CN116738853A - Method and system for evaluating importance coefficient of planar frame structural member - Google Patents
Method and system for evaluating importance coefficient of planar frame structural member Download PDFInfo
- Publication number
- CN116738853A CN116738853A CN202310749170.9A CN202310749170A CN116738853A CN 116738853 A CN116738853 A CN 116738853A CN 202310749170 A CN202310749170 A CN 202310749170A CN 116738853 A CN116738853 A CN 116738853A
- Authority
- CN
- China
- Prior art keywords
- sample
- frame structure
- importance
- plane frame
- component
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000013210 evaluation model Methods 0.000 claims abstract description 46
- 238000005452 bending Methods 0.000 claims description 29
- 230000003068 static effect Effects 0.000 claims description 16
- 238000012549 training Methods 0.000 claims description 16
- 238000010606 normalization Methods 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 10
- 230000009467 reduction Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 5
- 230000006866 deterioration Effects 0.000 claims 4
- 238000010801 machine learning Methods 0.000 abstract description 3
- 239000011159 matrix material Substances 0.000 description 16
- 230000015556 catabolic process Effects 0.000 description 14
- 238000006731 degradation reaction Methods 0.000 description 14
- 238000004364 calculation method Methods 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000003062 neural network model Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 2
- 238000005096 rolling process Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005489 elastic deformation Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004643 material aging Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 230000035882 stress Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/042—Knowledge-based neural networks; Logical representations of neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/0985—Hyperparameter optimisation; Meta-learning; Learning-to-learn
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Geometry (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)
Abstract
The invention discloses a method and a system for evaluating importance coefficients of a planar frame structural member, which relate to the field of structural engineering and machine learning, and comprise the following steps: obtaining an undirected graph corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the edges of the undirected graph represent that members corresponding to two graph nodes connected by the edges are connected; inputting node characteristics and edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain an importance coefficient of each component in the target plane frame structure; the node characteristics comprise component information of components corresponding to each graph node; the edge feature includes connection information for each component to all components. According to the invention, the importance coefficient of each component in the target plane frame structure is evaluated by adopting the trained importance coefficient evaluation model, so that the artificial subjective influence is avoided, and the calculated amount is small.
Description
Technical Field
The invention relates to the field of structural engineering and machine learning, in particular to a method and a system for evaluating importance coefficients of planar frame structural members.
Background
In the long-term service process, the key parts or components of the planar frame structure can be damaged and accumulated due to environmental erosion, load action, material aging and other factors, so that the resistance of the planar frame structure is attenuated, and even the planar frame structure is totally invalid under extreme conditions, and catastrophic accidents are caused. Therefore, the stress performance of the in-service planar frame structure needs to be known or mastered in time, the safety state evaluation and the targeted maintenance and reinforcement are carried out, and the component importance coefficient is an important component for evaluating the overall safety of the planar frame structure. The existing component importance coefficient evaluation method includes an expert experience evaluation method or a failure path search method, however, the former is highly subjective and the latter is excessively calculated. Therefore, there is a need for a method for evaluating the importance coefficient of a member that avoids artificial subjective influence and is small in calculation amount.
Disclosure of Invention
The invention aims to provide a method and a system for evaluating importance coefficients of a planar frame structure member, which can avoid artificial subjective influence and reduce the calculated amount.
In order to achieve the above object, the present invention provides the following solutions:
a method of evaluating importance coefficients of planar frame structural members, the method comprising:
obtaining an undirected graph corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the sides of the undirected graph represent that members corresponding to two graph nodes connected by the sides are connected;
inputting node characteristics and edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain an importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking sample node characteristics and sample edge characteristics of a sample undirected graph of a sample plane frame structure as input and taking sample importance coefficients of each sample component in the sample plane frame structure as labels.
Optionally, before inputting the node features and the edge features of the undirected graph into the trained importance coefficient evaluation model, the method further includes: training an importance coefficient evaluation model, which specifically comprises the following steps:
acquiring a data set; the data set comprises sample node characteristics and sample edge characteristics of a sample undirected graph of each sample plane frame structure in a plurality of sample plane frame structures and sample importance coefficients of each sample member;
and training the importance coefficient evaluation model by using the data set to obtain a trained importance coefficient evaluation model.
Optionally, the acquiring the data set specifically includes:
acquiring a plurality of sample plane frame structures;
for each sample plane frame structure, representing the sample plane frame structure based on a graph theory mode to obtain a sample undirected graph of the sample plane frame structure; a sample graph node of the sample undirected graph corresponds to a sample member of the sample planar frame structure; the sample edges of the sample undirected graph represent that sample members corresponding to two sample graph nodes connected by the sample edges are connected; the sample node characteristics of the sample undirected graph comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the sample component and the foundation and the position information of the sample component in the sample plane frame structure of the sample component corresponding to each sample graph node; the sample edge features of the sample undirected graph comprise connection information of each sample member and all sample members;
calculating the total elastic strain energy of the sample plane frame structure after the sample member is removed under the preset static load condition for each sample member in each sample plane frame structure; calculating a sample importance coefficient of the sample member according to the total elastic strain energy; all sample node features and sample edge features of the sample undirected graph of the sample plane frame structure and the sample importance coefficients of each sample member constitute the dataset.
Optionally, the preset static load condition includes load types and load sizes of all sample components in the sample plane frame structure; the load types comprise vertical concentrated load, vertical uniformly distributed load, horizontal concentrated load and/or horizontal linearly distributed load.
Optionally, the expression of the sample importance coefficient is:
wherein, gamma i Sample importance coefficient for the i-th sample member, i=1, 2, n; n is the total number of sample members of the sample plane frame structure; u is the total elastic strain energy of the sample plane frame structure under the preset static load condition; u (U) i The total elastic strain energy of the sample plane frame structure after the ith sample member is removed under the preset dead load condition.
Optionally, after calculating the sample importance coefficient of the sample member according to the total elastic strain energy, further comprising:
and carrying out normalization processing on the sample importance coefficient to obtain a normalization coefficient, and taking the normalization coefficient as a new sample importance coefficient.
Optionally, when the member of the target planar frame structure has member resistance degradation, replacing the axial rigidity and the bending rigidity of the member having member resistance degradation in the node characteristics with a degraded axial rigidity and a degraded bending rigidity, wherein the degraded axial rigidity is obtained by multiplying the axial rigidity of the member having member resistance degradation by a reduction coefficient, and the degraded bending rigidity is obtained by multiplying the bending rigidity of the member having member resistance degradation by a reduction coefficient.
Optionally, the trained importance coefficient evaluation model is a GCN model.
Optionally, the trained importance coefficient evaluation model includes three GCNConv layers connected in sequence.
The invention also provides a system for evaluating importance coefficients of planar frame structural members, which comprises:
the image acquisition module is used for acquiring an undirected image corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the sides of the undirected graph represent that members corresponding to two graph nodes connected by the sides are connected;
the importance coefficient evaluation module is used for inputting the node characteristics and the edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain the importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking sample node characteristics and sample edge characteristics of a sample undirected graph of a sample plane frame structure as input and taking sample importance coefficients of each sample component in the sample plane frame structure as labels.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects: the invention provides a method and a system for evaluating importance coefficients of a planar frame structural member, wherein the method comprises the following steps: obtaining an undirected graph corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the edges of the undirected graph represent that members corresponding to two graph nodes connected by the edges are connected; inputting node characteristics and edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain an importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking the sample node characteristics and the sample edge characteristics of a sample undirected graph of the sample plane frame structure as input and taking the sample importance coefficient of each sample component in the sample plane frame structure as a label. According to the invention, the importance coefficient of each component in the target plane frame structure is evaluated by adopting the trained importance coefficient evaluation model, so that the artificial subjective influence is avoided, and the calculated amount is small.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for evaluating importance coefficients of a planar frame structure member according to embodiment 1 of the present invention;
fig. 2 is a flowchart for evaluating importance coefficients of planar frame structural members based on a graph convolutional neural network according to embodiment 1 of the present invention;
fig. 3 is a schematic view of a three-span four-layer planar frame structure provided in embodiment 1 of the present invention;
fig. 4 is a schematic diagram of an undirected graph model corresponding to the planar frame structure provided in embodiment 1 of the present invention;
FIG. 5 is a diagram showing the comparison result between the model predicted value and the actual value of the importance coefficient of the component according to embodiment 1 of the present invention;
fig. 6 is a block diagram of a system for evaluating importance coefficients of a planar frame structure member according to embodiment 2 of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention aims to provide a method and a system for evaluating importance coefficients of components of a planar frame structure, wherein the importance coefficients of each component in a target planar frame structure are evaluated through a graph convolutional neural network (namely a trained importance coefficient evaluation model), so that artificial subjective influence is avoided, and the machine learning method has strong data analysis capability and excellent calculation efficiency in the aspect of processing multiple sample problems, and reduces the calculated amount.
In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.
Example 1
As shown in fig. 1, the present invention provides a method for evaluating importance coefficients of a planar frame structural member, the method comprising:
s1: obtaining an undirected graph corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the edges of the undirected graph represent that members corresponding to two graph nodes connected by the edges are connected.
S2: inputting node characteristics and edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain an importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking sample node characteristics and sample edge characteristics of a sample undirected graph of a sample plane frame structure as input and taking sample importance coefficients of each sample component in the sample plane frame structure as labels.
As shown in fig. 2, before inputting the node features and the edge features of the undirected graph into the trained importance coefficient evaluation model, the method further includes: training an importance coefficient evaluation model, which specifically comprises the following steps:
acquiring a data set; the dataset includes sample node features and sample edge features of a sample undirected graph for each of a number of sample plane frame structures and sample importance coefficients for each sample member.
And training the importance coefficient evaluation model by using the data set to obtain a trained importance coefficient evaluation model.
The method for acquiring the data set specifically comprises the following steps:
acquiring a plurality of sample plane frame structures;
for each sample plane frame structure, representing the sample plane frame structure based on a graph theory mode to obtain a sample undirected graph of the sample plane frame structure; a sample graph node of the sample undirected graph corresponds to a sample member of the sample planar frame structure; the sample edges of the sample undirected graph represent that sample members corresponding to two sample graph nodes connected by the sample edges are connected; the sample node characteristics of the sample undirected graph comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the sample component and the foundation and the position information of the sample component in the sample plane frame structure of the sample component corresponding to each sample graph node; the sample edge features of the sample undirected graph include connection information for each of the sample members to all of the sample members.
Calculating the total elastic strain energy of the sample plane frame structure after the sample member is removed under the preset static load condition for each sample member in each sample plane frame structure; calculating a sample importance coefficient of the sample member according to the total elastic strain energy; all sample node features and sample edge features of the sample undirected graph of the sample plane frame structure and the sample importance coefficients of each sample member constitute the dataset.
The training process of the importance coefficient evaluation model is divided into three steps: (1) generating a dataset; (2) building an importance coefficient evaluation model; and (3) model training and verification. The following specifically describes the above steps:
(1) Generating a data set: generating the dataset includes two parts, a first part being to build the graph model and a second part being to calculate a sample importance coefficient of the sample member.
Constructing a graph model:
in the embodiment, a PyTorch Geometric library in Python language is adopted to construct a graph model, and for each sample plane frame structure, sample graph nodes of a sample undirected graph represent sample members of the sample plane frame structure; the sample side represents the connectivity between two sample members, the sample side has no directionality. Each sample member is used as a sample graph node, and a sample edge is drawn between the sample graph nodes corresponding to the two sample members with the connected relation.
Sample node features of the sample undirected graph include: the sample graph node corresponds to a member Type (beam or Column) (Type), a member Axial Stiffness (Axial static), a bending Stiffness (Bending Stiffness), a Load Type (Load Type), a Load size (Load), connection information (group) of the member and a foundation, and position information of the sample member in the sample plane frame structure, wherein the position information comprises a Layer (Layer) and a Column (Column) of the sample member in the sample plane frame structure. The load types include vertical concentrated load, vertical uniform load, horizontal concentrated load and/or horizontal linear distributed load.
In this embodiment, the 8 kinds of information are input in the form of feature matrix in the order of table 1. Before inputting the importance coefficient evaluation model, normalization processing is further needed to be performed on the information, wherein the Type (Type) of the component, the connection information (group) of the sample component and the foundation and the load Type (LoadType) are used as category characteristics, and one-hot coding is adopted; the Layer (Layer), column (Column), axial Stiffness (Axial Stiffness), flexural Stiffness (Bending Stiffness), load size (Load) of the component are treated with a min-max scale.
TABLE 1 feature matrix example
Wherein the first row is specific information of each type of node feature, the second row and the third row are specific information of two sample members in a sample plane frame structure in table 1. In a column of the Type (Type) of the member, 1 represents the Type of the member as a beam, and 0 represents the Type of the member as a column. In a column of connection information (group) of a sample member to a foundation, 1 indicates that the member is connected to the foundation, and 0 indicates that the member is not connected to the foundation. The Load Type (Load Type) is shown in a column 1, and the working condition is that the vertical concentrated Load and the horizontal concentrated Load act simultaneously.
Sample edge characteristics of sample edges of the sample undirected graph and input forms are as follows: if there is a connection between the two sample members, a sample edge is provided between the sample graph nodes representing the two sample members, and the element values of the elements of the ith row and the jth column of the adjacency matrix represent whether the ith sample member is connected with the jth sample member or not, taking the adjacency matrix as an input form.
Edge Features (Edge Features) are used in this embodiment to describe the connection relationships between nodes in an undirected graph, which are input in the form of an Adjacency Matrix (Adjacency Matrix).
Generating a training data set of an importance coefficient evaluation model: the data set for the importance coefficient assessment model training includes a sample importance coefficient for each sample member in the sample plane frame structure. In the present embodiment, the importance coefficient of the sample member in the sample plane frame structure is the importance coefficient in the case of small linear elastic deformation. In the present embodiment, the sample member importance coefficient of the sample plane frame structure is calculated by a member removal method and an energy method, that is, the sample importance coefficient γ of the sample member i In order to remove the ratio of the total elastic strain energy stored by the sample plane frame structure under the preset static load condition and the total elastic strain energy of the sample plane frame structure without removing the sample member under the same preset static load condition, the expression of the sample importance coefficient is as follows:
wherein, gamma i Sample importance coefficient for the i-th sample member, i=1, 2, n; n is the total number of sample members of the sample plane frame structure; u is the total elastic strain energy of a perfect sample plane frame structure without dismantling the sample member under the preset static load condition; u (U) i The total elastic strain energy of the sample plane frame structure after the ith sample member is removed under the preset dead load condition. The preset static load condition comprises load types and load sizes of all sample components in the sample plane frame structure.
After calculating the sample importance coefficient of the sample member from the total elastic strain energy, further comprising: and carrying out normalization processing on the sample importance coefficient to obtain a normalization coefficient, and taking the normalization coefficient as a new sample importance coefficient. Specifically:
wherein beta is i The normalized coefficient of the i-th sample member, i.e., the normalized sample importance coefficient.
In this embodiment, when there is a member resistance degradation in the sample member of the sample plane frame structure, the axial rigidity and the bending rigidity of the member in which there is a member resistance degradation in the sample node characteristics are replaced with a degraded axial rigidity obtained by multiplying the axial rigidity of the sample in which there is a member resistance degradation by a reduction coefficient, and a degraded bending rigidity obtained by multiplying the bending rigidity of the sample in which there is a member resistance degradation by a reduction coefficient.
(2) Building an importance coefficient evaluation model;
in this embodiment, the trained importance coefficient evaluation model is a GCN model, i.e., a graph convolution neural network model. The trained importance coefficient evaluation model comprises three GCNConv layers which are connected in sequence.
In this embodiment, the input feature dimension of the node feature is 8, the edge feature is input in the form of an adjacency matrix, the output feature dimension is 1, the Mean Square Error (MSE) is adopted as the loss function, and the Adam optimizer is adopted to optimize the model. The parameters of the model are iteratively updated by back propagation and gradient descent to reduce the loss value. The mean square error calculation formula is as follows:
wherein MSE is the mean square error loss value, m is the number of samples, y i For the actual value of the sample importance coefficient of the i-th sample member,a predicted value of the sample importance coefficient for the i-th sample member.
(3) Model training and verification:
dividing the data set generated in the step (1) into a training set and a verification set, inputting the graph rolling neural network model (importance coefficient evaluation model) built in the step (2), firstly training the graph rolling neural network model by adopting the training set in each training period (epoch), and then evaluating the performance of the model by using the verification set. The neural network was convolved with a sample set training chart, with a final train loss of 0.000975,val loss (validation loss) of 0.00028.
Based on the above, a trained importance coefficient evaluation model can be obtained. And (3) for the target plane frame structure to be evaluated, determining a feature matrix of a graph node of the undirected graph corresponding to the target plane frame structure according to the step (1), taking the feature matrix as a node feature of the undirected graph, taking an adjacent matrix of the edge as an edge feature of the undirected graph, inputting the edge feature of the undirected graph into a trained importance coefficient evaluation model, and evaluating the importance of the components of the target plane frame structure under the corresponding static load to obtain the importance coefficient of each component in the target plane frame structure.
The following describes the evaluation process of the importance coefficient of the components thereof, taking the three-span four-layer planar frame structure shown in fig. 3 as an example:
according to the planar frame structure shown in fig. 3, there are 28 members in total, the frame structure parameters of the planar frame structure are shown in table 2, and the undirected graph corresponding to the planar frame structure is shown in fig. 4. Assuming that the vertical concentrated load and the horizontal concentrated load act simultaneously, the specific load sizes are shown in table 3, and the resistance degradation (simultaneous degradation of EA and EI) of the No. 3 beam member is 85% to the initial intact state.
When the member of the target plane frame structure has member resistance degradation, the axial rigidity and the bending rigidity of the member with member resistance degradation in the node characteristics are replaced by the degraded axial rigidity and the degraded bending rigidity, wherein the degraded axial rigidity is obtained by multiplying the axial rigidity of the member with member resistance degradation by a reduction coefficient, and the degraded bending rigidity is obtained by multiplying the bending rigidity of the member with member resistance degradation by a reduction coefficient.
Therefore, the feature matrix of the node feature of the undirected graph shown in fig. 4 can be obtained, and as shown in table 4, each parameter in table 4 is a parameter that is not subjected to normalization pretreatment. Based on the connection information of each member in the planar frame structure shown in fig. 3, the adjacency matrix of the edge features of the undirected graph corresponding to the planar frame structure is obtained as shown in table 5.
Table 2 frame structure parameters
The cross sections of the beam and the column members are rectangular, and when no member resistance is degraded, the elastic modulus of the planar frame structure shown in fig. 3 is 48000MPa.
Table 3 type and size for 5 loads shown in fig. 3
| Load name | Load type | Load size/KN |
| F1 | Vertical concentrated load | 4.2 |
| F2 | Horizontal concentrated load | 40 |
| F3 | Horizontal concentrated load | 30 |
| F4 | Horizontal concentrated load | 20 |
| F5 | Horizontal concentrated load | 10 |
TABLE 4 characterization matrix
In table 4: type 1 represents a beam and 0 represents a column; in group, 1 indicates that the component is connected with the foundation, and 0 indicates that the component is not connected with the foundation; 1 in Load Type indicates that the working condition (preset Load condition) is that the vertical concentrated Load and the horizontal concentrated Load act simultaneously.
TABLE 5 adjacency matrix
Inputting node characteristics and edge characteristics of the undirected graph corresponding to the planar frame structure in fig. 3 into the trained importance coefficient evaluation model, and predicting the importance coefficient of each component of the planar frame structure in fig. 3. And comparing the predicted value of the model with the actual value calculated by the energy method, and the comparison result is shown in figure 5. The maximum relative error is 0.00295, the error mean value is 0.000999, and the standard deviation of the error is 0.000131, so that the method can be used for efficiently and accurately evaluating the importance coefficient of the component of the planar frame structure under the static load.
The method for evaluating the importance coefficient of the planar frame structure member is suitable for the planar frame structure with only rigid nodes, and can input the feature matrix and the adjacent matrix of the undirected graph corresponding to the planar frame structure under the action of static load which is arbitrarily assumed into a trained graph convolutional neural network (an importance coefficient evaluation model) to quickly obtain the member importance evaluation result of the given intact planar frame structure (undetached member) under the working condition. Considering the condition that the resistance of the component in the planar frame structure is degraded, replacing the axial rigidity and the bending rigidity of the component with the degraded resistance with the degraded axial rigidity and the degraded bending rigidity to obtain new node characteristics, and carrying out importance coefficient evaluation by adopting the new node characteristics, so that the obtained component importance coefficient evaluation result is more accurate. The method provided by the invention can be used for efficiently and reasonably quantitatively evaluating the importance of the components of the planar frame structure under the action of static load, and the defects of insufficient objectivity of expert experience, large evaluation delay and low efficiency of failure path search calculation amount are overcome.
Example 2
As shown in fig. 6, the present invention also provides a system for evaluating importance coefficients of planar frame structural members, the system comprising:
the image acquisition module T1 is used for acquiring an undirected image corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the edges of the undirected graph represent that members corresponding to two graph nodes connected by the edges are connected.
The importance coefficient evaluation module T2 is used for inputting the node characteristics and the edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain the importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking sample node characteristics and sample edge characteristics of a sample undirected graph of a sample plane frame structure as input and taking sample importance coefficients of each sample component in the sample plane frame structure as labels. .
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.
Claims (10)
1. A method for evaluating importance coefficients of planar frame structural members, the method comprising:
obtaining an undirected graph corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the sides of the undirected graph represent that members corresponding to two graph nodes connected by the sides are connected;
inputting node characteristics and edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain an importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking sample node characteristics and sample edge characteristics of a sample undirected graph of a sample plane frame structure as input and taking sample importance coefficients of each sample component in the sample plane frame structure as labels.
2. The method of evaluating importance coefficients of planar frame structural members according to claim 1, further comprising, before inputting the node features and the edge features of the undirected graph into a trained importance coefficient evaluation model: training an importance coefficient evaluation model, which specifically comprises the following steps:
acquiring a data set; the data set comprises sample node characteristics and sample edge characteristics of a sample undirected graph of each sample plane frame structure in a plurality of sample plane frame structures and sample importance coefficients of each sample member;
and training the importance coefficient evaluation model by using the data set to obtain a trained importance coefficient evaluation model.
3. The method for evaluating importance coefficients of planar frame structural members according to claim 2, wherein said acquiring a data set specifically comprises:
acquiring a plurality of sample plane frame structures;
for each sample plane frame structure, representing the sample plane frame structure based on a graph theory mode to obtain a sample undirected graph of the sample plane frame structure; a sample graph node of the sample undirected graph corresponds to a sample member of the sample planar frame structure; the sample edges of the sample undirected graph represent that sample members corresponding to two sample graph nodes connected by the sample edges are connected; the sample node characteristics of the sample undirected graph comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the sample component and the foundation and the position information of the sample component in the sample plane frame structure of the sample component corresponding to each sample graph node; the sample edge features of the sample undirected graph comprise connection information of each sample member and all sample members;
calculating the total elastic strain energy of the sample plane frame structure after the sample member is removed under the preset static load condition for each sample member in each sample plane frame structure; calculating a sample importance coefficient of the sample member according to the total elastic strain energy; all sample node features and sample edge features of the sample undirected graph of the sample plane frame structure and the sample importance coefficients of each sample member constitute the dataset.
4. The method for evaluating importance coefficients of planar frame structural members according to claim 3, wherein the preset static load condition includes load types and load magnitudes of all sample members in the sample planar frame structure; the load types comprise vertical concentrated load, vertical uniformly distributed load, horizontal concentrated load and/or horizontal linearly distributed load.
5. The method for evaluating importance coefficients of planar frame structural members according to claim 3, wherein the expression of the sample importance coefficients is:
wherein, gamma i Sample importance coefficient for the i-th sample member, i=1, 2, n; n is the total number of sample members of the sample plane frame structure; u is the total elastic strain energy of the sample plane frame structure under the preset static load condition; u (U) i The total elastic strain energy of the sample plane frame structure after the ith sample member is removed under the preset dead load condition.
6. The method of evaluating the importance coefficient of a planar frame structural member according to claim 3, characterized by further comprising, after calculating the sample importance coefficient of the sample member from the total elastic strain energy:
and carrying out normalization processing on the sample importance coefficient to obtain a normalization coefficient, and taking the normalization coefficient as a new sample importance coefficient.
7. The method according to claim 1, wherein when there is a member resistance deterioration of the member of the target planar frame structure, the axial rigidity and the bending rigidity of the member having the member resistance deterioration in the node characteristic are replaced with a deteriorated axial rigidity obtained by multiplying the axial rigidity of the member having the member resistance deterioration by a reduction coefficient, and a deteriorated bending rigidity obtained by multiplying the bending rigidity of the member having the member resistance deterioration by a reduction coefficient.
8. The method for evaluating importance coefficients of planar frame structural members according to claim 1, wherein the trained importance coefficient evaluation model is a GCN model.
9. The method for evaluating importance coefficients of planar frame structural members according to claim 1, wherein said trained importance coefficient evaluation model comprises three GCNConv layers connected in sequence.
10. A planar frame structural member importance coefficient assessment system, the system comprising:
the image acquisition module is used for acquiring an undirected image corresponding to the target plane frame structure; one graph node of the undirected graph corresponds to one member of the target plane frame structure; the sides of the undirected graph represent that members corresponding to two graph nodes connected by the sides are connected;
the importance coefficient evaluation module is used for inputting the node characteristics and the edge characteristics of the undirected graph into a trained importance coefficient evaluation model to obtain the importance coefficient of each component in the target plane frame structure; the node characteristics comprise the component type, the axial rigidity, the bending rigidity, the load type, the load size, the connection information of the component and the foundation and the position information of the component in the target plane frame structure of the component corresponding to each graph node; the edge features comprise connection information of each component and all components; the trained importance coefficient evaluation model is a model which is obtained by taking sample node characteristics and sample edge characteristics of a sample undirected graph of a sample plane frame structure as input and taking sample importance coefficients of each sample component in the sample plane frame structure as labels.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310749170.9A CN116738853A (en) | 2023-06-25 | 2023-06-25 | Method and system for evaluating importance coefficient of planar frame structural member |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310749170.9A CN116738853A (en) | 2023-06-25 | 2023-06-25 | Method and system for evaluating importance coefficient of planar frame structural member |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116738853A true CN116738853A (en) | 2023-09-12 |
Family
ID=87902646
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310749170.9A Pending CN116738853A (en) | 2023-06-25 | 2023-06-25 | Method and system for evaluating importance coefficient of planar frame structural member |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116738853A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117350160A (en) * | 2023-10-18 | 2024-01-05 | 河海大学 | Single-layer reticulated shell member replacement optimal sequence determining method based on deep learning |
-
2023
- 2023-06-25 CN CN202310749170.9A patent/CN116738853A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117350160A (en) * | 2023-10-18 | 2024-01-05 | 河海大学 | Single-layer reticulated shell member replacement optimal sequence determining method based on deep learning |
| CN117350160B (en) * | 2023-10-18 | 2024-04-26 | 河海大学 | Single-layer reticulated shell member replacement optimal sequence determining method based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Nguyen et al. | Prediction of bridge deck condition rating based on artificial neural networks | |
| CN116738853A (en) | Method and system for evaluating importance coefficient of planar frame structural member | |
| CN111767398A (en) | Secondary equipment fault short text data classification method based on convolutional neural network | |
| CN110738349A (en) | Power grid fault first-aid repair duration prediction method based on multi-model fusion | |
| CN111814964A (en) | Air pollution treatment method based on air quality condition prediction and storage medium | |
| CN114692507B (en) | Soft measurement modeling method for count data based on stacked poisson self-encoder network | |
| CN111882198A (en) | Project performance evaluation method and system | |
| US20240257011A1 (en) | Risk assessment method and system for windage yaw flashover, device, and readable storage medium | |
| CN117875195B (en) | Crack propagation twinning prediction method for structural life assessment | |
| CN115456306A (en) | Bus load prediction method, system, equipment and storage medium | |
| CN116050600A (en) | CNN-GA-BP-based combined model spare part demand prediction method and system | |
| CN109101734B (en) | Prediction method for continuous rigid frame bridge downwarping risk | |
| DE112016007068T5 (en) | EQUIPMENT STATE ESTIMATOR, EQUIPMENT STATE ESTABLISHMENT METHOD, AND EQUIPMENT STATE SUPERVISION SYSTEM | |
| CN108280299B (en) | Truss structure sensitivity analysis method | |
| CN111061708A (en) | Electric energy prediction and restoration method based on LSTM neural network | |
| CN109255638B (en) | Mathematical model for mining potential customers | |
| CN114548585A (en) | Urban building earthquake damage prediction method based on neural network | |
| CN114647904A (en) | Self-adaptive sparse graph learning bearing life prediction method based on digital twin dictionary | |
| CN113343468A (en) | Method, device and equipment for carrying out multi-step prediction by SARIMA model | |
| CN112330046A (en) | Power demand prediction method based on multi-dimensional gray-neural network hybrid coordination | |
| CN112528510A (en) | Method for predicting repairable aviation material spare parts based on life-extinction process model | |
| CN118445905B (en) | Method for evaluating service life of corrugated steel web combined box girder | |
| CN114186627B (en) | QR-GRU-based power plant statistical data prediction and verification method | |
| CN112364995B (en) | Transformer fault rate assessment method and system and readable storage medium | |
| Arifin et al. | Education Building Maintenance Priority Strategy Consider Safety Condition Using Analytical Hierarchy Process (AHP) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |