient of the tower and the reliability index corresponding to the standard structural parameters, specifically as follows: ; ; In the formula, The first reliability index represents the structural serviceability limit state of the transmission tower to be evaluated. The second reliability index represents the ultimate reliability state of the structural bearing capacity of the transmission tower to be evaluated. This represents the first reliability correction factor for the serviceability limit state of the transmission tower under normal structural conditions. The second reliability correction factor represents the ultimate limit state of the structural bearing capacity of the transmission tower to be evaluated.

[0068] In summary, the structural reliability assessment method for transmission towers described above utilizes a pre-built reliability database to train a deep neural network model. Subsequent assessments of the transmission tower's structural reliability do not require complex parameter calculations; simply inputting the structural parameters of the transmission tower to be assessed into the trained deep neural network model yields a reliability correction coefficient. Using this correction coefficient and the reliability index corresponding to the standard structural parameters, the reliability index of the transmission tower to be assessed can be calculated. By combining the reliability database with the deep neural network model, the reliability assessment of transmission tower structures can be achieved accurately and quickly, thereby improving the accuracy of the assessment and reducing its complexity. Furthermore, the deep neural network model includes an input layer, hidden layers, a feature separation layer, a grouped feature attention layer, and an output layer, based on prior mechanical knowledge... The correlation between input layer nodes and reliability correction coefficients corresponding to varying structural parameters is analyzed by grouping the input layer nodes. These grouped nodes are then sequentially fed into the hidden layer, feature separation layer, grouped feature attention layer, and output layer for training. By incorporating prior mechanical knowledge, an attention mechanism, and a grouped network structure, irrelevant connections are minimized, effectively reducing sample requirements and avoiding overfitting issues caused by insufficient samples in existing fully connected deep neural network structures. This improves model training efficiency and ensures that the trained deep neural network model outputs more accurate reliability correction coefficients. Furthermore, Monte Carlo simulation combined with finite element analysis is used to calculate the reliability indices corresponding to standard structural parameters, varying structural parameters, and their corresponding reliability correction coefficients for transmission towers. This accurately quantifies the impact of structural parameters on reliability indices.

[0069] According to another aspect of the invention, Figure 2 This is a schematic diagram illustrating a structural reliability assessment system for a transmission tower according to an embodiment of the present invention. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the various steps of the structural reliability assessment method for transmission towers as described above.

[0070] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for assessing the structural reliability of transmission towers, characterized in that, Including the following steps: A reliability database for transmission towers is established, which includes standard structural parameters of transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. The deep neural network model is trained based on the changing structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model. Obtain the structural parameters of the transmission tower to be evaluated; The structural parameters are input into the trained deep neural network model, and the reliability correction coefficient of the transmission tower to be evaluated is output. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the transmission tower to be evaluated and the reliability index corresponding to the standard structural parameters.

2. The structural reliability assessment method for transmission towers according to claim 1, characterized in that, The deep neural network model is trained based on the changed structural parameters and their corresponding reliability correction coefficients, resulting in the following trained deep neural network model: The deep neural network model includes an input layer, hidden layers, a feature separation layer, a grouped feature attention layer, and an output layer; The changed structural parameters are input into the input layer as input layer nodes; Based on prior mechanical knowledge, the input layer nodes are grouped according to the correlation between the input layer nodes and the reliability correction coefficients corresponding to the changing structural parameters, resulting in grouped input layer nodes. The grouped input layer nodes are sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer for training to obtain the trained deep neural network model.

3. The structural reliability assessment method for transmission towers according to claim 1, characterized in that, Establishing a reliability database for transmission towers includes: Obtain the standard structural parameters of the transmission tower; Multiple sets of standard structural parameter samples were generated using Monte Carlo simulation based on the aforementioned standard structural parameters. A finite element model of the first transmission tower was established based on the aforementioned multiple sets of standard structural parameter samples. The reliability index corresponding to the standard structural parameters is calculated based on the finite element model of the first transmission tower. Generate varying structural parameters based on the standard structural parameters; Based on the aforementioned changing structural parameters, Monte Carlo simulation is used to generate multiple sets of samples of changing structural parameters; A finite element model of the second transmission tower was established based on the aforementioned multiple sets of varying structural parameter samples. The reliability index corresponding to the changing structural parameters is calculated based on the finite element model of the second transmission tower; Calculate the reliability correction coefficient corresponding to the changed structural parameters based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters; A reliability database is generated based on the standard structural parameters, the reliability index corresponding to the standard structural parameters, the variable structural parameters, and the reliability index and reliability correction coefficient corresponding to the variable structural parameters.

4. The structural reliability assessment method for a transmission tower according to claim 3, characterized in that, The reliability indices corresponding to the standard structural parameters calculated based on the finite element model of the first transmission tower include: Based on the finite element model of the first transmission tower, the structural response under each standard structural parameter sample was calculated to obtain multiple sets of structural responses; Calculate the first functional function value under the serviceability limit state and the second functional function value under the bearing capacity limit state based on the multiple sets of structural responses; Calculate the first mean and first variance of the first function value; The first reliability index of the transmission tower structure under normal service limit state under standard structural parameters is calculated based on the first mean and the first variance. Calculate the second mean and second variance of the second function value; The second reliability index of the ultimate bearing capacity of the transmission tower structure under standard structural parameters is calculated based on the second mean and the second variance.

5. The structural reliability assessment method for a transmission tower according to claim 3, characterized in that, The reliability correction coefficient for the changed structural parameters is calculated based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters, specifically as follows: ； ； In the formula, The first reliability correction factor represents the serviceability limit state of the transmission tower structure under varying structural parameters. The first reliability index representing the serviceability limit state of a transmission tower structure under varying structural parameters. The first reliability index representing the serviceability limit state of a transmission tower structure under standard structural parameters. m This indicates the number of samples with varying structural parameters. The second reliability correction factor represents the ultimate limit state of the load-bearing capacity of the transmission tower structure under varying structural parameters. The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under varying structural parameters. The second reliability index represents the ultimate limit state of the load-bearing capacity of a transmission tower structure under standard structural parameter conditions.

6. The structural reliability assessment method for a transmission tower according to claim 5, characterized in that, The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the tower and the reliability index corresponding to the standard structural parameters, specifically as follows: ； ； In the formula, The first reliability index represents the serviceability limit state of the transmission tower structure to be evaluated. The second reliability index represents the ultimate reliability state of the structural bearing capacity of the transmission tower to be evaluated. The first reliability correction factor represents the structural serviceability limit state of the transmission tower to be evaluated. The second reliability correction factor represents the ultimate limit state of the structural bearing capacity of the transmission tower to be evaluated.

7. A structural reliability assessment system for transmission towers, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it performs the following steps: A reliability database for transmission towers is established, which includes standard structural parameters of transmission towers and their corresponding reliability indices, as well as variable structural parameters and their corresponding reliability correction coefficients. The deep neural network model is trained based on the changing structural parameters and their corresponding reliability correction coefficients to obtain the trained deep neural network model. Obtain the structural parameters of the transmission tower to be evaluated; The structural parameters are input into the trained deep neural network model, and the reliability correction coefficient of the transmission tower to be evaluated is output. The reliability index of the transmission tower to be evaluated is calculated based on the reliability correction coefficient of the transmission tower to be evaluated and the reliability index corresponding to the standard structural parameters.

8. The structural reliability assessment system for transmission towers according to claim 7, characterized in that, The deep neural network model is trained based on the changed structural parameters and their corresponding reliability correction coefficients, resulting in the following trained deep neural network model: The deep neural network model includes an input layer, hidden layers, a feature separation layer, a grouped feature attention layer, and an output layer; The changed structural parameters are input into the input layer as input layer nodes; Based on prior mechanical knowledge, the input layer nodes are grouped according to the correlation between the input layer nodes and the reliability correction coefficients corresponding to the changing structural parameters, resulting in grouped input layer nodes. The grouped input layer nodes are sequentially input into the hidden layer, the feature separation layer, the grouped feature attention layer, and the output layer for training to obtain the trained deep neural network model.

9. The structural reliability assessment system for transmission towers according to claim 7, characterized in that, Establishing a reliability database for transmission towers includes: Obtain the standard structural parameters of the transmission tower; Multiple sets of standard structural parameter samples were generated using Monte Carlo simulation based on the aforementioned standard structural parameters. A finite element model of the first transmission tower was established based on the aforementioned multiple sets of standard structural parameter samples. The reliability index corresponding to the standard structural parameters is calculated based on the finite element model of the first transmission tower. Generate varying structural parameters based on the standard structural parameters; Based on the aforementioned changing structural parameters, Monte Carlo simulation is used to generate multiple sets of samples of changing structural parameters; A finite element model of the second transmission tower was established based on the aforementioned multiple sets of varying structural parameter samples. The reliability index corresponding to the changing structural parameters is calculated based on the finite element model of the second transmission tower; Calculate the reliability correction coefficient corresponding to the changed structural parameters based on the reliability index corresponding to the standard structural parameters and the reliability index corresponding to the changed structural parameters; A reliability database is generated based on the standard structural parameters, the reliability index corresponding to the standard structural parameters, the variable structural parameters, and the reliability index and reliability correction coefficient corresponding to the variable structural parameters.

10. The structural reliability assessment system for transmission towers according to claim 9, characterized in that, The reliability indices corresponding to the standard structural parameters calculated based on the finite element model of the first transmission tower include: Based on the finite element model of the first transmission tower, the structural response under each standard structural parameter sample was calculated to obtain multiple sets of structural responses; Calculate the first functional function value under the serviceability limit state and the second functional function value under the bearing capacity limit state based on the multiple sets of structural responses; Calculate the first mean and first variance of the first function value; The first reliability index of the transmission tower structure under normal service limit state under standard structural parameters is calculated based on the first mean and the first variance. Calculate the second mean and second variance of the second function value; The second reliability index of the ultimate bearing capacity of the transmission tower structure under standard structural parameters is calculated based on the second mean and the second variance.

Images