A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components

Abstract

The present invention belongs to the field of intelligent disaster prevention technology, and specifically relates to a digital twin intelligent operation and maintenance and evaluation system for energy dissipation and shock absorption components. The system includes an intelligent perception module, a fidelity reduction model, and a virtual-reality interactive intelligent decision-making module. By integrating digital twin technology, Internet of Things technology, edge computing and other technologies, it provides solutions for the intelligent operation and maintenance of energy dissipation and shock absorption components. The present invention uses digital twin technology to establish a fidelity virtual twin of the energy dissipation and shock absorption component, and uses virtual sensors to "see through" the overall performance status of the component, providing new technology for the full life cycle operation and maintenance of such components. With the help of this technology, the performance status of the core components of the engineering structure can be grasped in real time, so that it can maintain a full-time online transparent state throughout its life cycle, thereby improving the disaster resilience of the engineering structure as a whole.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent disaster prevention technology, and specifically relates to a digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components. Background Art

[0002] Energy dissipation and shock (vibration) components are widely used in engineering structures and play a core role in improving structural safety and comfort. In the field of earthquake resistance and disaster prevention, energy dissipation and shock absorption components such as mild steel dampers, replaceable coupling beams, buckling restrained braces and viscous dampers have been verified in improving seismic toughness. At the same time, energy dissipation and vibration absorption components such as viscous dampers and viscoelastic dampers also have excellent performance in vibration control and improving comfort. However, real earthquakes in recent years have shown that the performance of energy dissipation and shock absorption components that have been in service for a long time under earthquakes will not achieve the predetermined design goals due to lack of maintenance. More importantly, mastering the performance status of these core components before, during and after the earthquake is of great practical significance for evaluating the seismic toughness level of the structure. For the golden rescue time after the earthquake, if the remaining capacity of the core components can be quickly evaluated, more resources can be provided for disaster relief, such as personnel placement and life rescue. At present, most of these energy dissipation and shock absorption components are in a state of no maintenance or very little maintenance, and the maintenance method mostly adopts the method of visual inspection, which is inefficient and poor in precision. It is impossible to form an objective and continuous historical data archive, and it is even more impossible to detect damage or potential failures inside the components. At the same time, large-scale precision equipment is only used for inspection after an accident or earthquake, and it cannot be used for a long time due to high costs. Therefore, it is very necessary to explore the use of new technologies to establish low-cost and high-efficiency intelligent operation and maintenance methods for energy dissipation and shock absorption components in order to improve the seismic resilience of engineering structures. Summary of the invention

[0003] The present invention discloses a digital twin intelligent operation and maintenance and evaluation system for energy dissipation and shock absorption components, which provides a solution for the intelligent operation and maintenance of energy dissipation and shock absorption components by integrating digital twin technology, Internet of Things technology, edge computing and other technologies.

[0004] To achieve the above purpose, the technical solution of the present invention is:

[0005] A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components, including:

[0006] Intelligent sensing module: installed on the energy dissipation and shock absorption component, detects the physical data of the energy dissipation and shock absorption component in real time through sensors, obtains the key response field of the energy dissipation and shock absorption component through edge computing, and transmits the physical data and key response field to the fidelity reduction model;

[0007] Virtual twin of energy dissipation and shock absorption components - Fidelity reduced-order model: A response snapshot database is established through finite element model calculation, and a Fidelity reduced-order model module of energy dissipation and shock absorption components is further established. The Fidelity reduced-order model module is used to conduct real-time evaluation and visualization of the damage of energy dissipation and shock absorption components.

[0008] Virtual-reality interactive intelligent decision-making module: Receives data from the fidelity reduction model module, makes maintenance decisions on the energy dissipation and shock absorption components based on the real-time assessment and visualization of the damage of the energy dissipation and shock absorption components, and pushes maintenance suggestions and visualization images to the cloud for maintenance personnel to make manual judgments and arrange personnel to conduct on-site maintenance as appropriate; at the same time, it provides an assessment of the remaining capacity of the energy dissipation and shock absorption components and gives suggestions on the temporary application of engineering structures;

[0009] The intelligent perception module, the fidelity reduction model, and the virtual-reality interactive intelligent decision-making module are connected in sequence to form a digital twin intelligent operation and maintenance and evaluation closed-loop system for the energy dissipation and shock absorption components.

[0010] Preferably, the intelligent sensing module includes a sensor, a communication module, and a power module integrated with the single-chip microcomputer mainboard. The sensor is used to detect the three-dimensional deformation and working environment information of the energy dissipation and shock absorption components, and includes a displacement sensor, a gyroscope sensor, and a temperature and humidity sensor; the communication module is a 4G module or a WIFI module or a Bluetooth module, which is used to transmit information to the fidelity reduction model; the single-chip microcomputer is provided with an edge computing module, and is electrically connected to the sensor, the communication module, and the power module through wires respectively; the power module is used for the normal power supply of the intelligent sensing module and the emergency power supply in case of disasters.

[0011] Preferably, in the intelligent sensing module, the key response fields include the stress field, strain field and displacement field of the energy dissipation and shock absorption components.

[0012] Preferably, the fidelity reduced-order model includes: a finite element model, a response snapshot database, a reduced-order algorithm model and a fidelity reduced-order model module. The finite element model is used to establish a parameterized model that reflects the response of the energy dissipation and shock absorption component, calculate the response under various external excitations, and establish a corresponding snapshot database; based on the snapshot database, a reduced-order algorithm model is used to establish a fidelity reduced-order model module.

[0013] Preferably, the method by which the fidelity reduced-order model module performs real-time evaluation and visualization of the damage of the energy dissipation and shock absorption components is: based on the key response field information of the intelligent perception module, the historical damage information and real-time damage information of the energy dissipation and shock absorption components are integrated to locate the damaged area and perform quantitative evaluation, and the damage is evaluated and the location and degree of the damage are visualized based on the provided trained AI model.

[0014] Preferably, the maintenance decisions of the virtual-reality interactive intelligent decision-making module include: continued use without maintenance, use under certain maintenance conditions, and continued use that cannot be performed through maintenance.

[0015] Preferably, the maintenance decision also includes predictive maintenance recommendations. Based on the probability of aftershocks occurring after the main shock, the fidelity reduced-order model module calculates the key response fields of the energy dissipation and shock absorption components when aftershocks occur, and predicts the usage functions of the energy dissipation and shock absorption components. The virtual-reality interactive intelligent decision-making module gives predictive maintenance recommendations based on the function predictions.

[0016] Preferably, the fidelity reduced-order model is modeled by hybrid of POD-Galerkin and Autoencoder methods. First, the snapshot data obtained by finite element model analysis is subjected to singular value decomposition to obtain POD modes, and then POD-Galerkin projection reduction is performed to obtain intermediate reduced-order data. This data is used as input to the Encoder module of the Autoencoder to obtain reduced-order coordinates of the latent space; then, a parameterized reduced-order model is obtained by adopting a sparse regression algorithm, and the parameterized reduced-order model is decoded and reconstructed by the Decoder module of the Autoencoder to obtain a fidelity reduced-order model module.

[0017] Preferably, the sensor data is assimilated by a fidelity reduction model module, and a probabilistic graphical model is used to consider various uncertainties to dynamically update and predict the fidelity reduction model module, and enable the virtual-reality interactive intelligent decision-making module to make corresponding maintenance decisions.

[0018] Preferably, the probabilistic graphical model is a dynamic Bayesian network model including decision nodes, which uses 6 variables to describe the energy dissipation and shock absorption components, including physical state, digital state, key quantities, observation data, control input and evaluation decisions; the probabilistic graphical model is used to drive the interaction between the physical entity and the virtual twin, and the virtual-reality interactive intelligent decision-making module integrates multiple uncertainties to make maintenance decisions based on the prediction of the fidelity reduction model module and the real-time input of sensor data.

[0019] The beneficial effects of the digital twin intelligent operation, maintenance and evaluation system of the energy dissipation and shock absorption components of the present invention are:

[0020] 1. Intelligent operation and maintenance of energy dissipation and shock absorption components throughout their life cycle. Digital twin technology is used to establish a fidelity virtual twin of energy dissipation and shock absorption components, and virtual sensors are used to "see through" the overall performance status of the components, providing new technology for the full life cycle operation and maintenance of such components. With the help of this technology, the performance status of the core components of the engineering structure can be mastered in real time, so that they can remain online and transparent throughout their life cycle, thereby improving the disaster resilience of the engineering structure as a whole.

[0021] 2. The advantage of low cost can promote a large number of applications. With the help of mature low-cost sensor products and single-chip microcomputer products, large-scale deployment can still effectively control costs. This is very convincing for a large number of small and medium-sized owners. At the same time, with the fidelity response of the digital twin, it is avoided to deploy a large number of sensors, especially expensive sensors. The virtual sensor using digital twin technology can view the performance information of any location in real time. This information can be stored and archived in real time for future reference, which is convenient for establishing the health file of the component.

[0022] 3. High timeliness is conducive to post-earthquake rescue. As key components supporting the seismic resistance of engineering structures, energy dissipation and shock absorption components are responsible for dissipating the input energy of the earthquake when an earthquake occurs. After an earthquake, a rapid assessment of the damage and remaining capacity of these key components is crucial to the seismic resistance of engineering structures. Different from traditional post-earthquake assessments that require a large number of professionals and equipment, the rapid assessment technology based on digital twins of the present invention can obtain the performance status of components in real time. This timeliness cannot be achieved by traditional means. At the same time, due to more detailed data support, a quantitative assessment of the traffic capacity of important lifeline projects can be given. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 Schematic diagram of the arrangement of replaceable coupling beams in shear wall structures.

[0024] Figure 2 Schematic diagram of the digital twin of the replaceable coupling beam.

[0025] Figure 3 Schematic diagram of the digital twin of the replaceable wall base energy dissipation and shock absorption components.

[0026] Figure 4 Schematic diagram of the digital twin of a buckling restrained brace or a self-centering brace.

[0027] Figure 5 Schematic diagram of the digital twin of a metal mild steel damper.

[0028] Figure 6 Schematic diagram of the digital twin of the viscoelastic damper.

[0029] Figure 7 Schematic diagram of the digital twin of the viscous damper.

[0030] Figure 8 This is a schematic diagram of the layout of the energy dissipation and shock absorption components of the self-resetting frame.

[0031] Fig. 9 Schematic diagram of the digital twin of the self-righting frame energy dissipation and shock absorption components.

[0032] Fig.10 Schematic diagram of the digital twin of the bridge's energy dissipation and shock absorption components.

[0033] Fig.11 Schematic diagram of the digital twin of the energy dissipation and shock absorption components of the rocking-frame structure.

[0034] Fig.12 This is the framework diagram of the digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components.

[0035] In the figure: 1. Shear wall limb; 2. Replaceable wall foot; 3. Replaceable connecting beam; 4. Connecting steel beam; 5. End plate 1; 6. End plate 2; 7. IoT edge computing module; 8. Sensor 1; 9. Shear wall root part; 10. Sensor 2; 11. Buckling restraint support or self-resetting support; 12. Column; 13. Beam; 14. Sensor 3; 15. Data line; 16. Sensor 4; 17. Upper end plate; 18. Soft steel damper; 19. Sensor 5; 20. Lower end plate; 21. Sensor six; 22. Viscoelastic damper; 23. Viscous damper; 24. Sensor seven; 25. Frame beam; 26. Frame column; 27. Anchor block; 28. Prestressed cable; 29. ​​Replaceable energy dissipation and vibration reduction components; 30. Sensor eight; 31. Upper flange plate; 32. Lower flange plate; 33. Web plate; 34. Precast beam; 35. Seismic isolation bearing; 36. Sensor nine; 37. Bridge pier; 38. Sensor ten; 39. Rocking wall or rocking frame. DETAILED DESCRIPTION

[0036] The following description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0037] The following embodiments may be understood as individually expressing a part of a local structure or method of the present invention, or may be understood as a combination of the embodiments to explain the connotation of a larger structure or method of the present invention.

[0038] In the initial embodiment, the present invention provides a digital twin intelligent operation and maintenance and evaluation system for energy dissipation and shock absorption components, such as Fig.12 As shown, including:

[0039] Intelligent sensing module: installed on the energy dissipation and shock absorption component, detects the physical data of the energy dissipation and shock absorption component in real time through sensors, obtains the key response field of the energy dissipation and shock absorption component through edge computing, and transmits the physical data and key response field to the fidelity reduction model;

[0040] Virtual twin of energy dissipation and shock absorption components - Fidelity reduced-order model: A response snapshot database is established through finite element model calculation, and a Fidelity reduced-order model module of energy dissipation and shock absorption components is further established. The Fidelity reduced-order model module is used to conduct real-time evaluation and visualization of the damage of energy dissipation and shock absorption components.

[0041] Virtual-reality interactive intelligent decision-making module: Receives data from the fidelity reduction model module, makes maintenance decisions on the energy dissipation and shock absorption components based on the real-time assessment and visualization images of the damage of the energy dissipation and shock absorption components, and pushes maintenance suggestions and visualization images to the cloud for maintenance personnel to complete manual judgment and arrange personnel to conduct on-site maintenance as appropriate; at the same time, gives an assessment of the remaining capacity of the energy dissipation and shock absorption components and gives suggestions on the temporary application of engineering structures (for example, some lifeline projects (important bridges) are less damaged and have very important functions, so they can be used appropriately under real-time monitoring, which has important practical value for disaster relief);

[0042] The intelligent perception module, the fidelity reduction model, and the virtual-reality interactive intelligent decision-making module are connected in sequence to form a digital twin intelligent operation and maintenance and evaluation closed-loop system for the energy dissipation and shock absorption components.

[0043] In this embodiment, the intelligent sensing module is installed on the energy dissipation and shock absorption component, such as Figure 1-11 As shown, a schematic diagram of installing energy dissipation and shock absorption components in various building structures is given. At the same time, it also shows the principle of coordinated cooperation between the intelligent perception module, the fidelity reduction model, and the virtual-reality interactive intelligent decision-making module.

[0044] In a further embodiment, the intelligent sensing module includes a sensor, a communication module, and a power module integrated with the single-chip microcomputer mainboard. The sensor is used to detect the three-dimensional deformation and working environment information of the energy dissipation and shock absorption components, including a displacement sensor, a gyroscope sensor, and a temperature and humidity sensor; the communication module is a 4G module or a WIFI module or a Bluetooth module, which is used to transmit information to the fidelity reduction model; the single-chip microcomputer is provided with an edge computing module, and is electrically connected to the sensor, communication module, and power module through wires respectively; the power module is used for the normal power supply of the intelligent sensing module and the emergency power supply in case of disasters.

[0045] In this embodiment, due to the different types of sensors, the installation methods and locations of various sensors are prior art and are not described in detail here. The edge computing module is used to calculate the key response field of the energy dissipation and shock absorption component.

[0046] In a further embodiment, in the intelligent sensing module, Figure 1-11 As shown, the key response fields include the stress field, strain field and displacement field of the energy dissipation and shock absorption components.

[0047] In a further embodiment, the fidelity reduced-order model includes: a finite element model, a response snapshot database, a reduced-order algorithm model and a fidelity reduced-order model module. The finite element model is used to establish a parameterized model that reflects the response of the energy dissipation and shock absorption component, calculate the response under various external excitations, and establish a corresponding snapshot database; according to the snapshot database, the reduced-order algorithm model is used to establish a fidelity reduced-order model module. Compared with the finite element model, the fidelity reduced-order model can realize the rapid calculation of the response field (such as stress field, strain field and displacement field) while maintaining a high degree of accuracy, providing support for the high timeliness of rescue.

[0048] In a further embodiment, the method for the fidelity reduced-order model module to perform real-time evaluation and visualization of the damage of the energy dissipation and shock absorption components is: based on the key response field information of the intelligent perception module, the historical damage information and real-time damage information of the energy dissipation and shock absorption components are integrated to locate the damaged area and perform quantitative evaluation, and the damage is evaluated and the damage location and degree are visualized according to the provided trained AI model.

[0049] In a further embodiment, the maintenance decisions of the virtual-reality interactive intelligent decision-making module include: continued use without maintenance, use under certain maintenance conditions, and continued use that cannot be performed through maintenance.

[0050] In a further embodiment, the maintenance decision also includes predictive maintenance suggestions. Based on the probability of aftershocks after the main shock, the fidelity reduction model module calculates the key response field of the energy dissipation and shock absorption components if aftershocks occur, and predicts the use function of the energy dissipation and shock absorption components. The virtual-reality interactive intelligent decision module gives predictive maintenance suggestions based on the function prediction. Such predictive maintenance suggestions can provide reserve information for emergency response of disaster relief, which helps to improve the efficiency of rescue.

[0051] In a further embodiment, the fidelity reduction model adopts a hybrid modeling of POD-Galerkin and Autoencoder methods, firstly, the snapshot data obtained by finite element model analysis is subjected to singular value decomposition to obtain POD mode, and then POD-Galerkin projection reduction is performed to obtain intermediate reduced order data, and this data is used as input to the Encoder module of the Autoencoder to obtain the reduced order coordinates of the latent space, and accordingly, a parameterized reduced order model is obtained by adopting a sparse regression algorithm, and the parameterized reduced order model is decoded and reconstructed by the Decoder module of the Autoencoder to obtain a fidelity reduction model module. Different from the black box neural network model, the present invention reduces the data to the latent space, and obtains a reduced order model with strong interpretability through regression, which is a reduced order model with high accuracy, strong generalization ability and interpretability.

[0052] In a further embodiment, the sensor data is assimilated by a fidelity reduction model module, and a probabilistic graphical model is used to consider multiple uncertainties to dynamically update and predict the fidelity reduction model module, and enable the virtual-reality interactive intelligent decision-making module to make corresponding maintenance decisions.

[0053] In a further embodiment, the probabilistic graphical model is a dynamic Bayesian network model including decision nodes, which uses 6 variables to describe the energy dissipation and shock absorption components, including physical state, digital state, key quantities, observation data, control input and evaluation decisions; the probabilistic graphical model is used to drive the interaction between the physical entity and the virtual twin, and according to the prediction of the fidelity reduction model module and the real-time input of sensor data, the virtual-reality interactive intelligent decision-making module integrates multiple uncertainties to make maintenance decisions.

[0054] In this embodiment, six variables are used to describe the energy dissipation and shock absorption components, among which the physical state is such as the material properties of the steel plate of the replaceable connecting beam, the height of the flange and the web, the thickness of the steel, the presence or absence of damage, etc.; the digital state is the representation of the physical entity calculation model, such as the calculation model parameters, initial conditions, boundary conditions and mesh division of the replaceable connecting beam, etc., the key quantity is the output of the fidelity reduced-order model, such as the stress field, strain field or displacement field of the replaceable connecting beam, the observation data is the data obtained from the physical entity by the intelligent perception module, such as sensor data, inspection data, etc., the control input refers to the response made by the operation and maintenance personnel according to the fidelity reduced-order model and the observation data, such as manual inspection, component replacement, etc., and the evaluation decision refers to the comprehensive observation data, the fidelity reduced-order model and the control input, and the probability graphical model is used to make inference decisions to quantitatively evaluate the health status and service capability of the energy dissipation and shock absorption components, such as whether the replaceable connecting beam is no longer suitable for continued use.

[0055] Working principle of the present invention: Aiming at the pain point of the operation and maintenance of energy dissipation and shock absorption components throughout their life cycle, the present invention adopts digital twin technology and Internet of Things technology to propose intelligent operation and maintenance and rapid evaluation technology for normal use scenarios and earthquake scenarios within the entire life cycle of the components. The digital twin architecture of the present invention adopts the form of a triple, that is, the real-virtual-real closed loop of intelligent perception of the physical entity of the energy dissipation and shock absorption components-fidelity virtual twin mapping-virtual-real interactive intelligent decision-making. First, the real-time state of the energy dissipation and shock absorption components is perceived by the sensors arranged at their local positions, and transmitted to the data storage unit of the single-chip microcomputer through the data transmission line for storage and recording. The information is regularly summarized and uploaded to the cloud through the Internet of Things. On the other hand, the sensor data is transmitted to the fidelity reduced-order model. Secondly, the fidelity reduced-order model of the energy dissipation and shock absorption components is a virtual mapping of its physical entity, accepting a small amount of sensor data as input. With the help of the computing power of the single-chip microcomputer, it can quickly calculate the high-precision response of the entire component in real time, such as the stress field, strain field and displacement field, which are very important for evaluating damage and residual capacity. Therefore, the overall response is obtained through a small number of sensors arranged at the local position of the component through the digital twin model. By using these overall responses as virtual sensors, we can "see through" the damage status inside the components and avoid the drawbacks of using heavy flaw detection equipment, which may not be available for a long time after the earthquake. Finally, the virtual-reality interactive intelligent decision-making module conducts comprehensive reasoning on the sensor data and the calculated responses obtained by the fidelity digital twin reduced-order model. The deep learning model trained offline is used to perform qualitative classification and quantitative grading of damage, and the intelligent decision-making uses a probabilistic graphical model for reasoning and decision-making to give the final functional status of the component.

[0056] like Figure 1-11 As shown, the energy dissipation and shock absorbing components of the present invention include but are not limited to various metal dampers, replaceable components such as replaceable connecting beams, viscous dampers, viscoelastic dampers and buckling restrained supports, which are key components used for shock / vibration absorption in engineering.

[0057] The fidelity reduction model of the present invention adopts the POD-Galerkin and Autoencoder hybrid reduction algorithm. First, the high-precision finite element model snapshot data with extremely high degrees of freedom (ranging from tens of thousands to hundreds of thousands) is subjected to singular value decomposition to obtain the first r-order POD modes containing 95% of the energy, and the snapshot data is linearly reduced in the basis space composed of POD modes. At this time, the degrees of freedom of the POD linear reduction model are sharply reduced, ranging from tens to thousands.

[0058] Furthermore, based on the POD linear reduced-order model, the reduced-order data is input into the Encoder module of the Autoencoder. The output of the Encoder module is a nonlinear latent space with very few degrees of freedom (ranging from several to more than a dozen) spanned by latent variables. Thus, the POD reduced-order model data is further reduced through the Encoder module. In this latent space, the sparse regression algorithm can be used to obtain the structural dynamics expressed by ordinary differential equations (groups). These differential equations (groups) have high physical interpretability due to the small number of variables. Then, the POD reduced-order model is reconstructed through the Decoder module of the Autoencoder to obtain the fidelity reduced-order model data.

[0059] Furthermore, the Autoencoder model used is composed of two modules, Encoder and Decoder. Depending on the data being processed, the two modules can use different multi-layer neural network combinations, such as fully connected neural network MLP, convolutional neural network CNN, recurrent neural network RNN, etc. The number of neurons in each layer is optimized and adjusted according to the specific application. The number of neurons at the output of the Encoder module or the input of the Decoder module needs to be adjusted according to the accuracy requirements.

[0060] Furthermore, in the virtual-reality interactive intelligent decision-making module, the sensor data is assimilated through the reduced-order model to obtain the corrected response field (deformation, stress or strain field, etc.) of the energy dissipation and shock absorption components, which is input into the damage detection deep neural network model trained offline in advance to locate and quantify the damage, analyze the damage status of the components in detail, and provide support for the comprehensive decision-making of the probabilistic graphical model.

Claims

1. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components, characterized by: include: Intelligent sensing module: installed on the energy dissipation and shock absorption component, detects the physical data of the energy dissipation and shock absorption component in real time through sensors, obtains the key response field of the energy dissipation and shock absorption component through edge computing, and transmits the physical data and key response field to the fidelity reduction model; Virtual twin of energy dissipation and shock absorption components - Fidelity reduced-order model: A response snapshot database is established through finite element model calculation, and a Fidelity reduced-order model module of energy dissipation and shock absorption components is further established. The Fidelity reduced-order model module is used to conduct real-time evaluation and visualization of the damage of energy dissipation and shock absorption components. Virtual-reality interactive intelligent decision-making module: Receives data from the fidelity reduction model module, makes maintenance decisions on the energy dissipation and shock absorption components based on the real-time assessment and visualization of the damage of the energy dissipation and shock absorption components, and pushes maintenance suggestions and visualization images to the cloud for maintenance personnel to make manual judgments and arrange personnel to conduct on-site maintenance as appropriate; at the same time, it provides an assessment of the remaining capacity of the energy dissipation and shock absorption components and gives suggestions on the temporary application of engineering structures; The intelligent perception module, the fidelity reduction model, and the virtual-reality interactive intelligent decision-making module are sequentially connected to form a digital twin intelligent operation and maintenance and evaluation closed-loop system for energy dissipation and shock absorption components; The fidelity reduced-order model includes: a finite element model, a response snapshot database, a reduced-order algorithm model and a fidelity reduced-order model module. The finite element model is used to establish a parameterized model reflecting the response of the energy dissipation and shock absorption component, calculate the response under various external excitations, and establish a corresponding snapshot database; According to the snapshot database, a reduced-order algorithm model is used to establish a fidelity reduced-order model module; The method for the fidelity reduced-order model module to perform real-time evaluation and visualization of the damage of the energy dissipation and shock absorption components is: based on the key response field information of the intelligent perception module, the historical damage information and real-time damage information of the energy dissipation and shock absorption components are integrated to locate the damaged area and perform quantitative evaluation, and the damage is evaluated and the damage location and degree are visualized according to the provided trained AI model; The fidelity reduced-order model is modeled by hybrid modeling of POD-Galerkin and Autoencoder methods. First, the snapshot data obtained by finite element model analysis is subjected to singular value decomposition to obtain POD modes, and then POD-Galerkin projection reduction is performed to obtain intermediate reduced-order data. This data is used as input to the Encoder module of the Autoencoder to obtain reduced-order coordinates of the latent space; then, a parameterized reduced-order model is obtained by adopting a sparse regression algorithm, and the parameterized reduced-order model is decoded and reconstructed by the Decoder module of the Autoencoder to obtain a fidelity reduced-order model module.

2. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components as claimed in claim 1, characterized in that: The intelligent sensing module includes sensors, communication modules, and power modules integrated with the single-chip microcomputer mainboard. The sensors are used to detect the three-dimensional deformation and working environment information of the energy dissipation and shock absorption components, including displacement sensors, gyroscope sensors, and temperature and humidity sensors; the communication module is a 4G module or a WIFI module or a Bluetooth module, which is used to transmit information to the fidelity reduction model; the single-chip microcomputer is provided with an edge computing module, and is electrically connected to the sensor, communication module, and power module through wires respectively; the power module is used for the normal power supply of the intelligent sensing module and the emergency power supply in disasters.

3. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components as claimed in claim 2, characterized in that: In the intelligent sensing module, the key response fields include the stress field, strain field and displacement field of the energy dissipation and shock absorption components.

4. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components as claimed in claim 3, characterized in that: The maintenance decisions of the virtual-reality interactive intelligent decision-making module include: continued use without maintenance, use under certain maintenance conditions, and continued use that cannot be maintained.

5. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components as claimed in claim 4, characterized in that: The maintenance decision also includes predictive maintenance recommendations. Based on the probability of aftershocks occurring after the main shock, the fidelity reduced-order model module calculates the key response fields of the energy dissipation and shock absorption components when aftershocks occur, and predicts the usage functions of the energy dissipation and shock absorption components. The virtual-reality interactive intelligent decision-making module gives predictive maintenance recommendations based on the function predictions.

6. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components as claimed in claim 5, characterized in that: The sensor data is assimilated by the fidelity reduction model module, and a probabilistic graphical model is used to consider multiple uncertainties to dynamically update and predict the fidelity reduction model module, and enable the virtual-reality interactive intelligent decision-making module to make corresponding maintenance decisions.

7. A digital twin intelligent operation, maintenance and evaluation system for energy dissipation and shock absorption components as claimed in claim 6, characterized in that: The probabilistic graphical model is a dynamic Bayesian network model including decision nodes, which uses 6 variables to describe the energy dissipation and shock absorption components, including physical state, digital state, key quantities, observation data, control input and evaluation decision; the probabilistic graphical model is used to drive the interaction between the physical entity and the virtual twin. According to the prediction of the fidelity reduction model module and the real-time input of sensor data, the virtual-reality interactive intelligent decision-making module integrates multiple uncertainties to make maintenance decisions.

Images