SOM zoning method, device, equipment and medium considering wind pressure correlation
Through the combination of wind tunnel test and SOM neural network, the wind pressure correlation coefficient is processed and the neuron position is adaptively adjusted, which solves the problem of excessive classification of stroke pressure partitions in structural wind-resistant design and discrete results, and achieves the accuracy of wind pressure partitions An intuitive display of improvements and results.
Patent Information
- Application Number
- CN202411330434.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-09-24
AI Technical Summary
The existing technology has problems such as excessive classification conditions in structural wind resistance design, large workload and discrete results in finding optimal classification results, and it is difficult to effectively use the self-organized mapping (SOM) algorithm for wind pressure partitioning.
The wind pressure time course of the structural surface is obtained through wind tunnel tests, the original wind pressure coefficient data set is constructed, the wind pressure correlation coefficient is processed, the optimal classification number is determined, the wind pressure partition is used using SOM neural network, and the neuron position is adaptively adjusted to match the wind Pressure distribution characteristics.
It realizes the rationality and accuracy of wind pressure partitioning results, quickly and accurately determines the optimal classification number, eliminates discrete partition points, and intuitively displays the wind pressure partitioning situation, which is suitable for high-dimensional data processing and dimensional reduction analysis.
Smart Images

Figure CN118885815B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of structural surface wind pressure zoning, and in particular to a SOM zoning method, device, equipment and medium taking into account wind pressure correlation. Background Art
[0002] In the process of wind-resistant structural design, it is usually necessary to divide the surface wind load of the structure into zones. For example, in the Code for Loads on Building Structures GB50009-2012 and ASCE7-2016, the local average wind pressure coefficient and the extreme wind pressure coefficient of the building surface enclosure structure are divided into zones considering the wind load distribution characteristics, and there are also zone requirements for the post-data processing of the wind tunnel test of the building structure. In the wind pressure zoning method, scholars believe that the use of mathematical clustering algorithms can effectively perform wind pressure zoning, aggregate areas with similar wind pressure distribution characteristics together to form a cluster, and thus obtain the zoned wind pressure.
[0003] Traditional clustering methods have the disadvantages of too many classifications, large workload in finding the optimal classification results, and discrete wind pressure zoning results, which bring inconvenience to engineering applications. SOM (Self-Organizing Map) is an unsupervised learning method based on neural networks. It actively finds the laws and essential attributes of sample data, and self-organizes and adaptively changes the connection weights between networks to carry out zoning. By introducing the SOM algorithm in wind pressure zoning and considering the wind pressure correlation between various areas on the building surface, the rationality and accuracy of the zoning results can be improved, providing a more scientific basis for wind-resistant design.
[0004] However, research and application in this field are still in the initial exploration stage. How to effectively apply the SOM algorithm to wind pressure zoning still requires further research and technological breakthroughs. Summary of the invention
[0005] In order to solve the technical problem of how to effectively apply the SOM algorithm to wind pressure zoning in the prior art, the embodiment of the present invention provides a SOM zoning method, device, equipment, and medium that considers wind pressure correlation. The technical solution is as follows:
[0006] On the one hand, a SOM partitioning method considering wind pressure correlation is provided, the method is implemented by a partitioning device, and the method includes:
[0007] S1. Obtain the wind pressure time history of each measuring point on the structure surface through wind tunnel test, and construct the original wind pressure coefficient data set based on the wind pressure time history.
[0008] S2. Construct a new wind pressure coefficient data set based on the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the normalized original wind pressure coefficient data set.
[0009] S3. Determine the optimal number of classifications based on the original wind pressure coefficient data set.
[0010] S4. Based on the new wind pressure coefficient data set, the optimal number of classifications and the SOM method, the SOM wind pressure zoning results that consider the wind pressure correlation in the structural surface wind pressure zoning are obtained.
[0011] Optionally, in S1, the wind pressure time history of each measuring point on the surface of the structure is obtained through a wind tunnel test, and an original wind pressure coefficient data set is constructed according to the wind pressure time history, including:
[0012] S11. Obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and calculate the wind pressure coefficient time history of each measuring point based on the wind pressure time history.
[0013] S12. Calculate the wind pressure correlation coefficient between each measuring point based on the wind pressure coefficient time history.
[0014] S13. Construct an original wind pressure coefficient data set based on the wind pressure correlation coefficient.
[0015] Optionally, in S2, constructing a new wind pressure coefficient data set based on the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the normalized original wind pressure coefficient data set includes:
[0016] S21. Construct a correlation coefficient matrix based on the original wind pressure coefficient data set.
[0017] S22. Calculate the mean of each row of the correlation coefficient matrix to obtain a new vector.
[0018] S23, normalizing the original wind pressure coefficient data set, and forming a new wind pressure coefficient data set according to the normalized original wind pressure coefficient data set and the new vector.
[0019] Optionally, determining the optimal number of classifications according to the original wind pressure coefficient data set in S3 includes:
[0020] S31. Based on the original wind pressure coefficient data set, multiple value, and calculate each The sum of squared errors (SSE) of the values.
[0021] S32. Based on the original wind pressure coefficient data set, calculate the average silhouette coefficient of all measuring points.
[0022] S33, according to Value, each The sum of squared errors (SSE) and the average silhouette coefficient of the values are plotted. Relationship curve and The silhouette coefficient curve is based on Relationship curve and The silhouette coefficient curve determines the optimal number of clusters, where represents the sum of squared errors.
[0023] Optionally, in S4, based on the new wind pressure coefficient data set, the optimal number of classifications, and the SOM method, a SOM wind pressure zoning result that considers wind pressure correlation in the structural surface wind pressure zoning is obtained, including:
[0024] S41, randomly initialize according to the optimal number of classifications The weight vector of the center points.
[0025] S42. Input the wind pressure coefficient in the new wind pressure coefficient data set into the improved SOM neural network model, calculate the Euclidean distance between the wind pressure coefficient in the new wind pressure coefficient data set and the weight vector, and obtain the neuron with the smallest Euclidean distance as the winning neuron.
[0026] S43, updating the neurons in the topological neighborhood of the winning neuron to obtain a renormalized learned weight vector.
[0027] S44, determine whether the change of the weight vector is within a preset change range; if so, execute step S46; if not, execute step S45.
[0028] S45, determine whether the preset number of iterations has been reached; if so, execute step S46; if not, go to execute step S42.
[0029] S46. According to the weight vector, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning.
[0030] Optionally, the calculation formula of the renormalized learned weight vector is as shown in the following formula (1):
[0031] (1)
[0032] In the formula, represents the weight vector after renormalization learning, represents the input dataset, represents the output neuron, represents the initial value of the weight vector, represents the iterative learning rate.
[0033] On the other hand, a SOM partitioning device considering wind pressure correlation is provided, and the device is applied to the SOM partitioning method considering wind pressure correlation, and the device includes:
[0034] The acquisition module is used to obtain the wind pressure time history of each measuring point on the structure surface through wind tunnel tests, and construct the original wind pressure coefficient data set according to the wind pressure time history.
[0035] The construction module is used to construct a new wind pressure coefficient data set based on the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the unitized original wind pressure coefficient data set.
[0036] The determination module is used to determine the optimal number of classifications according to the original wind pressure coefficient data set.
[0037] The output module is used to obtain the SOM wind pressure zoning result considering the wind pressure correlation in the structural surface wind pressure zoning according to the new wind pressure coefficient data set, the optimal classification number and the SOM method.
[0038] Optionally, the acquisition module is further used to:
[0039] S11. Obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and calculate the wind pressure coefficient time history of each measuring point based on the wind pressure time history.
[0040] S12. Calculate the wind pressure correlation coefficient between each measuring point based on the wind pressure coefficient time history.
[0041] S13. Construct an original wind pressure coefficient data set based on the wind pressure correlation coefficient.
[0042] Optionally, the building block is further used to:
[0043] S21. Construct a correlation coefficient matrix based on the original wind pressure coefficient data set.
[0044] S22. Calculate the mean of each row of the correlation coefficient matrix to obtain a new vector.
[0045] S23, normalizing the original wind pressure coefficient data set, and forming a new wind pressure coefficient data set according to the normalized original wind pressure coefficient data set and the new vector.
[0046] Optionally, the determination module is further configured to:
[0047] S31. Based on the original wind pressure coefficient data set, multiple value, and calculate each The sum of squared errors (SSE) of the values.
[0048] S32. Based on the original wind pressure coefficient data set, calculate the average silhouette coefficient of all measuring points.
[0049] S33, according to Value, each The sum of squared errors (SSE) and the average silhouette coefficient of the values are plotted. Relationship curve and The silhouette coefficient curve is based on Relationship curve and The silhouette coefficient curve determines the optimal number of clusters, where represents the sum of squared errors.
[0050] Optionally, the output module is further configured to:
[0051] S41, randomly initialize according to the optimal number of classifications The weight vector of the center points.
[0052] S42. Input the wind pressure coefficient in the new wind pressure coefficient data set into the improved SOM neural network model, calculate the Euclidean distance between the wind pressure coefficient in the new wind pressure coefficient data set and the weight vector, and obtain the neuron with the smallest Euclidean distance as the winning neuron.
[0053] S43, updating the neurons in the topological neighborhood of the winning neuron to obtain a renormalized learned weight vector.
[0054] S44, determine whether the change of the weight vector is within a preset change range; if so, execute step S46; if not, execute step S45.
[0055] S45, determine whether the preset number of iterations has been reached; if so, execute step S46; if not, go to execute step S42.
[0056] S46. According to the weight vector, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning.
[0057] Optionally, the calculation formula of the renormalized learned weight vector is as shown in the following formula (1):
[0058] (1)
[0059] In the formula, represents the weight vector after renormalization learning, represents the input dataset, represents the output neuron, represents the initial value of the weight vector, represents the iterative learning rate.
[0060] On the other hand, a partitioning device is provided, comprising: a processor; a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, any one of the above-mentioned SOM partitioning methods that consider wind pressure correlation is implemented.
[0061] On the other hand, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement any one of the above-mentioned SOM partitioning methods considering wind pressure correlation.
[0062] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:
[0063] In the present invention, a SOM wind pressure zoning method is provided which considers the wind pressure correlation in the building surface wind pressure zoning. and The silhouette coefficient curve can quickly and accurately determine the optimal number of classifications ; The correlation coefficient of each measuring point is introduced, and the processed constraint variable distribution also reflects the wind pressure distribution characteristics. After introducing the constraint variable, the discrete points of the partition can be eliminated.
[0064] Through the competitive learning mechanism, SOM can adaptively adjust the position of neurons to better match the distribution characteristics of input data; through the two-dimensional or three-dimensional mapping results of SOM, the wind pressure zoning situation can be intuitively displayed, making it easier to understand the distribution and clustering results of the data; compared with traditional clustering algorithms, the SOM algorithm is good at processing high-dimensional data and can map complex data sets to low-dimensional space for analysis through dimensionality reduction. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0066] Figure 1 This is a basic flow chart of a SOM partitioning method considering wind pressure correlation provided by an embodiment of the present invention;
[0067] Figure 2 It is a technical flow chart of a SOM partitioning method considering wind pressure correlation provided by an embodiment of the present invention;
[0068] Figure 3 It is a diagram of an improved SOM neural network model provided by an embodiment of the present invention;
[0069] Figure 4 It is a wind tunnel test model and a wind direction angle definition diagram provided by an embodiment of the present invention;
[0070] Figure 5 is a wind pressure coefficient cloud map provided by an embodiment of the present invention;
[0071] Figure 6 The embodiment of the present invention provides Relationship curve and Silhouette coefficient curve plot;
[0072] Figure 7 is a partition result diagram provided by an embodiment of the present invention;
[0073] Figure 8 It is a block diagram of a SOM partitioning device considering wind pressure correlation provided by an embodiment of the present invention;
[0074] Fig. 9 It is a structural schematic diagram of a partitioning device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0075] The technical solution of the present invention is described below in conjunction with the accompanying drawings.
[0076] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.
[0077] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.
[0078] In the embodiments of the present invention, sometimes the subscripts such as W 1 It may be written in non-subscript form such as W1. When the difference is not emphasized, the meaning is the same.
[0079] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.
[0080] The embodiment of the present invention provides a SOM partitioning method considering wind pressure correlation, which can be implemented by a partitioning device, which can be a terminal or a server. Figure 1 , Figure 2The flowchart of the SOM partitioning method considering wind pressure correlation is shown in FIG. 1 . The processing flow of the method may include the following steps:
[0081] S1. Obtain the wind pressure time history of each measuring point on the structure surface through wind tunnel test, and construct the original wind pressure coefficient data set based on the wind pressure time history.
[0082] Optionally, the above step S1 may include the following steps S11-S13:
[0083] S11. Obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and calculate the wind pressure coefficient time history of each measuring point based on the wind pressure time history.
[0084] S12. Calculate the wind pressure correlation coefficient between each measuring point based on the wind pressure coefficient time history.
[0085] S13. Construct an original wind pressure coefficient data set based on the wind pressure correlation coefficient.
[0086] In a feasible implementation, firstly, the wind pressure coefficient time history is calculated through a wind tunnel test to obtain the wind pressure correlation coefficient between each measuring point.
[0087] Specifically, the structure surface measurement points The wind pressure is processed dimensionlessly to obtain the wind pressure coefficient time history of each measuring point. Defined as:
[0088] (1)
[0089] In formula (1), Measuring points on the structure surface The wind pressure time course, is the static pressure measured by the Pitot tube, is the air density, , is the average wind speed at the reference height of the building top.
[0090] Furthermore, after obtaining the wind pressure coefficient time history of each measuring point, the wind pressure correlation coefficient between each measuring point is obtained by the following formula (2): :
[0091] (2)
[0092] In formula (2), For measuring point At the moment The wind pressure coefficient time history, For measuring point At the moment The wind pressure coefficient time history, For measuring point and measuring points The covariance of the wind pressure coefficient time history, and The measuring points and measuring points Standard deviation of the wind pressure coefficient time history.
[0093] S2. Construct a new wind pressure coefficient data set based on the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the normalized original wind pressure coefficient data set.
[0094] Optionally, the above step S2 may include the following steps S21-S23:
[0095] S21. Construct a correlation coefficient matrix based on the original wind pressure coefficient data set.
[0096] S22. Calculate the mean of each row of the correlation coefficient matrix to obtain a new vector.
[0097] S23, normalizing the original wind pressure coefficient data set, and forming a new wind pressure coefficient data set according to the normalized original wind pressure coefficient data set and the new vector.
[0098] In a feasible implementation manner, the correlation coefficients between the measurement points form a correlation coefficient matrix, and a new set of vectors V is obtained by averaging each row of the matrix.
[0099] Furthermore, the wind pressure coefficient data set is normalized to the minimum and maximum values, and a new data set is formed with the vector. The minimum and maximum normalization formula is as follows:
[0100] (3)
[0101] In the formula, represents the minimum-maximum normalized data, and are the maximum and minimum values of the variable.
[0102] Furthermore, the data set is normalized and combined with the vector V to form a new data set P.
[0103] S3. Determine the optimal number of classifications based on the original wind pressure coefficient data set.
[0104] Optionally, the above step S3 may include the following steps S31-S33:
[0105] S31. Based on the original wind pressure coefficient data set, multiple value, and calculate each The sum of squared errors (SSE) of the values.
[0106] In a feasible implementation, the original wind pressure coefficient data set is selected in sequence according to K-means Perform clustering.
[0107] Furthermore, we calculate each The SSE (Sum of Square of Error) value corresponding to the value is:
[0108] (4)
[0109] In the formula, For the No. Wind pressure coefficient value of each measuring point; For the The average value of wind pressure coefficient of all measuring points in the class; For the The number of all measurement points in the class.
[0110] S32. Based on the original wind pressure coefficient data set, calculate the average silhouette coefficient of all measuring points.
[0111] In a feasible implementation mode, The calculation formula of the silhouette coefficient of each measuring point is as follows:
[0112] (5)
[0113] In the formula, Indicates The silhouette coefficient of the measuring points, It is the measuring point The average distance from the coefficient of to the coefficients of other points in the cluster to which it belongs, It is the measuring point The minimum value of the average distance from the coefficient of to all the measured point coefficients in other clusters.
[0114] Furthermore, the silhouette coefficients of all measuring points Take the average and get the final average silhouette coefficient.
[0115] S33, according to Value, each The sum of squared errors (SSE) and the average silhouette coefficient of the values are plotted. Relationship curve and The silhouette coefficient curve is based on Relationship curve and The silhouette coefficient curve determines the optimal number of clusters, where represents the sum of squared errors.
[0116] In a feasible implementation manner, according to each The SSE value obtained by formula (4) and the average silhouette coefficient value obtained by formula (5) are plotted. Relationship curve and Silhouette coefficient curve.
[0117] Furthermore, comprehensive The elbow position of the relationship curve and The first rise and fall point of the silhouette coefficient curve determines the optimal number of clusters .
[0118] S4. Based on the new wind pressure coefficient data set, the optimal number of classifications and the SOM method, the SOM wind pressure zoning results that consider the wind pressure correlation in the structural surface wind pressure zoning are obtained.
[0119] Optionally, the above step S4 may include the following steps S41-S46:
[0120] S41, randomly initialize according to the optimal number of classifications The weight vector of the center points.
[0121] In a feasible implementation, the optimal number of classifications is obtained And the new data set P composed of wind pressure correlation is randomly initialized The weight vector of the center points.
[0122] S42. Input the wind pressure coefficient in the new wind pressure coefficient data set into the improved SOM neural network model, calculate the Euclidean distance between the wind pressure coefficient in the new wind pressure coefficient data set and the weight vector, and obtain the neuron with the smallest Euclidean distance as the winning neuron.
[0123] In a feasible implementation, the improved SOM neural network model is shown in FIG. Figure 3 As shown, a new data set sample is input into the network, the Euclidean distance between the sample and the weight vector is calculated, and the neuron with the smallest distance is the winning neuron.
[0124] S43, updating the neurons in the topological neighborhood of the winning neuron to obtain a renormalized learned weight vector.
[0125] In one feasible implementation, the neurons in the topological neighborhood of the winning neuron are updated and the learned weights are renormalized. The basic formula is as follows:
[0126] (6)
[0127] In the formula, represents the weight vector after renormalization learning, represents the input dataset, represents the output neuron, represents the initial value of the weight vector, represents the iterative learning rate.
[0128] S44, determine whether the change of the weight vector is within a preset change range; if so, execute step S46; if not, execute step S45.
[0129] S45, determine whether the preset number of iterations has been reached; if so, execute step S46; if not, go to execute step S42.
[0130] In a feasible implementation manner, steps S42 to S43 are repeated until the weight vector of the center point no longer changes significantly or a preset number of iterations is reached.
[0131] S46. According to the weight vector, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning.
[0132] In a feasible implementation, the nodes are divided into classes, each class contains the data samples represented by the nodes closest to the class, stops the iterative calculation, and outputs the final partition result.
[0133] For example, this embodiment takes the wind pressure coefficient zoning of a high-rise building structure surface as an example. The wind tunnel test is carried out at the Wind Engineering Test Center of Hunan University of Science and Technology. The test is set as a typical square high-rise building. The test model and wind direction angle are defined as follows: Figure 4 As shown in the figure, the model size is 140mm (length) × 140mm (width) × 580mm (height), the scale ratio is 1:300, and each facade is set with 16 rows and a total of 90 measuring points. The test simulates the Class B landform specified in GB50009-2012 "Code for Loads on Building Structures". Based on the wind tunnel test, the wind pressure coefficient of the model surface is as follows: Figure 5 As shown, the maximum wind pressure coefficient at a wind direction angle of 0° and the minimum wind pressure coefficient at a wind direction angle of 90° of the S facade of the model are partitioned.
[0134] The detailed steps of this embodiment are as follows:
[0135] 1) The wind pressure at the measuring points on the model surface is dimensionlessly processed using formula (1) to obtain the wind pressure coefficient time history of each measuring point.
[0136] 2) Calculate the wind pressure correlation coefficient between each measuring point according to formula (2).
[0137] 3) Take the mean of each row of the obtained correlation coefficient matrix to obtain a new set of variables.
[0138] 4) The wind pressure coefficient of the S facade of the model obtained from the wind tunnel test is normalized according to formula (3) and combined with the variables in (3) to form a new data set.
[0139] 5) According to formula (4) and formula (5), we can get and The silhouette coefficient curve, such as Figure 6 As shown, the optimal classification number of the maximum wind pressure coefficient in this example is determined 3, the optimal number of classifications for the minimum wind pressure coefficient is 4.
[0140] 6) Random Initialization The weight vector of the center point is input into the network, and the Euclidean distance between the sample and the weight vector is calculated. The neuron with the smallest distance is the winning neuron.
[0141] 7) Renormalize the learned weights according to formula (6).
[0142] 8) When the preset number of iterations is reached, the final partition result is output as follows Figure 7 shown.
[0143] In an embodiment of the present invention, a SOM wind pressure zoning method is provided that considers wind pressure correlation in building surface wind pressure zoning. and The silhouette coefficient curve can quickly and accurately determine the optimal number of classifications ; The correlation coefficient of each measuring point is introduced, and the processed constraint variable distribution also reflects the wind pressure distribution characteristics. After introducing the constraint variable, the discrete points of the partition can be eliminated.
[0144] Through the competitive learning mechanism, SOM can adaptively adjust the position of neurons to better match the distribution characteristics of input data; through the two-dimensional or three-dimensional mapping results of SOM, the wind pressure zoning situation can be intuitively displayed, making it easier to understand the distribution and clustering results of the data; compared with traditional clustering algorithms, the SOM algorithm is good at processing high-dimensional data and can map complex data sets to low-dimensional space for analysis through dimensionality reduction.
[0145] Figure 8 1 is a block diagram of a SOM partitioning device considering wind pressure correlation according to an exemplary embodiment, wherein the device is used in a SOM partitioning method considering wind pressure correlation. Figure 8 The device includes an acquisition module 310, a construction module 320, a determination module 330 and an output module 340. Among them:
[0146] The acquisition module 310 is used to obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and construct an original wind pressure coefficient data set according to the wind pressure time history.
[0147] The construction module 320 is used to construct a new wind pressure coefficient data set according to the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the normalized original wind pressure coefficient data set.
[0148] The determination module 330 is used to determine the optimal number of classifications according to the original wind pressure coefficient data set.
[0149] The output module 340 is used to obtain the SOM wind pressure zoning result considering the wind pressure correlation in the structural surface wind pressure zoning according to the new wind pressure coefficient data set, the optimal classification number and the SOM method.
[0150] Optionally, the acquisition module 310 is further configured to:
[0151] S11. Obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and calculate the wind pressure coefficient time history of each measuring point based on the wind pressure time history.
[0152] S12. Calculate the wind pressure correlation coefficient between each measuring point based on the wind pressure coefficient time history.
[0153] S13. Construct an original wind pressure coefficient data set based on the wind pressure correlation coefficient.
[0154] Optionally, the construction module 320 is further configured to:
[0155] S21. Construct a correlation coefficient matrix based on the original wind pressure coefficient data set.
[0156] S22. Calculate the mean of each row of the correlation coefficient matrix to obtain a new vector.
[0157] S23, normalizing the original wind pressure coefficient data set, and forming a new wind pressure coefficient data set according to the normalized original wind pressure coefficient data set and the new vector.
[0158] Optionally, the determination module 330 is further configured to:
[0159] S31. Based on the original wind pressure coefficient data set, multiple value, and calculate each The sum of squared errors (SSE) of the values.
[0160] S32. Based on the original wind pressure coefficient data set, calculate the average silhouette coefficient of all measuring points.
[0161] S33, according to Value, each The sum of squared errors (SSE) and the average silhouette coefficient of the values are plotted. Relationship curve and The silhouette coefficient curve is based on Relationship curve and The silhouette coefficient curve determines the optimal number of clusters, where represents the sum of squared errors.
[0162] Optionally, the output module 340 is further configured to:
[0163] S41, randomly initialize according to the optimal number of classifications The weight vector of the center points.
[0164] S42. Input the wind pressure coefficient in the new wind pressure coefficient data set into the improved SOM neural network model, calculate the Euclidean distance between the wind pressure coefficient in the new wind pressure coefficient data set and the weight vector, and obtain the neuron with the smallest Euclidean distance as the winning neuron.
[0165] S43, updating the neurons in the topological neighborhood of the winning neuron to obtain a renormalized learned weight vector.
[0166] S44, determine whether the change of the weight vector is within a preset change range; if so, execute step S46; if not, execute step S45.
[0167] S45, determine whether the preset number of iterations has been reached; if so, execute step S46; if not, go to execute step S42.
[0168] S46. According to the weight vector, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning.
[0169] Optionally, the calculation formula of the renormalized learned weight vector is as shown in the following formula (1):
[0170] (1)
[0171] In the formula, represents the weight vector after renormalization learning, represents the input dataset, represents the output neuron, represents the initial value of the weight vector, represents the iterative learning rate.
[0172] In an embodiment of the present invention, a SOM wind pressure zoning method is provided that considers wind pressure correlation in building surface wind pressure zoning. and The silhouette coefficient curve can quickly and accurately determine the optimal number of classifications ; The correlation coefficient of each measuring point is introduced, and the processed constraint variable distribution also reflects the wind pressure distribution characteristics. After introducing the constraint variable, the discrete points of the partition can be eliminated.
[0173] Through the competitive learning mechanism, SOM can adaptively adjust the position of neurons to better match the distribution characteristics of input data; through the two-dimensional or three-dimensional mapping results of SOM, the wind pressure zoning situation can be intuitively displayed, making it easier to understand the distribution and clustering results of the data; compared with traditional clustering algorithms, the SOM algorithm is good at processing high-dimensional data and can map complex data sets to low-dimensional space for analysis through dimensionality reduction.
[0174] Fig. 9 is a schematic diagram of the structure of a partitioning device provided by an embodiment of the present invention, such as Fig. 9 As shown, the partition device may include the above Figure 8 The SOM partitioning device considering wind pressure correlation is shown. Optionally, the partitioning device 410 may include a first processor 2001 .
[0175] Optionally, the partition device 410 may further include a memory 2002 and a transceiver 2003 .
[0176] The first processor 2001, the memory 2002 and the transceiver 2003 may be connected via a communication bus.
[0177] Combine the following Fig. 9 The components of the partition device 410 are described in detail:
[0178] The first processor 2001 is the control center of the partition device 410, and may be a processor or a general term for multiple processing elements. For example, the first processor 2001 is one or more central processing units (CPUs), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (field programmable gate arrays, FPGAs).
[0179] Optionally, the first processor 2001 may execute various functions of the partition device 410 by running or executing a software program stored in the memory 2002 and calling data stored in the memory 2002 .
[0180] In a specific implementation, as an embodiment, the first processor 2001 may include one or more CPUs, such as Fig. 9 CPU0 and CPU1 are shown in FIG.
[0181] In a specific implementation, as an embodiment, the partition device 410 may also include multiple processors, such as Fig. 9 The first processor 2001 and the second processor 2004 are shown in FIG. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). The processor here may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).
[0182] The memory 2002 is used to store the software program for executing the solution of the present invention, and is controlled to be executed by the first processor 2001. The specific implementation method can refer to the above method embodiment, which will not be repeated here.
[0183] Optionally, the memory 2002 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 2002 may be integrated with the first processor 2001, or may exist independently, and access the first processor 2001 through the interface circuit ( Fig. 9 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.
[0184] The transceiver 2003 is used to communicate with a network device or a terminal device.
[0185] Optionally, the transceiver 2003 may include a receiver and a transmitter ( Fig. 9 The receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.
[0186] Optionally, the transceiver 2003 may be integrated with the first processor 2001, or may exist independently and communicate with the first processor 2001 through the interface circuit ( Fig. 9 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.
[0187] It should be noted that Fig. 9 The structure of the partition device 410 shown in the figure does not constitute a limitation on the router, and the actual knowledge structure recognition device may include more or less components than those shown in the figure, or combine certain components, or arrange the components differently.
[0188] In addition, the technical effects of the partitioning device 410 can refer to the technical effects of the SOM partitioning method considering wind pressure correlation described in the above method embodiment, which will not be repeated here.
[0189] It should be understood that the first processor 2001 in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0190] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).
[0191] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.
[0192] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.
[0193] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.
[0194] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.
[0195] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0196] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0197] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0198] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0199] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
[0200] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.
[0201] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
Claims
1. A SOM partitioning method considering wind pressure correlation, characterized in that: The method comprises: S1. Obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and construct an original wind pressure coefficient data set according to the wind pressure time history; S2, constructing a new wind pressure coefficient data set according to the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the normalized original wind pressure coefficient data set; S3, determining the optimal number of classifications according to the original wind pressure coefficient data set; S4. According to the new wind pressure coefficient data set, the optimal number of classifications and the SOM method, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning; Wherein, determining the optimal number of classifications according to the original wind pressure coefficient data set in S3 includes: S31, based on the original wind pressure coefficient data set, select multiple values in sequence according to the K-means algorithm, and calculate each The sum of squared errors of the values ; S32, calculating the average contour coefficient of all measuring points based on the original wind pressure coefficient data set; S33, according to Value, each The sum of squared errors of the values And the average silhouette coefficient, plot Relationship curve and The contour coefficient curve is based on the Relationship curve and The silhouette coefficient curve determines the optimal number of clusters, where represents the sum of squared errors.
2. The SOM partitioning method considering wind pressure correlation according to claim 1 is characterized in that: The step S1 of obtaining the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test and constructing an original wind pressure coefficient data set according to the wind pressure time history includes: S11, obtaining the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and calculating the wind pressure coefficient time history of each measuring point according to the wind pressure time history; S12, calculating the wind pressure correlation coefficient between each measuring point according to the wind pressure coefficient time history; S13. Constructing an original wind pressure coefficient data set according to the wind pressure correlation coefficient.
3. The SOM partitioning method considering wind pressure correlation according to claim 1 is characterized in that: The step S2 of constructing a new wind pressure coefficient dataset based on the vector obtained by processing the wind pressure correlation coefficients between the measuring points in the original wind pressure coefficient dataset and the normalized original wind pressure coefficient dataset includes: S21, constructing a correlation coefficient matrix according to the original wind pressure coefficient data set; S22, calculate the average value of each row of the correlation coefficient matrix to obtain a new vector; S23, normalizing the original wind pressure coefficient data set, and forming a new wind pressure coefficient data set according to the normalized original wind pressure coefficient data set and the new vector.
4. The SOM partitioning method considering wind pressure correlation according to claim 1 is characterized in that: The S4 step of obtaining the SOM wind pressure partitioning result taking into account the wind pressure correlation in the structural surface wind pressure partitioning according to the new wind pressure coefficient data set, the optimal number of classifications and the SOM method includes: S41. Randomly initialize the weight vectors of the center points according to the optimal number of classifications; S42, inputting the wind pressure coefficient in the new wind pressure coefficient data set into the improved SOM neural network model, calculating the Euclidean distance between the wind pressure coefficient in the new wind pressure coefficient data set and the weight vector, and obtaining a neuron with the smallest Euclidean distance as a winning neuron; S43, updating the neurons in the topological neighborhood of the winning neuron to obtain a renormalized learned weight vector; S44, determining whether the change amount of the weight vector is within a preset change range; if so, executing step S46; if not, executing step S45; S45, determine whether the preset number of iterations has been reached; if so, execute step S46; if not, go to step S42; S46. According to the weight vector, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning.
5. The SOM partitioning method considering wind pressure correlation according to claim 4 is characterized in that: The calculation formula of the re-normalized learned weight vector is shown in the following formula (1): (1) In the formula, represents the weight vector after renormalization learning, represents the input dataset, represents the output neuron, represents the initial value of the weight vector, represents the iterative learning rate.
6. A SOM partitioning device considering wind pressure correlation, the SOM partitioning device considering wind pressure correlation is used to implement the SOM partitioning method considering wind pressure correlation as claimed in any one of claims 1 to 5, characterized in that: The device comprises: An acquisition module is used to obtain the wind pressure time history of each measuring point on the surface of the structure through a wind tunnel test, and to construct an original wind pressure coefficient data set according to the wind pressure time history; A construction module, used to construct a new wind pressure coefficient data set according to the vector obtained by processing the wind pressure correlation coefficient between each measuring point in the original wind pressure coefficient data set and the unitized original wind pressure coefficient data set; A determination module, used to determine the optimal number of classifications according to the original wind pressure coefficient data set; The output module is used to obtain the SOM wind pressure zoning result that takes into account the wind pressure correlation in the structural surface wind pressure zoning according to the new wind pressure coefficient data set, the optimal classification number and the SOM method.
7. The SOM partitioning device considering wind pressure correlation according to claim 6 is characterized in that: The SOM wind pressure zoning result considering the wind pressure correlation in the structural surface wind pressure zoning is obtained according to the new wind pressure coefficient data set, the optimal classification number and the SOM method, including: S41, randomly initialize according to the optimal number of classifications The weight vector of the center points; S42, inputting the wind pressure coefficient in the new wind pressure coefficient data set into the improved SOM neural network model, calculating the Euclidean distance between the wind pressure coefficient in the new wind pressure coefficient data set and the weight vector, and obtaining a neuron with the smallest Euclidean distance as a winning neuron; S43, updating the neurons in the topological neighborhood of the winning neuron to obtain a renormalized learned weight vector; S44, determining whether the change amount of the weight vector is within a preset change range; if so, executing step S46; if not, executing step S45; S45, determine whether the preset number of iterations has been reached; if so, execute step S46; if not, go to step S42; S46. According to the weight vector, a SOM wind pressure zoning result is obtained that takes into account the wind pressure correlation in the structural surface wind pressure zoning.
8. A partitioning device, characterized in that: The partitioning device comprises: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 5 is implemented.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, which can be called by a processor to execute the method according to any one of claims 1 to 5.
Citation Information
Patent Citations
Thermal power generating unit blower overall health state evaluation method, computer and storage medium
CN114167730A
Air cooling array partition control method and device based on artificial intelligence
CN115638671A
Wind power device power prediction method and system and storage medium
CN115952895A
KR20230109922A
