WO2023121612A1

WO2023121612A1 - A digital twin system for predicting response of structures to natural disasters

Info

Publication number: WO2023121612A1
Application number: PCT/TR2022/051180
Authority: WO
Inventors: Eda EROL
Original assignee: Erol Eda
Priority date: 2021-12-22
Filing date: 2022-10-25
Publication date: 2023-06-29
Also published as: TR2021020723A2

Abstract

The invention relates to a computer-based digital twin system for predicting the response of a structure (200) to natural disasters. Accordingly, it is characterized in that it comprises a plurality of structure sensors (210) to detect at least one structure parameter of the multiple components; at least one environmental sensor (220) to measure at least one environmental parameter; an information processing device (100) configured to receive the signals generated regarding the measurements made by the said environmental sensor (220) and the said structure sensors (210); the said information processing device (100) comprises an information processing device (100) configured to access a virtual structure model (500) of the structure (200) and simulate the responses of the components of the said structure (200) in the real world according to the said structure parameters and environmental parameters in the said virtual structure model (500); the said information processing device (100) is configured to access a trained machine learning model using the environmental parameter measurements and the structure parameters measurements at the time these parameters are measured; to apply the test environment parameter values, which are in magnitude measured when the natural disaster occurs and are predetermined, to the machine learning model as input and to determine the damage status of the structure (200) according to the response structure parameters of the machine learning model to this and to simulate the response structure parameters in the virtual structure model (500).

Description

A DIGITAL TWIN SYSTEM FOR PREDICTING RESPONSE OF STRUCTURES TO NATURAL DISASTERS

TECHNICAL FIELD

The invention relates to a computer-based digital twin system for predicting the response of a structure to natural disasters.

BACKGROUND

Seismic fluctuations in the earth's crust as a result of energy emerging at an unexpected moment and the shock of these waves to the earth are called earthquakes. Earthquakes can shake the earth and damage the structures on the earth. However, although it is in the same region and the same intensity, the damage caused by the earthquake to different structures may vary. For this reason, there is a need to predict the earthquake resistance of the structures built on the earth.

Many parameters are needed to predict the earthquake resistance of structures. The said parameters include information such as past earthquake data, past damage data, technical data of the structures damaged in the past, the natural period of the structure whose strength is to be predicted, and structural and positional data. The natural period of a structure is the vibration period due to the weight of the said structure and the rigidity of the structural system against horizontal loads.

In the state of the art, the patent document JP20170095088 discloses a structure effect prediction system for predicting the effect of seismic waves on a structure. The said structure effect prediction system has a structure that records the structure information, including the location information of the structure, the natural period, and the accuracy of the natural period. The structure effect prediction system includes an earthquake database that records earthquake information including the period of earthquake waves, a prediction unit to predict the degree of effect of structures due to earthquake motion on a structural basis based on the natural period of the structure and the period of the earthquake wave, and a notification unit to display the degree of effect of the structure on the map based on the location information of the structure and the accuracy of the natural period. The structure effect prediction system can calculate the degree of effect of an earthquake for any structure by using the location and structural data of the structure entered through a display and the historical earthquake data.

In general, the effect of an earthquake event or other natural disasters on a structure can be calculated based on known historical data and known data about the structure. However, the accessibility of the said data may be limited. It may not be possible for a person to access the historical earthquake data or to obtain the technical data of the structure. In addition, the technical characteristics of the structure may vary over time. Over time, there may be changes in the soil and the components of the structure, and natural disaster instability situations caused by such changes can only be detected by special inspection or external observation.

All the problems mentioned above have made it necessary to make an innovation in the relevant technical field as a result.

BRIEF DESCRIPTION OF THE INVENTION

The present invention relates to a digital twin system in order to eliminate the above- mentioned disadvantages and bring new advantages to the related technical field.

It is an object of the invention to provide a system for predicting the natural disaster resistance of structures.

Another object of the invention is to provide a system with improved accuracy of prediction.

In order to achieve all the objectives that will emerge from the abovementioned and the following detailed description, the present invention is a computer-based digital twin system for predicting the response of a structure to natural disasters. Accordingly, its novelty is that it comprises a plurality of structure sensors to detect at least one structure parameter of the multiple components; at least one environmental sensor to measure at least one environmental parameter; an information processing device configured to receive the signals generated regarding the measurements made by the said environmental sensor and the said structure sensors; the said information processing device comprises an information processing device configured to access a virtual structure model of the structure and simulate the responses of the components of the said structure in the real world according to the said structure parameters and environmental parameters in the said virtual structure model; the said information processing device is configured to access a trained machine learning model using the environmental parameter measurements and the structure parameters measurements at the time these parameters are measured; to apply the test environment parameter values, which are in magnitude measured when the natural disaster occurs and are predetermined, to the machine learning model as input and determine the damage status of the structure according to the response structure parameters of the machine learning model to this and simulate the response structure parameters in the virtual structure model. Thus, it is ensured that the parameters of the structures are observed while they are under construction or in use after the completion of the construction and the response to a possible natural disaster is determined with the machine learning model; thus, the structures that may be damaged as a result of the natural disaster are determined in advance.

A possible embodiment of the invention is characterized in that the information processing device is configured to simulate the virtual structure model according to the instant environmental parameter measurements and structure parameter measurements it receives to show the instant state of the structure. Thus, it is ensured that the current state of the structure is observed instantly.

Another possible embodiment of the invention is characterized in that it is configured to display the simulations made in the said virtual structure model on a user interface.

Another possible embodiment of the invention is characterized in that the said sensors are at least one of the strain sensors, force sensors, acceleration sensors, temperature sensors, and humidity sensors.

Another possible embodiment of the invention is characterized in that the said environmental sensors are at least one of the seismograph, humidity sensor, and temperature sensor.

Another possible embodiment of the invention is characterized in that the processor unit is configured to train the machine learning model with the structure parameters obtained in response to the test environment parameters values. Thus, machine learning is allowed to be self-trained and results can be obtained with increased accuracy.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 shows a representative view of the digital twin system.

Figure 2 shows a representative view of the physical structure, the virtual structure model, and the machine item.

Figure 3 shows a representative view of the information processing device.

DETAILED DESCRIPTION OF THE INVENTION

In this detailed description, the subject matter of the invention is described by using examples only for a better understanding, which will have no limiting effect.

Referring to Figure 1 , the present invention is a digital twin system comprising a plurality of structure sensors (210) associated with a plurality of components of a structure (200) and measuring at least one parameter of these components; at least one environmental sensor (220) measuring at least one environmental parameter; an information processing device (100) that receives measurements from the said environmental sensor (220) and the said structure sensor (210) and enables a virtual structure model (500) related to the structure (200) according to these measurements to simulate the responses of the structure (200) to environmental factors and structure parameters in the physical world, that is, to do similar. According to the information obtained from the instantaneous sensors, the instantaneous simulation of the virtual structure model (500) is known as the digital twin in the art.

The said structures (200) may be bridges, structures, roads, etc. The components of the structure (200) refer to countless exemplary components such as the wall, column, beam, structure body, roof, etc. that make up the structure (200).

In a possible embodiment of the invention, one of the structure parameters may be a strain. The instantaneous strain of the respective components of the structure (200) can be measured by means of strain sensors placed in components such as columns, walls, and beams. Structure parameters may be such as humidity, temperature, force applied to a component, acceleration, the load to which a component is exposed, etc. and structure sensors (210) may also be suitable sensors to measure these parameters. In a possible embodiment of the invention, the structure sensor (210) may be an image sensor and the structure parameter may be the form of the structure (200) obtained by the image. The structure sensor (210) may be a thermal camera, image camera, or LIDAR. Thus, the instantaneous view of the structure (200) can be obtained.

The information processing device (100) may comprise a processor unit (110). The processor unit (110) may be associated with a memory unit (120) to read and write data. The processor unit (110) may be a microprocessor, CPU, GPU, etc. The memory unit (120) may also include memory or memory combinations that enable data to be stored permanently and temporarily.

The information processing device (100) may be a general-purpose computer, a server, etc.

The memory unit (120) may include functional modules consisting of command lines that enable the operation of the invention when executed by the processor unit (110). The memory unit (120) may also include the said virtual structure model (500). The virtual structure model (500) can be a model created with model creation software known in the art. Examples of such software are AutoCAD software or the CAD platform. The virtual structure model (500) can also be a model created with data from LIDAR. The processor unit (110) can simulate the virtual structure model (500) under these conditions according to the structure parameters and environmental parameters. The processor unit (110) may simulate the responses of the components of the structure (200), for example, by using measurements such as the obtained strain, force, temperature, etc. in response to the shock measured during an earthquake. For example, it can create a simulation showing that a column oscillates against vibration, and it can do these simulations together with all the components and their effects on each other.

The processor unit (110) can update the virtual structure model (500) according to the characteristics of the structure (200) it receives as input and simulate the virtual structure model (500) accordingly. Simulation software is well known in the art, this software models mathematical formulas and physical events and visualizes the responses of the model according to the parameter inputs received.

In a possible embodiment of the invention, the environmental sensor (220) may be a seismograph-like sensor and it may measure earthquakes. In a possible embodiment of the invention, the environmental sensors (220) may be temperature sensors and humidity sensors.

The invention is characterized in that it uses a machine learning model. The processor unit (110) enables the training of the said machine learning model and can determine the response of the components of the structure (200) in the face of a natural disaster that may occur in the future according to this model and enable this response to be simulated. The processor unit (110) trains the machine learning model with the measurements taken from the environmental sensors (220) and the measurements taken from the structure sensors (210) at the moment these measurements are taken from the environmental sensors (220). The processor unit (110) then inputs the test environment parameters expected to be obtained during natural disasters to the trained machine learning model and takes the responses predicted to give these test environment parameters by the components of the structure (200) as output. These responses can be obtained in terms of structure parameters. For example, when an earthquake of 7 magnitudes occurs according to the Richter scale, in return for the vibration parameter given as input, the strain to be measured according to time from a column to be demolished can be obtained in terms of structure parameter values. These values are applied to the virtual structure model (500) and the response of the virtual structure model (500) when the test parameters are realized is simulated.

The said test parameters may be predetermined values or they may be entered by the user via an input unit (not shown in the figure).

In a possible embodiment of the invention, some features of the components of the structure (200) may be entered into the information processing device (100) manually. For this, the information processing device (100) comprises an input unit. These features may be the type, ratio, etc. of material contained in the components.

In a possible embodiment of the invention, the processor simulates the virtual structure model (500) according to the data it receives from the environmental sensors (220) and the structure sensors (210) and allows the instant state of the structure (200) to be observed.

Thanks to the invention, it is ensured that the parameters of the structures (200) are observed while they are under construction or in use after the completion of the construction and the response to a possible natural disaster is determined with the machine learning model; thus, the structures (200) that can be damaged as a result of the natural disaster are determined in advance.

The digital twin system may additionally include a communication unit (250) to enable the measurements received by the sensors to be transferred to the information processing device (100). The said communication unit (250) may be a gateway. The information processing device (100) may be associated with a plurality of client devices (150), and the input units and user interfaces (130) may be associated with the client devices (150). Client devices (150) may be general-purpose computers, mobile computers, smartphones, etc.

The client devices (150) may communicate with the information processing device (100) through a communication network (400).

An exemplary mode of operation of the inventive digital twin system detailed above is as follows:

The structure sensors (210) are associated with the components of the structure (200) during or after the construction. Measurements are taken from the environmental sensors (220) and the structure sensors (210). The processor unit (110) enables the virtual structure model (500) to be simulated according to the instantaneous measurements. The processor unit (110) also enables the machine-learning model to be trained. The training of the machine learning model can continue over instant data e.g. 1 week. The processor unit (110) then inputs the test environment parameters to the trained machine learning model. The test environment parameters can be predetermined values or values entered by the user via an input unit. The processor unit (110) obtains the response structure parameters from the machine learning model in response to the test parameters. The response determines the damage according to the structure parameters. The processor unit (110) also enables the simulation of the virtual structure model (500) by using the response structure parameters and the test environment parameters. Thus, the responses of the existing structure (200) in the event of a natural disaster can be predicted, components can be predicted to be damaged, and precautions can be taken accordingly.

The digital twin system of the invention can be applied to a single structure (200) or a settlement consisting of multiple structures (200), such as a neighborhood. The scope of protection of the invention is specified in the attached claims and cannot be limited to those explained for sampling purposes in this detailed description. It is evident that a person skilled in the art may exhibit similar embodiments in light of the above-mentioned facts without drifting apart from the main theme of the invention.

REFERENCE NUMBERS GIVEN IN THE FIGURE

10 Digital twin system

100 Information processing device 110 Processor unit

120 Memory unit

130 User interface

150 Client device

200 Structure 210 Structure sensor

220 Environmental sensor

250 Communication unit

400 Communication network

500 Virtual structure model

Claims

CLAIMS A computer-based digital twin system for predicting the response of a structure (200) to natural disasters, characterized in that it comprises a plurality of structure sensors (210) to detect at least one structure parameter of the multiple components; at least one environmental sensor (220) to measure at least one environmental parameter; an information processing device (100) configured to receive the signals generated regarding the measurements made by the said environmental sensor (220) and the said structure sensors (210); the said information processing device (100) comprises an information processing device (100) configured to access a virtual structure model (500) of the structure (200) and simulate the responses of the components of the said structure (200) in the real world according to the said structure parameters and environmental parameters in the said virtual structure model (500); the said information processing device (100) is configured to access a trained machine learning model using the environmental parameter measurements and the structure parameters measurements at the time these parameters are measured; to apply the test environment parameter values, which are in magnitude measured when the natural disaster occurs and are predetermined, to the machine learning model as input and determine the damage status of the structure (200) according to the response structure parameters of the machine learning model to this and simulate the response structure parameters in the virtual structure model (500). A digital twin system according to claim 1 , characterized in that the information processing device (100) is configured to simulate the virtual structure model (500) according to the instantaneous environmental parameter measurements and structure parameter measurements it receives to show the instantaneous state of the building (200). A digital twin system according to claims 1 or 2, characterized in that it is configured to display the simulations made in the said virtual structure model (500) on a user interface (130). A digital twin system according to claim 1 , characterized in that the said sensors are at least one of strain sensors, force sensors, acceleration sensors, temperature sensors, and humidity sensors.

5. A digital twin system according to claim 1 , characterized in that the said environmental sensors (220) are at least one of the seismograph, humidity sensor, and temperature sensor. 6. A digital twin system according to claim 1 , characterized in that the processor unit

(110) is configured to train the machine learning model with the structure parameters obtained in response to the test environment parameters.