CN115168934A - Building structure parameter identification and earthquake response prediction technology based on deep learning - Google Patents
Building structure parameter identification and earthquake response prediction technology based on deep learning Download PDFInfo
- Publication number
- CN115168934A CN115168934A CN202210589691.8A CN202210589691A CN115168934A CN 115168934 A CN115168934 A CN 115168934A CN 202210589691 A CN202210589691 A CN 202210589691A CN 115168934 A CN115168934 A CN 115168934A
- Authority
- CN
- China
- Prior art keywords
- building structure
- acceleration
- parameter identification
- deep learning
- loss function
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Computational Mathematics (AREA)
- Artificial Intelligence (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Structural Engineering (AREA)
- Civil Engineering (AREA)
- Architecture (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Analysis (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Buildings Adapted To Withstand Abnormal External Influences (AREA)
Abstract
The invention discloses a building structure parameter identification and earthquake response prediction technology based on deep learning, which comprises the following steps: A. constructing an implicit relation between acceleration excitation a and building structure acceleration response by adopting a multi-layer multi-channel long-short term memory neural network, expressing trainable parameters of the network by theta, and adding residual connection to enhance the learning capacity of the model; B. and integrating the acceleration predicted by the network by using a Newmark-beta method to obtain the speed and displacement response at different moments. Compared with the prior art, the invention has the advantages that: the method can realize building structure parameter identification and earthquake response prediction, and is used for building damage identification and safety state evaluation.
Description
Technical Field
The invention relates to the field of building safety, in particular to a building structure parameter identification and earthquake response prediction technology based on deep learning.
Background
The health monitoring of the building by using the sensing technology is one of important means for evaluating the safety state of the structure. Particularly, under disaster conditions such as earthquakes, how to effectively use the measurement data (such as acceleration response) of the sensor to identify important structural parameters (such as damping and interlayer rigidity) is the basis for quantitatively evaluating the structural state and diagnosing damage. The existing structural parameter identification method is mainly based on data assimilation, and an optimization strategy is applied to a given parameterized model, so that unknown parameters are identified. Although this method has proven effective, it is less efficient in processing large structural systems, for example when the structural degrees of freedom reach hundreds, thousands, or even more, and the computational efficiency and accuracy of the method are challenging. In addition, how to establish an effective and efficient proxy model by using the measurement data to realize the real-time prediction of the structural seismic response is also a key problem.
Therefore, research on building structure parameter identification and seismic response prediction technologies based on deep learning is urgently needed.
Disclosure of Invention
The invention aims to solve the technical problem of providing a building structure parameter identification and earthquake response prediction technology based on deep learning.
In order to solve the technical problems, the technical scheme provided by the invention is as follows: a building structure parameter identification and seismic response prediction technology based on deep learning comprises the following steps:
A. constructing an implicit relation between acceleration excitation a and building structure acceleration response by adopting a multi-layer multi-channel long-short term memory neural network, expressing trainable parameters of the network by theta, and adding residual connection to enhance the learning capacity of the model;
B. and integrating the acceleration predicted by the network by using a Newmark-beta method to obtain the speed and displacement responses at different moments.
As an improvement, the technology comprises the following specific steps:
seismic dynamic acceleration at different moments is input, and structural acceleration responses at different moments are obtained through a multi-channel long-short term memory neural network;
structural acceleration responses at different moments are integrated by a Newmark-beta method to obtain structural displacement and speed responses at different moments;
respectively substituting the structural displacement and the speed response at different moments into a physical loss function and a data loss function, and substituting the structural acceleration response at different moments into the data loss function;
a loss function is derived.
As an improvement, the loss function for training the network comprises data loss and physical loss, wherein the data loss function is formed by the mean square error of the predicted acceleration and the measured acceleration, and the physical loss function is formed by a residual version of a shear dynamic model equation.
As an improvement, the trainable parameters of the whole network comprise structural damping, rigidity and multi-channel long-short term memory neural network parameters.
As an improvement, based on the measurement data, after the network is trained, the damping and stiffness parameters of the structure can be identified.
As an improvement, the integral calculation formula of the Newmark-beta method is as follows:
as an improvement, the physical loss function formula is:
as an improvement, the data loss function formula is:
compared with the prior art, the invention has the advantages that: the method can realize building structure parameter identification and earthquake response prediction, and is used for building damage identification and safety state evaluation.
Drawings
FIG. 1 is a schematic structural diagram of a building structure parameter identification and seismic response prediction technology based on deep learning.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
When the invention is implemented specifically, the building structure parameter identification and earthquake response prediction technology based on deep learning comprises the following steps:
A. constructing an implicit relation between acceleration excitation a and building structure acceleration response by adopting a multi-layer multi-channel long-short term memory neural network, expressing trainable parameters of the network by theta, and adding residual connection to enhance the learning capacity of the model;
B. and integrating the acceleration predicted by the network by using a Newmark-beta method to obtain the speed and displacement response at different moments.
As an improvement, the technology comprises the following specific steps:
inputting earthquake dynamic acceleration at different moments, and obtaining structural acceleration responses at different moments through a multi-channel long-short term memory neural network;
structural acceleration responses at different moments are integrated by a Newmark-beta method to obtain structural displacement and speed responses at different moments;
respectively substituting the structural displacement and the speed response at different moments into a physical loss function and a data loss function, and substituting the structural acceleration response at different moments into the data loss function;
a loss function is derived.
As an improvement, the loss function for training the network comprises data loss and physical loss, wherein the data loss function is formed by the mean square error of the predicted acceleration and the measured acceleration, and the physical loss function is formed by a residual version of a shear dynamic model equation.
As an improvement, the trainable parameters of the whole network comprise structural damping, rigidity and multi-channel long-short term memory neural network parameters.
As an improvement, based on the measurement data, after the network is trained, the damping and stiffness parameters of the structure can be identified.
As an improvement, the integral calculation formula of the Newmark-beta method is as follows:
as an improvement, the physical loss function formula is:
as an improvement, the data loss function formula is:
the working principle of the invention is as follows: a building structure parameter identification and seismic response prediction technology based on deep learning is mainly composed of a deep recurrent neural network based on physical information.
Considering a typical building structure, the corresponding shear dynamic model of which is assumed to be known, the parameters include a damping matrix C and a stiffness matrix K, the seismic response of which includes displacement, velocity, acceleration in various degrees of freedom, and the measurement data is composed of acceleration time courses measured by acceleration sensors arranged on different floors.
The invention adopts a multilayer multi-channel long and short term memory neural network (LSTM) to construct the implicit relation between the acceleration excitation a and the building structure acceleration response, trainable parameters of the network are expressed by theta, and residual connection is added to enhance the learning capability of the model. And then, integrating the acceleration predicted by the network by using a Newmark-beta method to obtain the speed and displacement response at different moments. The loss function for training the network comprises data loss and physical loss, wherein the data loss function is formed by predicted and measured acceleration mean square error, and the physical loss function is formed by a shear dynamic model equation residual error model. Trainable parameters of the entire network include structural damping, stiffness, and LSTM network parameters. Based on the measurement data, after the network is trained, the damping and rigidity parameters of the structure can be identified; meanwhile, the trained neural network model can be used for structure response prediction under other seismic excitation.
Under the structure monitoring scene, the method can be used for building structure parameter identification and earthquake response prediction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the invention, "plurality" means two or more unless explicitly defined otherwise.
In the present invention, unless otherwise specifically stated or limited, the terms "mounted," "connected," "fixed," and the like are to be construed broadly and may, for example, be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood according to specific situations by those of ordinary skill in the art.
In the present invention, unless expressly stated or limited otherwise, the recitation of a first feature "on" or "under" a second feature may include the recitation of the first and second features being in direct contact, and may also include the recitation that the first and second features are not in direct contact, but are in contact via another feature between them. Also, the first feature being "on," "above" and "over" the second feature includes the first feature being directly on and obliquely above the second feature, or merely indicating that the first feature is at a higher level than the second feature. A first feature being "under," "below," and "beneath" a second feature includes the first feature being directly under and obliquely below the second feature, or simply meaning that the first feature is at a lesser elevation than the second feature.
In the description herein, reference to the terms "one embodiment," "some embodiments," "an example," "a specific example," or "some examples" or the like means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made in the above embodiments by those of ordinary skill in the art without departing from the principle and spirit of the present invention.
Claims (8)
1. A building structure parameter identification and seismic response prediction technology based on deep learning is characterized by comprising the following steps:
A. constructing an implicit relation between acceleration excitation a and building structure acceleration response by adopting a multi-layer multi-channel long-short term memory neural network, expressing trainable parameters of the network by theta, and adding residual connection to enhance the learning capacity of the model;
B. and integrating the acceleration predicted by the network by using a Newmark-beta method to obtain the speed and displacement response at different moments.
2. The building structure parameter identification and seismic response prediction technology based on deep learning of claim 1, characterized in that the technology comprises the following specific steps:
inputting earthquake dynamic acceleration at different moments, and obtaining structural acceleration responses at different moments through a multi-channel long-short term memory neural network;
structural acceleration responses at different moments are integrated by a Newmark-beta method to obtain structural displacement and speed responses at different moments;
respectively substituting the structural displacement and the speed response at different moments into a physical loss function and a data loss function, and substituting the structural acceleration response at different moments into the data loss function;
a loss function is derived.
3. The building structure parameter identification and seismic response prediction technology based on deep learning of claim 1, wherein: the loss function for training the network comprises data loss and physical loss, wherein the data loss function is formed by predicted and measured acceleration mean square error, and the physical loss function is formed by a shear dynamic model equation residual error model.
4. The building structure parameter identification and seismic response prediction technology based on deep learning of claim 1, wherein: trainable parameters of the entire network include structural damping, stiffness, and multi-channel long and short term memory neural network parameters.
5. The building structure parameter identification and seismic response prediction technology based on deep learning of claim 1, wherein: based on the measured data, after the network is trained, the damping and rigidity parameters of the structure can be identified.
6. The building structure parameter identification and seismic response prediction technology based on deep learning of claim 2, wherein a Newmark-beta method integral calculation formula is as follows:
7. the building structure parameter identification and seismic response prediction technology based on deep learning of claim 2, wherein the physical loss function formula is as follows:
8. the deep learning based building structure parameter identification and seismic response prediction technique of claim 2, wherein the data loss function formula is:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210589691.8A CN115168934A (en) | 2022-05-26 | 2022-05-26 | Building structure parameter identification and earthquake response prediction technology based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210589691.8A CN115168934A (en) | 2022-05-26 | 2022-05-26 | Building structure parameter identification and earthquake response prediction technology based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115168934A true CN115168934A (en) | 2022-10-11 |
Family
ID=83482856
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210589691.8A Withdrawn CN115168934A (en) | 2022-05-26 | 2022-05-26 | Building structure parameter identification and earthquake response prediction technology based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115168934A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116011125A (en) * | 2023-03-23 | 2023-04-25 | 成都理工大学 | Response prediction method for uncertain axle coupling system |
-
2022
- 2022-05-26 CN CN202210589691.8A patent/CN115168934A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116011125A (en) * | 2023-03-23 | 2023-04-25 | 成都理工大学 | Response prediction method for uncertain axle coupling system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Eftekhar Azam et al. | Damage detection in structural systems utilizing artificial neural networks and proper orthogonal decomposition | |
| JP5809174B2 (en) | Building safety verification system, building safety verification method and program | |
| Cho et al. | Smart wireless sensor technology for structural health monitoring of civil structures | |
| Noh et al. | Development of fragility functions as a damage classification/prediction method for steel moment‐resisting frames using a wavelet‐based damage sensitive feature | |
| Hernandez-Garcia et al. | An experimental investigation of change detection in uncertain chain-like systems | |
| CN102507121B (en) | Building structure seismic damage assessment system and method based on wireless sensor network | |
| CN104457681B (en) | Girder structure dynamic deflection monitoring method based on strain mode | |
| Balmes et al. | Statistical model‐based damage localization: A combined subspace‐based and substructuring approach | |
| CN109029589A (en) | Bridge structures safety condition monitoring system | |
| CN115168934A (en) | Building structure parameter identification and earthquake response prediction technology based on deep learning | |
| DE112012006296T5 (en) | Message based on an event identified from vibration data | |
| CN110006676A (en) | A kind of building performance appraisal procedure and system based on monitoring structural health conditions data | |
| Entezami et al. | Improving feature extraction via time series modeling for structural health monitoring based on unsupervised learning methods | |
| CN114707632A (en) | Sensor network sensor fault positioning method, system, equipment and medium | |
| Mantawy et al. | Convolutional neural network based structural health monitoring for rocking bridge system by encoding time‐series into images | |
| CN115994320A (en) | Intelligent friction pendulum vibration isolation support and state monitoring and fault diagnosis system | |
| Smarsly et al. | Autonomous structural condition monitoring based on dynamic code migration and cooperative information processing in wireless sensor networks | |
| CN111479235B (en) | Formalization method for detecting wireless sensor network node | |
| CN111255220A (en) | Safety monitoring method for super high-rise building construction steel platform formwork equipment | |
| JP5799183B2 (en) | Building safety verification system, building safety verification method and program | |
| KR102356403B1 (en) | Method and apparatus for predicting deformed shape of building | |
| CN202382915U (en) | Building structure shaking-damage evaluating system based on wireless sensing net | |
| Swartz et al. | Structural health monitoring system with the latest information technologies | |
| Dong et al. | High-speed heterogeneous data acquisition using Martlet-A next-generation wireless sensing node | |
| Pal et al. | Structural health monitoring using neural networks in IoT and cps paradigm—a review |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20221011 |