Building safety assessment method and system based on optical fiber strain detection

Abstract

The invention provides a building safety assessment method and system based on optical fiber strain detection, and relates to the technical field of force detection and data processing, and the method comprises the steps: carrying out the scene segmentation of a target building structure, configuring an optical fiber sensing array, receiving optical fiber sensing data, combining a stress-optical database, carrying out the signal preprocessing and strain calculation, and determining a four-dimensional strain field; a fuzzy hierarchy module is supervised and trained, a temperature drift compensation branch is introduced, temperature drift compensation correction and building risk assessment are performed, a risk distribution field is determined, and sub-level assessment fitting is performed, so that the technical problems of high difficulty in strain acquisition and data processing and limited analysis precision and accuracy in the prior art are solved; and the building safety assessment error is large. Data validity is improved through dynamic and static scene segmentation and optimized sampling, fuzzy analytic hierarchy process and temperature drift compensation are executed with a stress-light database as an analysis base line, and the precision and efficiency of strain analysis and building evaluation can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of force detection and data processing, and in particular to a building safety assessment method and system based on optical fiber strain detection. Background Art

[0002] Fiber optic strain sensors can monitor the health of structures and assess their safety in real time by measuring changes in strain under load. Fiber optic strain detection technology has been widely used in critical infrastructure such as bridges, dams, and high-rise buildings.

[0003] Currently, because sensing signals in fiber-optic sensor networks are generated through optical interference, data collection and processing are challenging. In practical applications, effectively extracting useful information from large amounts of data and accurately analyzing and assessing it remains a critical challenge. Furthermore, the application of fiber-optic sensing technology in building safety assessment is still in its infancy, and relevant technical standards and specifications are still incomplete.

[0004] In summary, existing technologies have technical problems such as difficulty in strain collection and data processing, and limited analysis precision and accuracy, which lead to large errors in building safety assessment. Summary of the Invention

[0005] This application provides a building safety assessment method and system based on optical fiber strain detection, which is used to solve the technical problems existing in the existing technology in strain collection and data processing, limited analysis precision and accuracy, and resulting in large errors in building safety assessment.

[0006] In view of the above problems, the present application provides a building safety assessment method and system based on optical fiber strain detection.

[0007] In a first aspect, the present application provides a building safety assessment method based on optical fiber strain detection, the method comprising:

[0008] Interact with the target building structure, perform scene segmentation, and configure the optical fiber sensor array, wherein the segmented scene includes at least a static building scene and a dynamic building scene;

[0009] receiving optical fiber sensing data from the optical fiber sensing array, performing signal preprocessing and strain calculation in combination with a strain-light database to determine a four-dimensional strain field, wherein the strain-light database includes a mapping relationship based on external strain, internal optical properties, and optical signal transmission characteristics, and the optical fiber sensing array optimizes sampling based on a Markov probability distribution;

[0010] Based on the building risk class, a fuzzy set based on element membership is determined, and a fuzzy hierarchical module is supervised and trained in combination with a hierarchical structure, wherein the hierarchical structure includes a target layer, a criterion layer, and a solution layer;

[0011] A temperature drift compensation branch is introduced to perform temperature drift compensation correction on the four-dimensional strain field. The building risk assessment is performed in combination with the fuzzy hierarchical module to determine the risk distribution field. The risk distribution field includes sub-level assessments based on multi-scenario partitioning and multi-risk categories.

[0012] Traversing the risk distribution field, performing sub-level assessment fitting on the target building structure, and determining a comprehensive assessment result;

[0013] The risk distribution field and the comprehensive assessment result are integrated to perform visualization on a terminal display device.

[0014] In a second aspect, the present application provides a building safety assessment system based on optical fiber strain detection, the system comprising:

[0015] A sensor configuration unit, the sensor configuration unit is used to interact with the target building structure and perform scene segmentation, and configure the optical fiber sensor array, wherein the segmented scene includes at least a static building scene and a dynamic building scene;

[0016] a strain analysis unit configured to receive optical fiber sensing data from the optical fiber sensing array, perform signal preprocessing and strain calculation in conjunction with a strain-light database, and determine a four-dimensional strain field, wherein the strain-light database includes a mapping relationship based on external strain, internal optical properties, and optical signal transmission characteristics, and the optical fiber sensing array optimizes sampling based on a Markov probability distribution;

[0017] A module construction unit, wherein the module construction unit is used to determine a fuzzy set based on element membership based on the building risk class, and supervise the training of a fuzzy hierarchical module in combination with a hierarchical structure, wherein the hierarchical structure includes a target layer, a criterion layer, and a solution layer;

[0018] a risk assessment unit, the risk assessment unit being configured to introduce a temperature drift compensation branch, perform temperature drift compensation correction on the four-dimensional strain field, and perform building risk assessment in combination with the fuzzy hierarchy module to determine a risk distribution field, the risk distribution field including sub-level assessments based on multi-scenario partitioning and multi-risk categories;

[0019] An evaluation and fitting unit, configured to traverse the risk distribution field, perform sub-level evaluation and fitting on the target building structure, and determine a comprehensive evaluation result;

[0020] A visualization unit is used to integrate the risk distribution field and the comprehensive assessment result to perform visualization on a terminal display device.

[0021] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0022] The present invention provides a building safety assessment method based on optical fiber strain detection. The method interacts with a target building structure, performs scenario-based segmentation, and configures an optical fiber sensor array. The method receives optical fiber sensor data from the optical fiber sensor array and, in combination with a signal-to-optical database, performs signal preprocessing and strain calculation to determine a four-dimensional strain field. Based on the building risk category, a fuzzy set based on element membership is determined. A fuzzy hierarchy module is supervised for training, and a temperature drift compensation branch is introduced to compensate for temperature drift in the four-dimensional strain field. The fuzzy hierarchy module is then used to perform building risk assessment, determine a risk distribution field, traverse the risk distribution field, perform sub-level assessment fitting on the target building structure, and determine a comprehensive assessment result. The risk distribution field and comprehensive assessment result are then integrated and visualized on a terminal display device. This method addresses the technical issues of existing technologies, such as the difficulty in strain acquisition and data processing, limited analysis precision and accuracy, and large errors in building safety assessments. By using dynamic and static scene segmentation and optimized sampling to improve data validity, and using the signal-to-optical database as the analysis baseline, fuzzy hierarchy analysis and temperature drift compensation are performed to effectively improve the accuracy and efficiency of strain analysis and building assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flow chart of a building safety assessment method based on optical fiber strain detection is provided for this application;

[0024] Figure 2 A schematic diagram of the structure of a building safety assessment system based on optical fiber strain detection is provided for this application.

[0025] Description of reference numerals:

[0026] Sensing configuration unit 11, strain analysis unit 12, module construction unit 13, risk assessment unit 14, assessment and fitting unit 15, visualization unit 16. DETAILED DESCRIPTION

[0027] This application provides a building safety assessment method and system based on optical fiber strain detection, performs scenario segmentation on the target building structure, configures an optical fiber sensor array, receives optical fiber sensor data, combines the strain-optical database, performs signal preprocessing and strain calculation, determines the four-dimensional strain field, supervises the training of fuzzy hierarchical modules, introduces a temperature drift compensation branch, performs temperature drift compensation correction and building risk assessment, determines the risk distribution field and performs sub-level assessment fitting, determines the comprehensive assessment results and performs terminal visualization, and is used to solve the technical problems existing in the existing technology in strain collection and data processing, limited analysis precision and accuracy, and resulting in large errors in building safety assessment.

[0028] Example 1

[0029] like Figure 1 As shown, the present application provides a building safety assessment method based on optical fiber strain detection, the method comprising:

[0030] S1: Interact with the target building structure and perform scene segmentation, configuring the fiber optic sensor array, where the segmentation scene includes at least a static building scene and a dynamic building scene;

[0031] Specifically, the target building is the entire building to be assessed for safety. The building's dimensions, material properties, and current structural state are determined as physical structural data and integrated into a 3D topological structure. Exemplary data is obtained through 3D scanning or Building Information Modeling (BIM). The target building structure is then divided into different monitoring areas based on its functional and structural characteristics. This typically involves the division of static and dynamic building scenes.

[0032] Furthermore, the classification criteria for static building scenes are determined, encompassing the main structural components of the building, such as columns, beams, and walls. These components primarily bear static gravity and undergo minimal change, but are crucial to the overall safety of the building. The classification criteria for dynamic building scenes are also determined, encompassing structural areas potentially affected by external dynamic factors, such as windows, doors, and exterior walls. These areas may experience dynamic strain due to wind, traffic, or other vibration sources. Based on this, the target building structure is divided into scenes, with the static and dynamic building scenes each comprising multiple structural partitions.

[0033] Finally, based on the aforementioned segmented scenes, appropriate locations are selected to configure the fiber optic sensor array. Preferably, adaptive sensors are configured for different scenarios, for example, fiber grating sensors for static building scenes and Bragg grating sensors for dynamic building scenes. In static scenes, sensors are primarily deployed at key locations on the load-bearing structure to continuously monitor possible minor displacements or cracks. In dynamic scenes, sensors are concentrated in areas significantly affected by external factors to monitor and record structural strain caused by environmental changes. This distributed sensor network can collect building status data in real time, providing real-time, accurate data support for subsequent structural health monitoring and safety assessments.

[0034] S2: Receive optical fiber sensing data from the optical fiber sensing array, perform signal preprocessing and strain calculation in combination with a strain-light database, and determine a four-dimensional strain field, wherein the strain-light database includes a mapping relationship based on external strain, internal optical properties, and optical signal transmission characteristics, and the optical fiber sensing array optimizes sampling based on a Markov probability distribution;

[0035] The optical fiber sensor array is based on Markov probability distribution optimization sampling, and step S2 of the present application further includes:

[0036] Determine multiple levels of strain states and set sampling probabilities; map the multiple levels of strain states to the sampling probabilities, perform positive serialization integration based on strain trends, and generate a Markov probability chain; if the upper sampling probability is greater than or equal to a first probability threshold, generate a lower sampling instruction; if the upper sampling probability is less than the first probability threshold and the external strain meets the first strain threshold, generate a lower sampling instruction.

[0037] Specifically, target building strain data is collected based on the fiber optic sensor array and transmitted through the optical fiber. When the optical fiber is subjected to external strain, its internal optical properties, such as refractive index and optical path, will change, causing the transmission characteristics of the optical signal in the optical fiber to also change. Based on this, the building strain condition is analyzed.

[0038] When performing optical fiber strain acquisition, continuous and uninterrupted acquisition can easily lead to data redundancy, and the processing volume is too large under massive data. In the embodiment of the present application, the data sampling strategy of the optical fiber sensor array is optimized using a method based on Markov probability distribution.

[0039] Among them, different building structure areas have different risk types, such as position offset, cracks, settlement, etc., and the strain trend characteristics of different risk types are different. A first risk class is determined, and multiple strain states under the first risk class are divided, wherein the sampling probability is positively correlated with the multiple strain states. Based on the strain trend, the strain states of the first risk class are positively serialized and integrated, and the sampling probabilities of the sequence nodes are identified as the first Markov probability chain of the first risk class. The building risk classes are traversed to determine the Nth Markov probability chain of the Nth risk class. The first Markov probability chain is integrated and regularized until the Nth Markov probability chain is determined.

[0040] The Markov probability chain guides the sampling process, ensuring that the sampling frequency is increased during critical strain states and decreased during stable or low-risk states. The first probability threshold is a rigid standard for executing sampling, and the first strain threshold is a threshold for the scale of building structural strain. These two thresholds can be customized based on expert experience. Specifically, if the probability of the strain state corresponding to the previous sampling node is higher than or equal to the set first probability threshold, the next level of sampling instruction will be triggered. Conversely, if the probability of the strain state corresponding to the previous sampling node is lower than the first probability threshold, but the external strain reaches the first strain threshold, the sampling instruction will also be triggered.

[0041] The Markov probability chain guides the sampling process and receives the returned fiber optic sensor data. Signal processing and strain calculation are performed in conjunction with the stress-light database. When an optical fiber is subjected to external strain, its internal optical properties, such as refractive index and optical path length, change. This causes changes in the optical signal's transmission characteristics within the fiber, such as phase, intensity, and frequency, to sense external stress quantities. The stress-light database is constructed by mining the relative mapping relationships between external strain, internal optical properties, and optical signal transmission characteristics. Exemplarily, this mining is performed in conjunction with historical sensor data, using a preset ratio as a confidence constraint.

[0042] The fiber optic sensing data is preprocessed, such as filtering for noise reduction and signal amplification, to improve optical signal accuracy. Furthermore, in conjunction with the stress-optical database, strain calculations are performed based on the relative relationship between the trend vectors of optical properties and transmission characteristics, determining stress data. This stress data is spatially mapped to the target building structure. The stress data is then distributed in three dimensions to determine a three-dimensional strain field. A time axis is introduced to establish a four-dimensional strain field, which is updated synchronously with sensor detection. This four-dimensional strain field can detail the strain state of the building structure at different points in time and space, which is crucial for understanding the structural health of the building.

[0043] In summary, resources can be used more efficiently and attention can be focused on those states that are most likely to show structural risks, thereby optimizing monitoring costs while ensuring monitoring quality.

[0044] S3: Based on the building risk class, determine the fuzzy set based on the element membership, combine the hierarchical structure, and supervise the training of the fuzzy hierarchical module, wherein the hierarchical structure includes the target layer, the criterion layer, and the solution layer;

[0045] Among them, the supervised training fuzzy hierarchy module, step S3 of this application also includes:

[0046] The building risk classes are traversed, subordinate risk elements are mined and membership evaluation is performed, multiple mapping trees are determined, and the fuzzy sets are integrated to determine the membership, wherein the membership ∈ [0, 1]; the hierarchical structure is constructed, and building safety assessment records are called for hierarchical integration to determine training samples; based on the fuzzy sets and the hierarchical structure, driven training based on the training samples is performed to generate the fuzzy hierarchical module.

[0047] The building risk categories represent the types of risks that may exist in the target building, such as structural risks (such as structural drift and settlement) and environmental risks. For each risk category, the associated risk elements are identified, specifically the specific factors that contribute to the risk, such as cracks, corrosion, and overload. Each risk element is assigned a membership degree, a value between 0 and 1 that indicates the strength of the element's association with the corresponding risk category. A higher membership degree indicates a stronger correlation between the element and the risk category.

[0048] The building risk classes are traversed, and for each risk class, a mapping tree is constructed. This is a mapping structure between risk classes and subordinate risk elements, with each risk element identified with a degree of membership. Multiple mapping trees correspond one-to-one with the building risk classes, organizing and classifying the subordinate risk elements according to their corresponding risk categories. These multiple mapping trees are integrated into the fuzzy set, which reflects the risk status of the entire building and the relationships between elements.

[0049] Furthermore, a hierarchical structure is constructed, comprising a target layer, a criteria layer, and a solution layer. The target layer represents the overall assessment objective, such as the overall safety of a building; the criteria layer includes the various standards or criteria used to assess the objective; and the solution layer involves specific assessment plans or measures. This hierarchical structure facilitates the systematic and standardized analysis and processing of risk assessment data.

[0050] Furthermore, existing building safety assessment records are integrated according to the constructed hierarchical structure to extract sample data for training. Based on this sample data, supervised training is performed using the organized fuzzy sets and hierarchical structure as a baseline. This training process utilizes a machine learning algorithm to train the model using historical data, enabling the model to output accurate safety assessments based on the input risk elements and hierarchical structure. Preferably, a convergence check is performed after training is complete. If convergence conditions are not met, sample reorganization and re-checking can be performed until convergence conditions are met, resulting in the completed fuzzy hierarchical module.

[0051] The trained fuzzy hierarchical model will be able to provide decision makers with clear and quantitative recommendations on building safety risks. Building safety assessments will become more scientific and precise, enabling better identification and management of potential risks, thereby improving the safety and durability of buildings.

[0052] The mining of the risk elements and the evaluation of the membership are carried out to determine multiple mapping trees. Step S3 of this application also includes:

[0053] Determine a first building risk class and a first subordinate risk element, wherein the risk element is a strain characteristic; determine building characteristics of a scene area corresponding to the first building risk class, wherein the building characteristics include structural characteristics, material characteristics, and load characteristics; using the building characteristics as a baseline, traverse the first subordinate risk element to perform a membership assessment and determine a membership layer; and determine a first mapping tree based on the first building risk class and the membership layer.

[0054] First, the first building risk category needs to be identified. This can be any of the main risk areas of the building, for example, structural safety risk. This risk typically involves the risk that the building may not be able to withstand the expected loads or may fail structurally during future use.

[0055] Further identify subordinate risk elements associated with the first risk category. For the structural safety risk category, these elements may include strain characteristics such as the appearance of cracks, concrete peeling, and steel corrosion. These elements are directly related to the integrity and long-term durability of the structure.

[0056] Next, the building characteristics of the scenario area corresponding to the first building risk class are identified, including structural characteristics, such as the design load of the building, the type and layout of the supporting structure; material characteristics, such as the concrete grade used, the type and quality of steel; load characteristics, such as the static load and possible dynamic loads borne by the building, such as wind load and snow load.

[0057] Using the identified building characteristics as a baseline, the first-level risk element is assessed for membership. This involves analyzing how each risk element interacts with the building characteristics and how these interactions may affect the building's structural safety. The membership assessment results reflect each element's contribution to the risk category. A higher membership indicates a greater potential threat to the building's structural safety.

[0058] Finally, based on the first building risk category and the assessed membership layer, the first mapping tree is constructed. A mapping tree is a structured diagram that illustrates the relationship between different risk elements and building characteristics, and their impact on overall building risk. This helps decision-makers understand which structural elements require priority attention and reinforcement to improve building safety and reliability. This allows for systematic assessment and management of building structural safety risks, effectively identifying and mitigating potential risk factors, and ensuring the long-term safe use of the building.

[0059] S4: Introducing a temperature drift compensation branch to perform temperature drift compensation correction on the four-dimensional strain field, combining the fuzzy hierarchical module to perform building risk assessment and determine a risk distribution field, which includes sub-level assessments based on multi-scenario partitioning and multi-risk categories;

[0060] Wherein, the introduction of the temperature drift compensation branch, step S4 of the present application further includes:

[0061] For multiple temperature levels, the cross-sensitivity coefficient based on temperature and stress is measured, the cross-sensitivity coefficient is traversed, and the strain compensation ratio is mined in combination with strain detection samples. A mapping baseline of the multiple temperature levels-cross-sensitivity coefficient-strain compensation ratio is established, and sample-driven training is performed based on the strain detection samples to generate the temperature drift compensation branch. The sensed temperature field is synchronously transmitted back and, in combination with the temperature drift compensation branch, compensation correction is performed on the four-dimensional strain field.

[0062] Since fiber optic sensors are sensitive to both temperature and stress, they cannot directly distinguish the source of changes in internal optical properties or changes in light transmission characteristics. Therefore, stress compensation is performed to improve the accuracy of strain data, taking into account the influence of temperature.

[0063] First, it's necessary to identify the different temperature levels to which the building may be exposed. These levels should cover the range of possible temperatures, from very low to very high, to ensure effective monitoring under all operating conditions. For each temperature level, the cross-sensitivity coefficient between temperature and stress is determined. This cross-sensitivity coefficient indicates how the stress response of a building material is affected by temperature changes under different temperature conditions. A high cross-sensitivity coefficient means that the stress measurement is more sensitive to temperature changes under those temperature conditions. Using actual strain gauges, the strain data under different temperature conditions is studied to identify the specific strain compensation ratio. The strain compensation ratio refers to the strain adjustment ratio required to offset the error caused by temperature drift.

[0064] Furthermore, a mapping baseline is established based on the relationship between multiple temperature levels, cross-sensitivity coefficients, and strain compensation ratios. This baseline provides a reference for subsequent automatic compensation, ensuring the accuracy of strain data under any temperature conditions.

[0065] The previously mined strain detection samples are used for sample-driven training to generate a temperature drift compensation branch. This branch uses machine learning to automatically adjust strain data at different temperatures to eliminate the effects of temperature changes.

[0066] The synchronously transmitted sensor temperature field data is combined with the temperature drift compensation branch to perform real-time compensation correction on the four-dimensional strain field. This ensures that the strain data is not erroneously affected by temperature changes when conducting building risk assessments.

[0067] The corrected four-dimensional strain field data is then combined with the fuzzy hierarchy module to perform building risk assessment. The risk distribution field is based on multi-scenario partitioning and multi-risk class assessment, providing detailed sub-level assessments for each part and level.

[0068] In summary, it not only effectively compensates and corrects the impact of temperature, but also provides a comprehensive and detailed building risk assessment to ensure the long-term safety and stability of the building.

[0069] Wherein, the step S4 of determining the risk distribution field in this application further includes:

[0070] The four-dimensional strain field is transmitted back and divided into a static strain field and a dynamic strain field, wherein the division criteria are determined by fitting a primary division based on the building scenario and a secondary division based on a preset strain frequency; combined with the fuzzy hierarchy module, guided by the assessment objectives based on the building risk class, a fuzzy hierarchy analysis is performed on the static strain field and the dynamic strain field, and the building structure spatial distribution is integrated and performed to determine the risk distribution field, wherein the analysis dimensions include strain risk assessment and linear risk trend assessment.

[0071] First, the four-dimensional strain data collected by the fiber optic sensing network, including three spatial dimensions plus the time dimension, is transmitted back to the data processing center in real time. This data provides detailed information about the strain state of the building structure at different time and spatial locations.

[0072] Strain data is categorized according to different building components, such as infrastructure, load-bearing walls, and floors. This categorization, based on the building's structural characteristics and functional areas, aims to identify static areas, such as fixed structures like columns and walls, and dynamic areas, such as exterior walls affected by wind loads, thereby determining a primary division based on the building's scenario.

[0073] Furthermore, based on this initial division, the frequency of change of the strain data within each region is further analyzed. Static strain fields typically exhibit a lower frequency of change, reflecting the long-term deformation or settlement of the structure; dynamic strain fields exhibit a higher frequency of change, which may be caused by external factors such as wind, traffic, and other dynamic loads. Under the same type, different strain frequencies may represent different risk states. Further segmentation is performed based on strain frequency. For example, multi-level frequencies are set, and the dynamic part is mapped and segmented, that is, the strain state of the segmented part is approximated, and the static strain field and dynamic strain field are integrated to determine.

[0074] Guided by assessment objectives for building risks, such as structural integrity, durability, and safety, and using fuzzy sets as analysis criteria, a detailed risk assessment of both static and dynamic strain fields is conducted through fuzzy hierarchical analysis, using time nodes and time zones as analysis criteria. This includes identifying the magnitude of risk, possible development trends, and the need for emergency response measures. Analysis results are then determined for multiple strain points. The analysis results for each strain point include both single-time node risk and time zone risk. These results are then integrated based on the spatial distribution of the building structure to determine the risk distribution field. This distribution field not only displays the risk level of each area but also reflects the spatial distribution and possible development trends of the risk.

[0075] Among them, the strain risk assessment based on time nodes and the linear risk trend assessment based on time zones can comprehensively evaluate the overall safety status of the building and provide data-based decision support, effectively monitor and evaluate the health status of the building structure, ensure the timely detection of potential risks and take corresponding preventive or repair measures, and improve the safety and reliability of the building.

[0076] Before performing the fuzzy hierarchical analysis, step S4 of the present application further includes:

[0077] The four-dimensional strain field is traversed to determine a fuzzy judgment matrix based on the strain characteristics, wherein the fuzzy judgment matrix corresponds to the partitioned strain field and the matrix items are identified with membership degrees; the fuzzy judgment matrix corresponding to the dynamic strain field is extracted, and a prediction linear function is identified; and a fuzzy hierarchical analysis is performed based on the fuzzy judgment matrix and the prediction linear function.

[0078] Specifically, the four-dimensional strain field is traversed to capture the strain characteristics of the building structure at different locations and time points. Key strain characteristics, such as strain magnitude, rate of change, and duration, are extracted from the traversed data. Based on these strain characteristics, the element values ​​of the risk elements in the fuzzy set are matched and integrated to determine the fuzzy judgment matrix.

[0079] Furthermore, special attention is paid to dynamic strain fields, which typically involve strains caused by environmental changes such as temperature fluctuations and wind loads. Data relevant to dynamic scenarios are extracted from the four-dimensional strain field. Based on the dynamic strain data, predictive linear functions are developed. These functions are designed to model future trends in the strain data and help predict possible structural behavior and potential risks. An in-depth fuzzy hierarchical analysis is conducted using the established fuzzy judgment matrix and predictive linear functions. This allows for a comprehensive assessment of the overall impact of different strain characteristics on building safety.

[0080] Simultaneously, decision-making is conducted based on the analysis results. By comparing the impact weights of different strain characteristics, it is determined which structural issues require priority treatment and which can be monitored over the long term. This combined fuzzy logic and hierarchical analysis method provides a scientific and systematic decision-making support tool for building safety assessments, making risk management more precise and efficient. Through this approach, building managers can better understand and respond to structural strains and potential risks caused by various external and internal factors.

[0081] Wherein, the identification prediction linear function, step S4 of this application further includes:

[0082] For the fuzzy judgment matrix corresponding to the dynamic strain field, a linear matrix part and a nonlinear matrix part are determined; the nonlinear matrix part is traversed and a cost function is introduced, wherein the cost function is an approximate linear law; the linear relationship of the linear matrix part is determined, and the prediction linear function is determined by integration and combination in combination with the cost function.

[0083] Specifically, according to the fuzzy judgment matrix corresponding to the dynamic strain field, a linear matrix part and a nonlinear matrix part are divided based on whether it can be directly determined that the change of strain data is in direct proportional relationship with time or other factors, wherein the linear matrix part can be described and predicted using a standard linear relationship.

[0084] For the nonlinear matrix part, the data changes covered are complex behaviors that do not conform to a directly proportional relationship. This may include nonlinear responses due to material fatigue, complex loading conditions, or environmental factors. For the nonlinear matrix part, a cost function is introduced, which is used to quantify the deviation between the prediction model and the actual observed data. In this context, the cost function is designed to approximate linear laws, quantizing the nonlinear matrix part to be approximately linear, and approximately describing these nonlinear behaviors through linearization. Exemplarily, the best linear approximation is found based on the least squares method.

[0085] Furthermore, the linear relationship of the linear matrix part is determined, combined with the cost function, and the prediction linear function is determined by integration, so that the linear and nonlinear behaviors in the dynamic strain field can be processed and predicted simultaneously based on linear means.

[0086] In summary, the behavior of building structures under various loads and environmental conditions can be understood and predicted more comprehensively and efficiently, thereby improving the safety and reliability of the structures.

[0087] S5: traverse the risk distribution field, perform sub-level assessment fitting on the target building structure, and determine a comprehensive assessment result;

[0088] S6: Integrate the risk distribution field and the comprehensive assessment result, and perform visualization on a terminal display device.

[0089] Specifically, the risk distribution field contains a detailed view of the risk assessment for each area within the building structure. Each area is assessed and labeled with a specific risk level. These areas may be categorized based on different risk categories, such as structural risk, environmental risk, and risk factors, such as strain and load variation.

[0090] The risk distribution field is traversed, risk data of each measuring point corresponding to the fiber optic sensor array is identified, and based on the global importance of the target building, a weight distribution is determined. The risk data is weighted by coefficient calculation to determine the comprehensive assessment result, which reflects the overall safety status of the entire building structure and points out key risk areas and issues that require special attention.

[0091] Preferably, during the comprehensive assessment process, risks are prioritized according to their severity and urgency. Based on these priorities, corresponding maintenance strategies and emergency response plans are developed.

[0092] Furthermore, the risk distribution field and comprehensive assessment results are integrated into an easy-to-understand format for presentation on terminal display devices. For example, the data can be converted into charts, heat maps, or dynamic simulations to visually display risk distribution and key problem areas. This allows maintenance teams and decision makers to intuitively understand the building's safety status. In summary, building safety assessments are no longer limited to static reports; instead, they become dynamic, interactive tools that update and display critical information in real time, greatly improving the efficiency and effectiveness of building maintenance and risk management.

[0093] The building safety assessment method based on optical fiber strain detection provided in this application has the following technical effects:

[0094] 1. Scenario-based segmentation of the target building structure is performed, and a fiber optic sensor array is configured to receive fiber optic sensor data. Signal preprocessing and strain calculation are performed based on the Ying-Guang database. Four-dimensional strain fields are determined, and supervised training of a fuzzy hierarchical model is conducted. A temperature drift compensation branch is introduced to perform drift compensation correction and building risk assessment. The risk distribution field is determined and sub-level assessment fitting is performed. This addresses the technical issues of existing technologies, such as the difficulty in strain collection and data processing, limited analysis precision and accuracy, and large errors in building safety assessments. By improving data validity through dynamic and static scene segmentation and optimized sampling, and using the Ying-Guang database as the analysis baseline, fuzzy hierarchical analysis and temperature drift compensation are performed, effectively improving the accuracy and efficiency of strain analysis and building assessment.

[0095] 2. By performing scene segmentation and conducting targeted configuration of sensors based on movement and stillness, and subsequent targeted analysis, the effectiveness of processing can be effectively improved.

[0096] 3. Introducing the strain-optical database, strain calculation is performed based on the mapping relationship between external strain, internal optical properties, and optical signal transmission characteristics, improving compliance and logic. Taking time and space as the three dimensions, a comprehensive analysis combining dynamic and static aspects is achieved, improving the completeness of the analysis.

[0097] 4. Introducing fuzzy sets and hierarchical structures, and conducting building risk assessment through modeling can effectively improve the logic and effectiveness of the analysis. At the same time, introducing temperature drift compensation and compensating for detection drawbacks based on the characteristics of fiber optic sensors can effectively improve assessment accuracy.

[0098] Example 2

[0099] Based on the same inventive concept as the building safety assessment method based on optical fiber strain detection in the aforementioned embodiment, Figure 2 As shown, the present application provides a building safety assessment system based on optical fiber strain detection, the system comprising:

[0100] A sensor configuration unit 11 is used to interact with the target building structure and perform scene segmentation, and configure a fiber optic sensor array, wherein the segmented scene includes at least a static building scene and a dynamic building scene;

[0101] a strain analysis unit 12, configured to receive optical fiber sensing data from the optical fiber sensing array, perform signal preprocessing and strain calculation in conjunction with a strain-light database, and determine a four-dimensional strain field, wherein the strain-light database includes a mapping relationship based on external strain, internal optical properties, and optical signal transmission characteristics, and the optical fiber sensing array optimizes sampling based on a Markov probability distribution;

[0102] A module construction unit 13 is used to determine a fuzzy set based on element membership based on the building risk class, and supervise the training of a fuzzy hierarchical module in combination with a hierarchical structure, wherein the hierarchical structure includes a target layer, a criterion layer, and a solution layer;

[0103] a risk assessment unit 14, which is configured to introduce a temperature drift compensation branch, perform temperature drift compensation correction on the four-dimensional strain field, and perform building risk assessment in combination with the fuzzy hierarchy module to determine a risk distribution field, wherein the risk distribution field includes sub-level assessments based on multiple scenario partitions and multiple risk classes;

[0104] An evaluation and fitting unit 15 is configured to traverse the risk distribution field, perform sub-level evaluation and fitting on the target building structure, and determine a comprehensive evaluation result;

[0105] The visualization unit 16 is used to integrate the risk distribution field and the comprehensive assessment result to perform visualization on a terminal display device.

[0106] Furthermore, the strain analysis unit 12 is further configured to perform the following steps:

[0107] Determine multiple levels of strain states and set sampling probabilities; map the multiple levels of strain states to the sampling probabilities, perform positive serialization integration based on strain trends, and generate a Markov probability chain; if the upper sampling probability is greater than or equal to a first probability threshold, generate a lower sampling instruction; if the upper sampling probability is less than the first probability threshold and the external strain meets the first strain threshold, generate a lower sampling instruction.

[0108] The module construction unit 13 is further configured to perform the following steps:

[0109] The building risk classes are traversed, subordinate risk elements are mined and membership evaluation is performed, multiple mapping trees are determined, and the fuzzy sets are integrated to determine the membership, wherein the membership ∈ [0, 1]; the hierarchical structure is constructed, and building safety assessment records are called for hierarchical integration to determine training samples; based on the fuzzy sets and the hierarchical structure, driven training based on the training samples is performed to generate the fuzzy hierarchical module.

[0110] The module construction unit 13 is further configured to perform the following steps:

[0111] Determine a first building risk class and a first subordinate risk element, wherein the risk element is a strain characteristic; determine building characteristics of a scene area corresponding to the first building risk class, wherein the building characteristics include structural characteristics, material characteristics, and load characteristics; using the building characteristics as a baseline, traverse the first subordinate risk element to perform a membership assessment and determine a membership layer; and determine a first mapping tree based on the first building risk class and the membership layer.

[0112] The risk assessment unit 14 is further configured to perform the following steps:

[0113] The four-dimensional strain field is transmitted back and divided into a static strain field and a dynamic strain field, wherein the division criteria are determined by fitting a primary division based on the building scenario and a secondary division based on a preset strain frequency; combined with the fuzzy hierarchy module, guided by the assessment objectives based on the building risk class, a fuzzy hierarchy analysis is performed on the static strain field and the dynamic strain field, and the building structure spatial distribution is integrated and performed to determine the risk distribution field, wherein the analysis dimensions include strain risk assessment and linear risk trend assessment.

[0114] The risk assessment unit 14 is further configured to perform the following steps:

[0115] The four-dimensional strain field is traversed to determine a fuzzy judgment matrix based on the strain characteristics, wherein the fuzzy judgment matrix corresponds to the partitioned strain field and the matrix items are identified with membership degrees; the fuzzy judgment matrix corresponding to the dynamic strain field is extracted, and a prediction linear function is identified; and a fuzzy hierarchical analysis is performed based on the fuzzy judgment matrix and the prediction linear function.

[0116] The risk assessment unit 14 is further configured to perform the following steps:

[0117] For the fuzzy judgment matrix corresponding to the dynamic strain field, a linear matrix part and a nonlinear matrix part are determined; the nonlinear matrix part is traversed and a cost function is introduced, wherein the cost function is an approximate linear law; the linear relationship of the linear matrix part is determined, and the prediction linear function is determined by integration and combination in combination with the cost function.

[0118] The risk assessment unit 14 is further configured to perform the following steps:

[0119] For multiple temperature levels, the cross-sensitivity coefficient based on temperature and stress is measured, the cross-sensitivity coefficient is traversed, and the strain compensation ratio is mined in combination with strain detection samples. A mapping baseline of the multiple temperature levels-cross-sensitivity coefficient-strain compensation ratio is established, and sample-driven training is performed based on the strain detection samples to generate the temperature drift compensation branch. The sensed temperature field is synchronously transmitted back and, in combination with the temperature drift compensation branch, compensation correction is performed on the four-dimensional strain field.

[0120] Through the above detailed description of the building safety assessment method based on optical fiber strain detection in this specification, those skilled in the art can clearly understand the building safety assessment method and system based on optical fiber strain detection in this embodiment. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0121] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A building safety assessment method based on optical fiber strain detection, characterized in that: The method comprises: Interact with the target building structure, perform scene segmentation, and configure the optical fiber sensor array, wherein the segmentation scene at least includes a static building scene and a dynamic building scene; Receiving optical fiber sensing data of the optical fiber sensing array, performing signal preprocessing and strain calculation in combination with a strain-light database, and determining a four-dimensional strain field, wherein the strain-light database includes a mapping relationship based on external strain, internal optical properties, and optical signal transmission characteristics, and the optical fiber sensing array optimizes sampling based on Markov probability distribution; Based on the building risk class, a fuzzy set based on element membership is determined, and a fuzzy hierarchical module is supervised and trained in combination with a hierarchical structure, wherein the hierarchical structure includes a target layer, a criterion layer, and a solution layer; A temperature drift compensation branch is introduced to perform temperature drift compensation correction on the four-dimensional strain field, and the building risk assessment is performed in combination with the fuzzy hierarchy module to determine the risk distribution field, which includes sub-level assessments based on multi-scenario partitions and multi-risk categories; Traversing the risk distribution field, performing sub-level assessment fitting on the target building structure, and determining a comprehensive assessment result; The risk distribution field and the comprehensive assessment result are integrated to perform visualization on a terminal display device.

2. The building safety assessment method based on optical fiber strain detection according to claim 1, characterized in that: The optical fiber sensor array optimizes sampling based on Markov probability distribution, including: Determine multi-level strain states and set sampling probabilities; Mapping the multi-level strain states and the sampling probabilities, performing positive serialization integration based on strain trends, and generating a Markov probability chain; If the upper sampling probability is greater than or equal to the first probability threshold, a lower sampling instruction is generated; If the upper sampling probability is less than the first probability threshold and the external strain satisfies the first strain threshold, a lower sampling instruction is generated.

3. The building safety assessment method based on optical fiber strain detection according to claim 1, characterized in that: The supervised training fuzzy hierarchy module comprises: Traversing the building risk classes, mining the affiliated risk elements and evaluating the membership, determining multiple mapping trees, integrating and determining the fuzzy set, wherein the membership ∈ [0, 1]; Constructing the hierarchical structure, calling the building safety assessment records for hierarchical integration, and determining training samples; Based on the fuzzy set and the hierarchical structure, driving training based on the training samples is performed to generate the fuzzy hierarchical module.

4. The building safety assessment method based on optical fiber strain detection according to claim 3, characterized in that: The mining of the affiliated risk elements and the evaluation of the affiliated degrees determine multiple mapping trees, including: Determine a first building risk class, determine a first subordinate risk element, wherein the risk element is a strain feature; Determining the building characteristics of the scene area corresponding to the first building risk class, wherein the building characteristics include structural characteristics, material characteristics and load characteristics; Taking the building characteristics as a baseline, traversing the first subordinate risk element to perform a subordinate degree assessment and determine a subordinate layer; Based on the first building risk class and the membership layer, a first mapping tree is determined.

5. The building safety assessment method based on optical fiber strain detection according to claim 4, characterized in that: The step of determining the risk distribution field comprises: Transmitting back the four-dimensional strain field, dividing the static strain field and the dynamic strain field, wherein the division standard is determined by fitting a primary division based on a building scene and a secondary division based on a preset strain frequency; Combined with the fuzzy hierarchy module, guided by the assessment target based on the building risk class, fuzzy hierarchy analysis is performed on the static strain field and the dynamic strain field, and the building structure spatial distribution is integrated and performed to determine the risk distribution field, wherein the analysis dimensions include strain risk assessment and linear risk trend assessment.

6. The building safety assessment method based on optical fiber strain detection according to claim 5, characterized in that: Before conducting fuzzy hierarchical analysis, including: Traversing the four-dimensional strain field, determining a fuzzy judgment matrix based on the strain characteristics, wherein the fuzzy judgment matrix corresponds to the partitioned strain field, and the matrix items are marked with membership degrees; Extracting the fuzzy judgment matrix corresponding to the dynamic strain field, and identifying a prediction linear function; Based on the fuzzy judgment matrix and the prediction linear function, fuzzy hierarchy analysis is performed.

7. The building safety assessment method based on optical fiber strain detection according to claim 6, characterized in that: The identification prediction linear function includes: For the fuzzy judgment matrix corresponding to the dynamic strain field, determining a linear matrix part and a nonlinear matrix part; Traversing the nonlinear matrix part, introducing a cost function, wherein the cost function is an approximate linear law; The linear relationship of the linear matrix part is determined, combined with the cost function, and integrated to determine the predicted linear function.

8. The building safety assessment method based on optical fiber strain detection according to claim 1, characterized in that: The introducing of the temperature drift compensation branch comprises: For multiple temperature levels, the cross-sensitivity coefficient based on temperature and stress is determined, the cross-sensitivity coefficient is traversed, and the strain compensation ratio is mined in combination with the strain detection samples; Establishing the mapping baseline of the multi-temperature level-cross sensitivity coefficient-strain compensation ratio, performing sample drive training based on the strain detection sample, and generating the temperature drift compensation branch; The sensed temperature field is synchronously transmitted back and combined with the temperature drift compensation branch to compensate and correct the four-dimensional strain field.

9. Building safety assessment system based on optical fiber strain detection, characterized in that: The system is used to implement the building safety assessment method based on optical fiber strain detection according to any one of claims 1 to 8, and the system comprises: A sensor configuration unit, the sensor configuration unit is used to interact with the target building structure, perform scene segmentation, and configure the optical fiber sensor array, wherein the segmented scene at least includes a static building scene and a dynamic building scene; A strain analysis unit, the strain analysis unit is used to receive the optical fiber sensing data of the optical fiber sensing array, perform signal preprocessing and strain calculation in combination with a strain-light database, and determine a four-dimensional strain field, wherein the strain-light database contains a mapping relationship based on external strain-internal optical properties-optical signal transmission characteristics, and the optical fiber sensing array optimizes sampling based on Markov probability distribution; A module construction unit, the module construction unit is used to determine a fuzzy set based on element membership based on the building risk class, and supervise the training of a fuzzy hierarchical module in combination with a hierarchical structure, wherein the hierarchical structure includes a target layer, a criterion layer, and a solution layer; A risk assessment unit, the risk assessment unit is used to introduce a temperature drift compensation branch, perform compensation correction based on temperature drift on the four-dimensional strain field, perform building risk assessment in combination with the fuzzy hierarchy module, and determine a risk distribution field, wherein the risk distribution field includes sub-level assessments based on multi-scenario partitions and multi-risk classes; An evaluation and fitting unit, the evaluation and fitting unit being used to traverse the risk distribution field, perform sub-level evaluation and fitting on the target building structure, and determine a comprehensive evaluation result; A visualization unit is used to integrate the risk distribution field and the comprehensive evaluation result to perform visualization on a terminal display device.

Images