A construction method of a floating platform digital twin system based on inverse finite elements

Abstract

The application discloses a kind of based on inverse finite element's floating platform digital twin system construction method, it is related to marine engineering structure digital twin field, including the following steps: constructing distributed strain sensing network on the main structure of floating platform;High-fidelity digital twin geometric model of floating platform is constructed, the mapping relationship of strain sensing unit and high-fidelity digital twin geometric model node is established;Strain data is preprocessed, and three-dimensional deformation field of floating platform is reconstructed based on inverse finite element method;Based on the three-dimensional deformation field of reconstruction, the stress field of platform is calculated, and multidimensional evaluation is carried out;According to multidimensional evaluation result, automatically generate comprehensive early warning level and output operation and maintenance suggestion;System full life cycle data is classified and stored and multilayer backup;Fusion sensing data and mechanics model, digital twin technology with high-precision inversion and online evaluation capability, realize the dynamic monitoring of floating platform, provide technical support for the safety management of platform full life cycle.

Description

Technical Field

[0001] This invention relates to the field of digital twins of marine engineering structures, and in particular to a method for constructing a digital twin system for a floating platform based on inverse finite element method. Background Technology

[0002] As the core equipment for deep-sea oil and gas development, floating platforms have been in operation for a long time in complex marine environments with multiple load coupling effects. During the entire life cycle of operation, the platform is subjected to multiple sources of coupling effects such as periodic wave loads, wind and current disturbances and operational loads, which can easily lead to a gradual evolution process of fatigue damage accumulation and structural stiffness degradation, resulting in a gradual weakening of the overall mechanical performance of the platform.

[0003] In existing technologies, although finite element analysis or empirical correction methods are widely used for structural strength assessment, they are limited by load assumptions and real-time calculations, making it difficult to meet the online monitoring needs in complex marine environments.

[0004] In recent years, digital twin technology has been gradually introduced into the field of marine engineering to achieve dynamic interaction and information fusion between physical entities and virtual models, thereby supporting structural health monitoring and life prediction. However, there are still significant limitations: on the one hand, due to limitations in sensor network layout and marine environmental interference, monitoring data suffers from anomalies, noise, and missing data, resulting in insufficient mapping accuracy between physical entities and twin models. On the other hand, pure data-driven models often rely on statistical correlation or empirical learning methods, while traditional physical models are limited by computational efficiency. Both lack the ability to efficiently and accurately characterize the real mechanical mechanisms, making it difficult to achieve interpretable inversion and dynamic extrapolation of structural deformation, stress, and fatigue damage.

[0005] Therefore, a method for constructing a digital twin system for a floating platform based on inverse finite element method is provided to solve the above problems. Summary of the Invention

[0006] The purpose of this invention is to provide a method for constructing a digital twin system for a floating platform based on inverse finite element method, which solves the problems of low accuracy, unclear mechanism, single evaluation dimension and insufficient real-time performance of traditional methods for reconstructing digital twin models of floating platforms.

[0007] To achieve the above objectives, this invention provides a method for constructing a digital twin system for a floating platform based on the inverse finite element method, comprising the following steps: S1: Construct a distributed strain sensing network on the main structure of the floating platform; S2: Construct a high-fidelity digital twin geometric model of the floating platform and establish a mapping relationship between strain sensing units and nodes of the high-fidelity digital twin geometric model; S3: Collect strain data through a distributed strain sensing network, preprocess the strain data, and reconstruct the three-dimensional deformation field of the floating platform based on the inverse finite element method; S4: Calculate the stress field of the floating platform based on the reconstructed three-dimensional deformation field, and conduct multi-dimensional evaluation of structural deformation, strength and fatigue damage; S5: Based on multi-dimensional assessment results, automatically generate a comprehensive early warning level and output operation and maintenance suggestions; S6: Classify and store system data throughout its entire lifecycle and perform multi-level backups.

[0008] Preferably, step S1 specifically includes the following steps: S11: Define the main structure of the floating platform, which includes the deck, columns, pontoons, and bulkheads; S12: Discretize the deck, pillars, pontoons and bulkheads into uniform grids respectively; S13: Strain sensing units are arranged on the front and back sides of each uniform grid. The strain sensing units are located at the centroid of the uniform grid and adopt fiber optic grating sensing arrays. S14: Calculate the measured strain of the neutral layer using a strain sensing unit. The measured strain includes plane strain. and bending strain plane strain and bending strain Set them to: ; ; in, Indicates the grid cell number, Indicates plate thickness. Indicates measurement at the top of the cell To adapt to change, Indicates measurement at the bottom of the unit To adapt to change, Indicates measurement at the top of the cell To adapt to change, Indicates measurement at the bottom of the unit To adapt to change, This represents the in-plane shear strain measured at the top of the element. This represents the in-plane shear strain measured at the bottom of the element.

[0009] Preferably, step S2 specifically includes the following steps: S21: Based on the original design data of the floating platform, construct a high-fidelity parametric model with geometric and material details, and import it into the digital twin interactive platform; S22: Define the deck, pillars, pontoons and bulkheads as structural subdomains, set corresponding material constitutive properties for each structural subdomain, and assign unique identifiers; S23: Establish a one-to-one mapping relationship between the spatial coordinates of the strain sensing unit and specific nodes on the corresponding structural subdomain, and extend it into a mapping link from physical space to digital space.

[0010] Preferably, step S3 specifically includes the following steps: S31: Strain data is collected through strain sensing units, and evaluation metrics are assigned to the strain data using the Isolation Forest algorithm. And remove evaluation indicators High data points, evaluation metrics Specifically set as follows: ; in, Representing data points height, Representing data points Average height among all isolated trees express The average height of the data points; S32: Predict and fill in the missing data of the strain sensing unit using the Gaussian process regression algorithm. The specific algorithm model of the Gaussian process regression algorithm is set as follows: ; ~N (0 ); in, This represents the algorithm's predicted value. Represents the input feature matrix The transpose of the corresponding feature mapping vector, Represents a mapping function. Represents the input feature matrix. This represents the weight vector of a linear model. Indicates Gaussian noise. N Indicates a normal distribution. Indicates the noise variance; S33: Based on the preprocessed strain data, perform inverse finite element method deformation inversion on the structure at the boundary, and solve the core reconstruction equation. The core reconstruction equation is specifically set as follows: ; ; ; ; ; ; in, This represents the stiffness matrix after introducing boundary condition constraints. Represents the deformation field of the reconstruction. This represents the measured strain vector after introducing boundary condition constraints. Indicates the unit number. Indicates the total number of units. Represents the coordinate transformation matrix. This represents the transpose of the coordinate transformation matrix. Represents the element stiffness matrix. Represents the measured strain matrix of the element. Represents the element deformation vector. Represents the area of ​​the neutral layer in the unit cell. , and All represent weighting constants; 1 is taken when there are measured values, and 0 is taken when there are no measured values. , Represents the plane strain matrix. This represents the transpose of the plane strain matrix. Represents the bending strain matrix. This represents the transpose of the bending strain matrix. Represents the transverse shear strain matrix. This represents the transpose of the transverse shear strain matrix. This indicates the number of monitoring points within a single grid cell. Indicates transverse shear strain; S34: The deformation field of the floating platform is reconstructed using a boundary recursion algorithm. The reconstruction starts from the boundary constraint structure, extracts the displacement response of the connection interface, and uses it as the boundary condition of the connection structure subdomain after coordinate transformation for recursive inversion until the floating platform is covered.

[0011] Preferably, step S4 specifically includes the following steps: S41: Calculate the stress field of the floating platform using the shape function differentiation method, and reconstruct the deformation field. By performing differentiation, the strain field is obtained, and the stress tensor of the entire field is calculated according to the generalized Hooke's law; S42: Real-time calculation of maximum displacement of decks, pillars, pontoons, and bulkheads. With design allowable deformation The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model; S43: Real-time calculation of maximum equivalent stress on decks, pillars, pontoons, and bulkheads. With the allowable stress of the material The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model; S44: Based on the SN curve of the material, load cycle statistics are performed by combining the rainflow counting method, and the damage degree is iteratively updated by Miner's linear cumulative damage theory. S45: Calculate cumulative fatigue damage With allowable fatigue damage The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model, accumulating fatigue damage. Specifically set as follows: ; ; in, Indicates the number of different stress amplitudes. Indicates the stress level of the floating platform structure Fatigue damage under the action of , Indicates constant amplitude stress The number of loops, This indicates that the floating platform structure is under constant amplitude stress. The number of cycles required to induce fatigue failure under certain conditions. This represents the fatigue strength safety factor.

[0012] Preferably, step S5 specifically includes the following steps: S51: Based on multi-dimensional assessment results, establish a real-time early warning indicator system, conduct continuous risk assessment and early warning at the second level, trigger corresponding early warning levels, and simultaneously output operation and maintenance suggestions. The specific operation and maintenance principles are set as follows: S511: At that time, the floating platform structure responded within the safe range, triggering a Level 1 warning, and routine inspections were carried out as originally planned; S512: At that time, the floating platform structure generated a large response, triggering a level-two early warning, increasing the monitoring frequency in the relevant area, and manually verifying the data; S513: When the floating platform structure approaches the design allowable value, a level three warning is triggered. Adjust the working load or platform orientation and prepare a maintenance plan. S514: When the floating platform structure has reached or exceeded the design limit, a level four warning is triggered, and load reduction or work stoppage operations are performed, and emergency inspection is initiated. S52: Based on the multi-dimensional assessment results, establish a periodic early warning indicator system to conduct monthly fatigue intensity risk assessments and early warnings, and simultaneously output operation and maintenance suggestions. The specific operation and maintenance principles are set as follows: S521: If fatigue damage is within the expected range, a Level 1 warning is triggered, and the existing inspection cycle is maintained. S522: When fatigue damage accumulates, a level-two warning is triggered, and the relevant floating platform structure is included in the key monitoring list, and the assessment cycle is shortened from monthly to weekly. S523: At that time, fatigue damage accumulates rapidly, triggering a level three warning, and a special fatigue inspection and maintenance plan is formulated and arranged. S524: When fatigue damage approaches a critical state, a level four warning is triggered, a fatigue safety assessment is conducted, and a reinforcement or component replacement plan is planned.

[0013] Preferably, step S6 specifically includes the following steps: S61: Divide the data into raw collected data and processed data, store them separately, and encapsulate them according to time series; S62: Based on access frequency and time range, the raw collected data and processed data are divided into hot data, warm data and cold data, and different materials are used for storage according to the classification, implementing a three-level layered storage of hot-warm-cold data; S63: Performs local multi-layer backup and off-site disaster recovery backup of data. Local backup uses real-time synchronous mirroring or scheduled backup, while an off-site backup center is established. Incremental data is transmitted daily through the network to build a local-off-site-cloud multimodal collaborative backup mechanism.

[0014] Therefore, the present invention employs the above-mentioned method for constructing a floating platform digital twin system, based on the inverse finite element method, and has the following beneficial effects: (1) This scheme achieves deep integration of sensing data and physical mechanism by introducing the inverse finite element method, overcoming the defect that traditional data-driven models are difficult to reflect the real mechanical state of the structure; (2) This solution constructs a closed-loop online process from perception, reconstruction, evaluation to storage, breaking through the bottleneck of traditional analysis methods that are computationally time-consuming and cannot be used for real-time monitoring; (3) This scheme establishes a multi-dimensional assessment system covering the deformation, strength and fatigue of floating platforms, and realizes a more comprehensive quantitative identification and early warning of structural risks; (4) This solution designs an intelligent storage and multimodal backup architecture based on data tiering, which ensures system robustness while achieving cost-effective full lifecycle data management.

[0015] The method of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] Figure 1This is a flowchart illustrating a method for constructing a digital twin system for a floating platform based on inverse finite element method according to the present invention. Figure 2 This is a schematic diagram of the in-plane and bending strain component measurement of the floating platform structure of the present invention; Figure 3 This is a schematic diagram of the grid discretization and sensor measurement point arrangement of the single-board structure in an embodiment of the present invention; Figure 4 This is a flowchart of the inverse finite element reconstruction of the three-dimensional deformation field across the entire platform according to the present invention; Figure 5 This is a schematic diagram illustrating the coordinate system definition and displacement recursive reconstruction of the three-dimensional wall panel structure in an embodiment of the present invention; Figure 6 This is a flowchart illustrating the multi-dimensional coupled evaluation process of structural deformation, strength, and fatigue damage in this invention. Detailed Implementation

[0017] The method of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0018] Unless otherwise defined, the methodological or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0019] The terms "comprising" or "including" as used in this invention mean that the element preceding the term encompasses the element listed after the term, and do not exclude the possibility of encompassing other elements. Terms such as "inner," "outer," "upper," and "lower" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. When the absolute position of the described object changes, the relative positional relationship may also change accordingly. In this invention, unless otherwise explicitly specified and limited, the term "attached" and similar terms should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral part; it can refer to a direct connection or an indirect connection through an intermediate medium; it can refer to the internal communication of two elements or the interaction relationship between two elements. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0020] Example like Figures 1-6 As shown, this invention provides a method for constructing a digital twin system for a floating platform based on the inverse finite element method, comprising the following steps: S1: Construct a distributed strain sensing network on the main structure of the floating platform to realize real-time synchronous acquisition of structural strain response. Step S1 specifically includes the following steps: S11: Define the main structure of the floating platform, which includes the deck, columns, pontoons, and bulkheads; S12: Discretize the deck, pillars, pontoons and bulkheads into uniform grids respectively; S13: Strain sensing units are deployed on the front and back sides of each uniform grid. The strain sensing units are located at the centroid of the uniform grid. The strain sensing units adopt fiber grating sensing arrays. All data are synchronized based on a unified time reference to ensure the consistency of the spatiotemporal data field. S14: Calculate the measured strain of the neutral layer using a strain sensing unit. The measured strain includes plane strain. and bending strain plane strain and bending strain Set them to: ; ; in, Indicates the grid cell number, Indicates plate thickness. Indicates measurement at the top of the cell To adapt to change, Indicates measurement at the bottom of the unit To adapt to change, Indicates measurement at the top of the cell To adapt to change, Indicates measurement at the bottom of the unit To adapt to change, This represents the in-plane shear strain measured at the top of the element. This represents the in-plane shear strain measured at the bottom of the element.

[0021] The grid discretization and sensor arrangement in this embodiment are taken from the side plate of a platform float. The plate is 37.8m long, 6.9m wide, and 18mm thick. The plate is discretized using iQS4 cells. The grid is 0.6m long and wide, with a total of 693 grid cells. The markers in the grid represent strain monitoring points, and the blank grids represent areas where sensors cannot be deployed due to space limitations.

[0022] S2: Construct a high-fidelity digital twin geometric model of the floating platform and establish a mapping relationship between strain sensing units and nodes of the high-fidelity digital twin geometric model; Step S2 specifically includes the following steps: S21: Based on the original design data of the floating platform, construct a high-fidelity parametric model with geometric and material details, and import it into the digital twin interactive platform; S22: Define the deck, pillars, pontoons and bulkheads as structural subdomains, set corresponding material constitutive properties for each structural subdomain, and assign unique identifiers; S23: Establish a one-to-one mapping relationship between the spatial coordinates of the strain sensing unit and specific nodes on the corresponding structural subdomain, and extend it into a mapping link from physical space to digital space.

[0023] S3: Collect strain data through a distributed strain sensing network, preprocess the strain data, and reconstruct the three-dimensional deformation field of the floating platform based on the inverse finite element method; Step S3 specifically includes the following steps: S31: Strain data is collected through strain sensing units, and evaluation metrics are assigned to the strain data using the Isolation Forest algorithm. And remove evaluation indicators High data points, evaluation metrics Specifically set as follows: ; in, Representing data points height, Representing data points Average height among all isolated trees express The average height of each data point; S32: Predict and fill in the missing data of the strain sensing unit using the Gaussian process regression algorithm. The specific algorithm model of the Gaussian process regression algorithm is set as follows: ; ~N (0 ); in, This represents the algorithm's predicted value. Represents the input feature matrix The transpose of the corresponding feature mapping vector, Represents a mapping function. Represents the input feature matrix. This represents the weight vector of a linear model. Indicates Gaussian noise. N Indicates a normal distribution. The noise variance is represented by a mapping function trained using existing monitoring data, and then the missing point features are input to obtain the predicted value of the missing point. S33: Based on the preprocessed strain data, perform inverse finite element method deformation inversion on the structure at the boundary, and solve the core reconstruction equation. The core reconstruction equation is specifically set as follows: ; ; ; ; ; ; in, This represents the stiffness matrix after introducing boundary condition constraints. Represents the deformation field of the reconstruction. This represents the measured strain vector after introducing boundary condition constraints. Indicates the unit number. Indicates the total number of units. Represents the coordinate transformation matrix. This represents the transpose of the coordinate transformation matrix. Represents the element stiffness matrix. Represents the measured strain matrix of the element. Represents the element deformation vector. Represents the area of ​​the neutral layer in the unit cell. , and All represent weighting constants; 1 is taken when there are measured values, and 0 is taken when there are no measured values. , Represents the plane strain matrix. This represents the transpose of the plane strain matrix. Represents the bending strain matrix. This represents the transpose of the bending strain matrix. Represents the transverse shear strain matrix. This represents the transpose of the transverse shear strain matrix. This indicates the number of monitoring points within a single grid cell. This represents transverse shear strain. In practical applications, transverse shear strain is difficult to measure using strain gauges. If its influence is small, a weighted method is used to control its proportion in the calculation, or it can be ignored directly. In the inverse finite element displacement reconstruction process of the same plate, only one calculation is needed at the initial stage. ,and It can be updated in real time based on monitoring data. As the only unknown in the system, it can be directly obtained through the simple product of a matrix and a vector. This solution process does not require iteration. With the real-time update of sensor data, the system can quickly solve for the updated deformation field, thereby realizing real-time monitoring of structural deformation.

[0024] Obtain the deformation vector Then, the deformation at any point within the structure is obtained through shape function interpolation. The displacement response at the interface connecting this structural subdomain and other structural subdomains is then obtained through interpolation. This displacement response is used as the boundary condition for other structural subdomains. In the local coordinate system of the structure, The axis runs along the thickness of the plate.

[0025] S34: The deformation field of the floating platform is reconstructed using a boundary recursion algorithm. The reconstruction starts from the boundary constraint structure, extracts the displacement response of the connection interface, and uses it as the boundary condition of the connection structure subdomain after coordinate transformation for recursive inversion until the floating platform is covered.

[0026] This embodiment is a three-dimensional wall panel structure composed of three plates, with a coordinate system... For the global coordinate system, coordinate system ( ), ( )and( The figures show the local coordinate systems of plates 1, 2, and 3, respectively. Plates 1 and 3 are constrained by fixed supports from the lower boundary. The displacements of the three plates are transformed using the following formula: ; ; in, Indicates the position of plate 1 at displacement boundary 1 Displacement, Indicator plate 2 at displacement boundary 1 Displacement, Indicator plate 2 at displacement boundary 2 Displacement, Indicates the position of plate 3 at displacement boundary 2 Displacement, Indicates the position of plate 1 at displacement boundary 1 Displacement, Indicator plate 2 at displacement boundary 1 Displacement, Indicator plate 2 at displacement boundary 2 Displacement, Indicates the position of plate 3 at displacement boundary 2 Displacement, Indicates the position of plate 1 at displacement boundary 1 Displacement, Indicator plate 2 at displacement boundary 1 Displacement, Indicator plate 2 at displacement boundary 2 Displacement, Indicates the position of plate 3 at displacement boundary 2 Displacement; The displacement fields of plates 1 and 3 located at the boundary are reconstructed first. Then, the reconstructed displacement data at the junction of each plate with plate 2 are extracted and transformed to the local coordinate system of plate 2 as the boundary conditions for the reconstruction of plate 2. Inverse finite element reconstruction of plate 2 is then performed.

[0027] For the iQS4 inverse finite element plate element used in this embodiment, the relationship between strain and displacement is obtained by using the shape function differentiation method: ; ; ; ; ; in, , , , , Both represent strain components. , , Both represent translational displacements. , , Each represents the total displacement component at any point. , Both represent angular displacement.

[0028] S4: Calculate the stress field of the floating platform based on the reconstructed three-dimensional deformation field, and conduct multi-dimensional evaluation of structural deformation, strength and fatigue damage; Step S4 specifically includes the following steps: S41: Calculate the stress field of the floating platform using the shape function differentiation method, and reconstruct the deformation field. By performing differentiation, the strain field is obtained, and the stress tensor of the entire field is calculated according to the generalized Hooke's law; S42: Real-time calculation of maximum displacement of decks, pillars, pontoons, and bulkheads. With design allowable deformation The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model; S43: Real-time calculation of maximum equivalent stress on decks, pillars, pontoons, and bulkheads. With the allowable stress of the material The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model; S44: Based on the SN curve of the material, load cycle statistics are performed by combining the rainflow counting method, and the damage degree is iteratively updated by Miner's linear cumulative damage theory. S45: Calculate cumulative fatigue damage With allowable fatigue damage The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model, accumulating fatigue damage. Specifically set as follows: ; ; in, Indicates the number of different stress amplitudes. Indicates the stress level of the floating platform structure Fatigue damage under the action of , Indicates constant amplitude stress The number of loops, This indicates that the floating platform structure is under constant amplitude stress. The number of cycles required to induce fatigue failure under certain conditions. The fatigue strength safety factor is used to represent different safety factors depending on the structure type, the environment in which the structure is located, whether it is accessible for maintenance, and the consequences of failure. The specific value should refer to the relevant specifications of the classification society.

[0029] S5: Based on multi-dimensional assessment results, automatically generate a comprehensive early warning level and output operation and maintenance suggestions; Step S5 specifically includes the following steps: S51: Based on multi-dimensional assessment results, establish a real-time early warning indicator system, conduct continuous risk assessment and early warning at the second level, trigger corresponding early warning levels, and simultaneously output operation and maintenance suggestions. The specific operation and maintenance principles are set as follows: S511: At that time, the floating platform structure responded within the safe range, triggering a Level 1 warning, and routine inspections were carried out as originally planned; S512: At that time, the floating platform structure generated a large response, triggering a level-two early warning, increasing the monitoring frequency in the relevant area, and manually verifying the data; S513: When the floating platform structure approaches the design allowable value, a level three warning is triggered. Adjust the working load or platform orientation and prepare a maintenance plan. S514: When the floating platform structure has reached or exceeded the design limit, a level four warning is triggered, and load reduction or work stoppage operations are performed, and emergency inspection is initiated. The system's final warning level is determined autonomously based on the highest principle and the combined upgrade principle. The final warning level is the highest single-dimensional level among the two dimensions. If both dimensions reach level three at the same time, the final warning level will be upgraded to level four.

[0030] S52: Based on the multi-dimensional assessment results, establish a periodic early warning indicator system to conduct monthly fatigue intensity risk assessments and early warnings, and simultaneously output operation and maintenance suggestions. The specific operation and maintenance principles are set as follows: S521: If fatigue damage is within the expected range, a Level 1 warning is triggered, and the existing inspection cycle is maintained. S522: When fatigue damage accumulates, a level-two warning is triggered, and the relevant floating platform structure is included in the key monitoring list, and the assessment cycle is shortened from monthly to weekly. S523: At that time, fatigue damage accumulates rapidly, triggering a level three warning, and a special fatigue inspection and maintenance plan is formulated and arranged. S524: When fatigue damage approaches a critical state, a level four warning is triggered, a fatigue safety assessment is conducted, and a reinforcement or component replacement plan is planned. Real-time dynamic early warning and periodic trend early warning operate in parallel and independently within the system.

[0031] S6: Classify and store system data throughout its entire lifecycle and perform multi-level backups.

[0032] Step S6 specifically includes the following steps: S61: Divide the data into raw collected data and processed data, store them separately, and encapsulate them according to time series; S62: Based on access frequency and time range, the raw collected data and processed data are divided into hot data, warm data and cold data, and different materials are used for storage according to the classification, implementing a three-level layered storage of hot-warm-cold data; For data from the past month and frequently accessed data (≥1 time / hour), SSD solid-state drives are used for storage to ensure high-frequency access requirements. For data within 1 month to 1 year, and for warm data accessed once per day to once per hour, SAS hard disk storage is used to balance access requirements and storage costs. For historical data exceeding one year and cold data accessed ≤ once per day, use SATA hard drives or tape libraries for storage to reduce storage costs.

[0033] S63: Performs local multi-layer backup and off-site disaster recovery backup for data. Local backup uses real-time synchronous mirroring or scheduled backup. Multi-layer backup of core data is implemented within the server or local area network to ensure business continuity. An off-site backup center is established to transmit incremental data daily through the network to prevent regional disasters. In addition, important archived data is regularly encrypted and uploaded to the cloud for cold backup and long-term storage, building a local-off-site-cloud multimodal collaborative backup mechanism.

[0034] Therefore, this invention adopts the above-mentioned method for constructing a floating platform digital twin system based on inverse finite element method, which integrates sensor data and mechanical model, and digital twin technology with high-precision inversion and online evaluation capabilities, to realize dynamic monitoring of floating platforms and provide technical support for the safety management of the platform throughout its entire life cycle.

[0035] Finally, it should be noted that the above embodiments are only used to illustrate the method of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the method of the present invention, and these modifications or equivalent substitutions should not cause the modified method to deviate from the spirit and scope of the method of the present invention.

Claims

1. A method for constructing a digital twin system for a floating platform based on inverse finite element method, characterized in that, Includes the following steps: S1: Construct a distributed strain sensing network on the main structure of the floating platform; S2: Construct a high-fidelity digital twin geometric model of the floating platform and establish a mapping relationship between strain sensing units and nodes of the high-fidelity digital twin geometric model; S3: Collect strain data through a distributed strain sensing network, preprocess the strain data, and reconstruct the three-dimensional deformation field of the floating platform based on the inverse finite element method; S4: Calculate the stress field of the floating platform based on the reconstructed three-dimensional deformation field, and conduct multi-dimensional evaluation of structural deformation, strength and fatigue damage; S5: Based on multi-dimensional assessment results, automatically generate a comprehensive early warning level and output operation and maintenance suggestions; S6: Classify, store, and back up system data throughout its entire lifecycle; Step S1 specifically includes the following steps: S11: Define the main structure of the floating platform, which includes the deck, columns, pontoons, and bulkheads; S12: Discretize the deck, pillars, pontoons and bulkheads into uniform grids respectively; S13: Strain sensing units are arranged on each uniform grid. The strain sensing units are located at the centroid of the uniform grid. The strain sensing units adopt fiber optic grating sensing arrays. S14: Calculate the measured strain of the neutral layer using a strain sensing unit. The measured strain includes plane strain. and bending strain plane strain and bending strain Set them to: ； ； in, Indicates the grid cell number, Indicates plate thickness. Indicates measurement at the top of the cell To adapt to change, Indicates measurement at the bottom of the unit To adapt to change, Indicates measurement at the top of the cell To adapt to change, Indicates measurement at the bottom of the unit To adapt to change, This represents the in-plane shear strain measured at the top of the element. This represents the in-plane shear strain measured at the bottom of the element; Step S2 specifically includes the following steps: S21: Based on the original design data of the floating platform, construct a high-fidelity parametric model with geometric and material details, and import it into the digital twin interactive platform; S22: Define the deck, pillars, pontoons and bulkheads as structural subdomains, set corresponding material constitutive properties for each structural subdomain, and assign unique identifiers; S23: Establish a one-to-one mapping relationship between the spatial coordinates of the strain sensing unit and specific nodes on the corresponding structural subdomain, and extend it to a mapping link from physical space to digital space. Step S3 specifically includes the following steps: S31: Strain data is collected through strain sensing units, and evaluation metrics are assigned to the strain data using the Isolation Forest algorithm. And remove evaluation indicators High data points, evaluation metrics Specifically set as follows: ； in, Representing data points height, Representing data points Average height among all isolated trees express The average height of the data points; S32: Predict and fill in the missing data of the strain sensing unit using the Gaussian process regression algorithm. The specific algorithm model of the Gaussian process regression algorithm is set as follows: ； ~N (0 )； in, This represents the algorithm's predicted value. Represents the input feature matrix The transpose of the corresponding feature mapping vector, Represents a mapping function. Represents the input feature matrix. This represents the weight vector of a linear model. Indicates Gaussian noise. N Indicates a normal distribution. Indicates the noise variance; S33: Based on the preprocessed strain data, perform inverse finite element method deformation inversion on the structure at the boundary, and solve the core reconstruction equation. The core reconstruction equation is specifically set as follows: ； ； ； ； ； ； in, This represents the stiffness matrix after introducing boundary condition constraints. Represents the deformation field of the reconstruction. This represents the measured strain vector after introducing boundary condition constraints. Indicates the unit number. Indicates the total number of units. Represents the coordinate transformation matrix. This represents the transpose of the coordinate transformation matrix. Represents the element stiffness matrix. Represents the measured strain matrix of the element. Represents the element deformation vector. Represents the area of ​​the neutral layer in the unit cell. , and All represent weighting constants; 1 is taken when there are measured values, and 0 is taken when there are no measured values. , Represents the plane strain matrix. This represents the transpose of the plane strain matrix. Represents the bending strain matrix. This represents the transpose of the bending strain matrix. Represents the transverse shear strain matrix. This represents the transpose of the transverse shear strain matrix. This indicates the number of monitoring points within a single grid cell. Indicates transverse shear strain; S34: The deformation field of the floating platform is reconstructed using a boundary recursion algorithm. The reconstruction starts from the boundary constraint structure, extracts the displacement response of the connection interface, and uses it as the boundary condition of the connection structure subdomain after coordinate transformation for recursive inversion until the floating platform is covered.

2. The method for constructing a digital twin system for a floating platform based on inverse finite element method according to claim 1, characterized in that, Step S4 specifically includes the following steps: S41: Calculate the stress field of the floating platform using the shape function differentiation method, and reconstruct the deformation field. By performing differentiation, the strain field is obtained, and the stress tensor of the entire field is calculated according to the generalized Hooke's law; S42: Real-time calculation of maximum displacement of decks, pillars, pontoons, and bulkheads. With design allowable deformation The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model; S43: Real-time calculation of maximum equivalent stress on decks, pillars, pontoons, and bulkheads. With the allowable stress of the material The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model; S44: Based on the SN curve of the material, load cycle statistics are performed by combining the rainflow counting method, and the damage degree is iteratively updated by Miner's linear cumulative damage theory. S45: Calculate cumulative fatigue damage With allowable fatigue damage The ratio, and dynamically drive the state update and visualization of the high-fidelity digital twin geometric model, accumulating fatigue damage. Specifically set as follows: ； ； in, Indicates the number of different stress amplitudes. Indicates the stress level of the floating platform structure Fatigue damage under the action of , Indicates constant amplitude stress The number of loops, This indicates that the floating platform structure is under constant amplitude stress. The number of cycles required to induce fatigue failure under certain conditions. This represents the fatigue strength safety factor.

3. The method for constructing a digital twin system for a floating platform based on inverse finite element method according to claim 2, characterized in that, Step S5 specifically includes the following steps: S51: Based on multi-dimensional assessment results, establish a real-time early warning indicator system, conduct continuous risk assessment and early warning at the second level, trigger corresponding early warning levels, and simultaneously output operation and maintenance suggestions. The specific operation and maintenance principles are set as follows: S511: At that time, the floating platform structure responded within the safe range, triggering a Level 1 warning, and routine inspections were carried out as originally planned; S512: At that time, the floating platform structure generated a large response, triggering a level-two early warning, increasing the monitoring frequency in the relevant area, and manually verifying the data; S513: When the floating platform structure approaches the design allowable value, a level three warning is triggered. Adjust the working load or platform orientation and prepare a maintenance plan. S514: When the floating platform structure has reached or exceeded the design limit, a level four warning is triggered, and load reduction or work stoppage operations are performed, and emergency inspection is initiated. S52: Based on the multi-dimensional assessment results, establish a periodic early warning indicator system to conduct monthly fatigue intensity risk assessments and early warnings, and simultaneously output operation and maintenance suggestions. The specific operation and maintenance principles are set as follows: S521: If fatigue damage is within the expected range, a Level 1 warning is triggered, and the existing inspection cycle is maintained. S522: When fatigue damage accumulates, a level-two warning is triggered, and the relevant floating platform structure is included in the key monitoring list, and the assessment cycle is shortened from monthly to weekly. S523: At that time, fatigue damage accumulates rapidly, triggering a level three warning, and a special fatigue inspection and maintenance plan is formulated and arranged. S524: When fatigue damage approaches a critical state, a level four warning is triggered, a fatigue safety assessment is conducted, and a reinforcement or component replacement plan is planned.

4. The method for constructing a digital twin system for a floating platform based on inverse finite element method according to claim 3, characterized in that, Step S6 specifically includes the following steps: S61: Divide the data into raw collected data and processed data, store them separately, and encapsulate them according to time series; S62: Based on access frequency and time range, the raw collected data and processed data are divided into hot data, warm data and cold data, and different materials are used for storage according to the classification, implementing a three-level layered storage of hot-warm-cold data; S63: Performs local multi-layer backup and off-site disaster recovery backup of data. Local backup uses real-time synchronous mirroring or scheduled backup, while an off-site backup center is established. Incremental data is transmitted daily through the network to build a local-off-site-cloud multimodal collaborative backup mechanism.

Images