A highway construction engineering quality digital detection method and system

Abstract

The application discloses a kind of highway construction engineering quality digitization detection method and system, belong to data processing field, this highway construction engineering quality digitization detection method includes following steps: the full life cycle performance data of key position of engineering facility is collected and is washed, alignment and feature engineering, the problem of prior art "heavy short-term qualified, light long-term performance" is systematically solved, by collecting and processing full life cycle performance data, long-term evaluation data foundation is laid;Build digital twin model fusing real-time data, and integrate mechanics and material degradation algorithm, realize the dynamic simulation of long-term performance evolution of structure under the action of environment and load, implant machine learning algorithm, realize real-time evaluation of structural damage and prediction of future performance degradation trend, so that potential risks can be identified and warned in early operation, so as to change quality control from passive acceptance to active performance guarantee throughout the life cycle.

Description

Technical Field

[0001] This invention belongs to the field of data processing, and in particular relates to a digital detection method and system for the quality of highway construction projects. Background Technology

[0002] In current tunnel construction, the pass rate of individual projects is typically used as the core criterion for acceptance, for example, requiring a pass rate of over 95% for key projects. However, this pass rate-oriented evaluation model, primarily focused on whether minimum quality thresholds are met upon completion, has significant limitations: it easily leads participating parties to prioritize short-term, static acceptance indicators, lacking sufficient consideration and evaluation of the long-term performance throughout the project's life cycle (such as material durability, structural fatigue resistance, and performance evolution under environmental influences). This may result in higher performance degradation and maintenance risks in the later stages of operation, risks that are difficult to effectively identify and prevent during final acceptance. Therefore, existing engineering quality inspection standards focus more on verifying immediate quality and compliance upon completion, failing to comprehensively assess the long-term performance and durability of the project throughout its life cycle, and require improvement. Summary of the Invention

[0003] Therefore, it is necessary to provide a digital inspection method and system for the quality of highway construction projects to address the above-mentioned problems.

[0004] The present invention is implemented as follows: a digital inspection method for the quality of highway construction projects includes the following steps:

[0005] Collect full life cycle performance data (stress, strain, displacement, cracks, temperature and humidity, corrosion, etc.) of key parts (tunnel lining, bridge bearings, etc.) of engineering facilities (such as tunnels and bridges), and clean, align and feature-engineer to obtain spatiotemporally consistent full life cycle performance data.

[0006] Construct a BIM model (Building Information Model) corresponding to the real engineering facility, control the integration of spatiotemporally consistent full life cycle performance data of the BIM model, and create a digital twin model that reflects the real engineering facility. The digital twin model can not only present the current state of the structure in three dimensions, but also dynamically simulate the long-term performance evolution process of the structure under real loads and environmental effects by integrating mechanical and material degradation algorithms.

[0007] In the digital twin model, machine learning algorithms are embedded to mine massive amounts of historical monitoring data and real-time streaming data. The machine learning algorithms establish a quantitative mapping relationship between data features (such as strain combinations under specific patterns and the rate of change of crack geometric parameters) and structural damage states (such as microcrack initiation, local yielding, and initial corrosion of steel bars) through training. Based on real-time data input, the algorithm automatically identifies and quantifies the type, location, and severity of current damage to obtain assessment results. On this basis, the machine learning algorithm further combines the current damage state, continuously input environmental load data, and embedded material time-varying degradation models to dynamically deduce the degradation trajectory of structural performance indicators (such as bearing capacity and stiffness) over a future set period (through recursive or iterative calculations). Based on preset safety thresholds or failure criteria, the algorithm calculates the probability and time of reaching the critical state and finally outputs the degradation trend and quantitative risk level of the long-term performance of the structure to obtain prediction results.

[0008] In one embodiment, the present invention provides a digital inspection method for the quality of highway construction projects, further comprising:

[0009] An intelligent decision-making and closed-loop feedback mechanism is established. The obtained evaluation results and prediction results are automatically generated into maintenance, repair or reinforcement decisions, including specific measures, priority ranking and implementation timing, through a preset expert rule base and optimization algorithm. The decisions guide the preventive maintenance plan and emergency response during the operation period. The effect data of the decision implementation is collected in real time and fed back to the digital twin model to continuously verify and correct the prediction accuracy of the digital twin model.

[0010] In one embodiment, the present invention provides a digital inspection method for the quality of highway construction projects, further comprising:

[0011] Based on the results of individual damage assessments or predictions, a multi-damage coupling analysis module is introduced to address the cumulative effects of interactions between damages of different locations and types. This module quantifies the synergistic effects between damages (e.g., adjacent microcracks may merge into macro-cracks, or local yielding may lead to stress redistribution and exacerbate corrosion in other areas) through structural topological relationships and mechanical transfer models, thus assessing the overall severity of the damage superposition. Calculated using the following formula:

[0012] ;

[0013] in, This represents the severity index of the i-th injury. Let be the independent weighting coefficients for the i-th damage. Let i be the interaction coefficient between damage i and j. Based on the distance between damages The spatial decay function.

[0014] In one embodiment, the present invention provides a digital inspection method for the quality of highway construction projects, further comprising:

[0015] A multi-damage interaction case database is constructed and continuously maintained to support the continuous optimization of the multi-damage coupling analysis module. This database stores damage combination instances captured in each assessment, corresponding actual monitoring data, and subsequently observed structural responses (such as changes in performance degradation rate or repair effects). Based on this database, data-driven methods (such as statistical regression or machine learning) are periodically used for analysis to correlate the actual long-term impact of different damage combinations with the current interaction coefficients used by the multi-damage coupling analysis module. The predicted results are compared and verified, and then the interaction coefficients are calibrated and updated. .

[0016] In one embodiment, the present invention provides a digital inspection method for the quality of highway construction projects, further comprising:

[0017] The deviation between the predicted mechanical data (such as the predicted strain value) of the digital twin model and the actual monitoring data is continuously compared. When the deviation exceeds the threshold, the parameter calibration process is automatically triggered. The latest monitoring data is used to fine-tune the material constitutive parameters, boundary conditions or load assumptions in the digital twin model through optimization algorithms, so that the simulated behavior of the digital twin model is synchronized with the behavior drift of the physical entity caused by material aging and changes in constraint conditions during long-term service.

[0018] In one embodiment, the present invention provides a digital inspection system for the quality of highway construction projects, comprising:

[0019] The data acquisition module is used to collect full life cycle performance data (stress, strain, displacement, cracks, temperature and humidity, corrosion, etc.) of key parts (tunnel lining, bridge bearings, etc.) of engineering facilities (such as tunnels and bridges), and to clean, align and feature-engineer the data to obtain spatiotemporally consistent full life cycle performance data.

[0020] The model building module is used to build a BIM model corresponding to the real engineering facility, control the integration of spatiotemporally consistent full life cycle performance data of the BIM model, and create a digital twin model that reflects the real engineering facility. The digital twin model can not only present the current state of the structure in three dimensions, but also dynamically simulate the long-term performance evolution process of the structure under real loads and environmental effects by integrating mechanics and material degradation algorithms.

[0021] The damage assessment and prediction module is used to embed machine learning algorithms into the digital twin model to mine massive amounts of historical monitoring data and real-time streaming data. The machine learning algorithm establishes a quantitative mapping relationship between data features (such as strain combinations under specific patterns and the rate of change of crack geometric parameters) and structural damage states (such as microcrack initiation, local yielding, and initial corrosion of steel bars) through training. Based on real-time data input, it automatically identifies and quantifies the type, location, and severity of current damage to obtain assessment results. On this basis, the machine learning algorithm further combines the current damage state, continuously input environmental load data, and the embedded material time-varying degradation model to dynamically deduce the degradation trajectory of the structure's performance indicators (such as bearing capacity and stiffness) over a future set period (through recursive or iterative calculation). Based on preset safety thresholds or failure criteria, it calculates the probability and time of reaching the critical state, and finally outputs the degradation trend and quantitative risk level of the structure's long-term performance to obtain prediction results.

[0022] In one embodiment, the present invention provides a digital inspection system for the quality of highway construction projects, further comprising:

[0023] The intelligent decision feedback module is used to establish an intelligent decision-making and closed-loop feedback mechanism. It automatically generates maintenance, repair or reinforcement decisions, including specific measures, priority ranking and implementation timing, based on the obtained evaluation results and prediction results through a preset expert rule base and optimization algorithm. The decisions guide the preventive maintenance plan and emergency response during the operation period. The effect data of the decision implementation is collected in real time and fed back to the digital twin model to continuously verify and correct the prediction accuracy of the digital twin model.

[0024] In one embodiment, the present invention provides a digital inspection system for the quality of highway construction projects, further comprising:

[0025] The comprehensive loss calculation module, based on the obtained individual damage assessment or prediction results, introduces a multi-damage coupling analysis module to address the superimposed effects of damage of different locations and types due to their interactions. This multi-damage coupling analysis module quantifies the synergistic effects between damages (e.g., adjacent microcracks may merge into macro-cracks, or local yielding may lead to stress redistribution and exacerbate corrosion in other areas) through structural topological relationships and mechanical transfer models, thus assessing the overall severity of the damage superposition. Calculated using the following formula:

[0026] ;

[0027] in, This represents the severity index of the i-th injury. Let be the independent weighting coefficients for the i-th damage. Let i be the interaction coefficient between damage i and j. Based on the distance between damages The spatial decay function.

[0028] In one embodiment, the present invention provides a digital inspection system for the quality of highway construction projects, further comprising:

[0029] The interaction coefficient calibration module is used to build and continuously maintain a multi-damage interaction case database to support the continuous optimization of the multi-damage coupling analysis module. This database stores damage combination instances captured in each assessment, corresponding actual monitoring data, and subsequently observed structural responses (such as changes in performance degradation rate or repair effects). Based on this database, data-driven methods (such as statistical regression or machine learning) are periodically used for analysis to correlate the actual long-term impact of different damage combinations with the current interaction coefficients used by the multi-damage coupling analysis module. The predicted results are compared and verified, and then the interaction coefficients are calibrated and updated. .

[0030] In one embodiment, the present invention provides a digital inspection system for the quality of highway construction projects, further comprising:

[0031] The model fine-tuning module is used to continuously compare the deviation between the predicted mechanical data (such as predicted strain values) of the digital twin model and the actual monitoring data. When the deviation exceeds the threshold, the parameter calibration process is automatically triggered. The latest monitoring data is used to fine-tune the material constitutive parameters, boundary conditions or load assumptions in the digital twin model through optimization algorithms, so that the simulated behavior of the digital twin model is kept in sync with the behavior drift of the physical entity caused by material aging and changes in constraint conditions during long-term service.

[0032] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention systematically solves the problem of existing technologies that "emphasize short-term compliance while neglecting long-term performance". By collecting and processing performance data throughout the entire life cycle, it lays the data foundation for long-term evaluation; it constructs a digital twin model that integrates real-time data and integrates mechanical and material degradation algorithms, realizing dynamic simulation of the long-term performance evolution of structures under environmental and load effects; it incorporates machine learning algorithms to achieve real-time assessment of structural damage and prediction of future performance degradation trends, enabling potential risks to be identified and warned in the early stages of operation, thereby transforming quality control from passive acceptance to proactive performance assurance throughout the entire life cycle. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the first part of a digital inspection method for the quality of highway construction projects provided in an embodiment of the present invention.

[0034] Figure 2 This is a schematic diagram of the second part of a digital inspection method for the quality of highway construction projects provided in an embodiment of the present invention.

[0035] Figure 3 This is a schematic diagram of the third part of a digital inspection method for the quality of highway construction projects provided in an embodiment of the present invention.

[0036] Figure 4 This is a schematic diagram of the fourth part of a digital inspection method for the quality of highway construction projects provided in an embodiment of the present invention.

[0037] Figure 5 This is a schematic diagram of the fifth part of a digital inspection method for the quality of highway construction projects provided in an embodiment of the present invention.

[0038] Figure 6 This is a schematic diagram of the first part of a digital inspection system for the quality of highway construction projects provided in an embodiment of the present invention.

[0039] Figure 7 This is a schematic diagram of the second part of a digital inspection system for the quality of highway construction projects provided in an embodiment of the present invention.

[0040] Figure 8 This is a schematic diagram of the third part of a digital inspection system for the quality of highway construction projects provided in an embodiment of the present invention.

[0041] Figure 9 This is a schematic diagram of the fourth part of a digital inspection system for the quality of highway construction projects provided in an embodiment of the present invention.

[0042] Figure 10 This is a schematic diagram of the fifth part of a digital inspection system for the quality of highway construction projects provided in an embodiment of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] In one embodiment, such as Figure 1 As shown, a digital inspection method for the quality of highway construction projects includes the following steps:

[0045] Step S1: Collect full life cycle performance data (stress, strain, displacement, cracks, temperature and humidity, corrosion, etc.) of key parts (tunnel lining, bridge bearings, etc.) of engineering facilities (such as tunnels and bridges), and clean, align and feature engineering to obtain spatiotemporally consistent full life cycle performance data.

[0046] Step S2: Construct a BIM model (Building Information Model) corresponding to the real engineering facility, control the BIM model to integrate spatiotemporally consistent full life cycle performance data, and create a digital twin model that reflects the real engineering facility. The digital twin model can not only present the current state of the structure in three dimensions, but also dynamically simulate the long-term performance evolution process of the structure under real loads and environmental effects by integrating mechanical and material degradation algorithms.

[0047] Step S3: In the digital twin model, a machine learning algorithm is embedded to mine massive amounts of historical monitoring data and real-time streaming data. The machine learning algorithm establishes a quantitative mapping relationship between data features (such as strain combinations under specific modes and the rate of change of crack geometric parameters) and structural damage states (such as microcrack initiation, local yielding, and initial corrosion of steel bars) through training. Based on the real-time data input, it automatically identifies and quantifies the type, location, and severity of the current damage to obtain the assessment results. On this basis, the machine learning algorithm further combines the current damage state, continuously input environmental load data, and the embedded material time-varying degradation model to dynamically deduce the degradation trajectory of the structure's performance indicators (such as bearing capacity and stiffness) over a future set period (through recursive or iterative calculation). Based on the preset safety threshold or failure criteria, it calculates the probability and time of reaching the critical state and finally outputs the degradation trend and quantitative risk level of the structure's long-term performance to obtain the prediction results.

[0048] Steps S1 to S3 together constitute the core technological closed loop of digital quality inspection, aiming to systematically address the fundamental flaw of traditional methods that "emphasize short-term compliance while neglecting long-term performance." Step S1, through standardized data collection and preprocessing, lays a reliable and spatiotemporally unified data foundation for the entire process analysis, solving the problem of information silos. Step S2 uses this to construct a dynamic evolutionary digital twin model that integrates real-time data, upgrading the static BIM model into a real model capable of simulating long-term performance degradation, overcoming the inability of traditional design models to reflect changes during service life. Step S3 then implants an intelligent assessment and prediction kernel into the digital twin model, using machine learning to mine damage patterns from massive amounts of data and quantitatively predict future risks, realizing a paradigm shift in quality judgment from post-inspection to in-process early warning and pre-prediction.

[0049] In one embodiment, such as Figure 2 As shown, a digital inspection method for the quality of highway construction projects also includes:

[0050] Step S4: Establish an intelligent decision-making and closed-loop feedback mechanism. The obtained evaluation results and prediction results are automatically used to generate maintenance, repair or reinforcement decisions that include specific measures, priority ranking and implementation timing through a preset expert rule base and optimization algorithm. The decisions guide the preventive maintenance plan and emergency response during the operation period. The effect data of the decision implementation is collected in real time and fed back to the digital twin model to continuously verify and correct the prediction accuracy of the digital twin model.

[0051] Step S4 automatically transforms the complex prediction results from step S3 into actionable decisions by establishing a rule base, ensuring that risk warnings directly drive precise maintenance and improve response efficiency. Simultaneously, this mechanism feeds back the effect data after decision execution to the digital twin model, forming a closed loop of "evaluation-decision-verification-optimization." This allows the entire system to continuously learn and improve using actual operational data, enabling the accumulation and reuse of engineering wisdom.

[0052] This continuously accumulated and validated long-term performance data and knowledge can also be systematically fed back to the design and construction phases of engineering projects. This can be used to reverse-optimize the calculation parameters and safety factors in the design phase of similar future projects, the durability indicators and selection standards in the material phase, and the key process control indicators in the construction phase, thereby forming a closed loop that drives the continuous improvement of engineering quality.

[0053] In one embodiment, such as Figure 3 As shown, a digital inspection method for the quality of highway construction projects also includes:

[0054] Step S5: Based on the obtained individual damage assessment or prediction results, a multi-damage coupling analysis module is introduced to address the superposition effect caused by the interaction of damages of different locations and types. This module quantifies the synergistic effects between damages (e.g., adjacent microcracks may merge into macro-cracks, or local yielding may lead to stress redistribution and exacerbate corrosion in other areas) and the overall severity of damage superposition through structural topology and mechanical transfer models. Calculated using the following formula:

[0055] ;

[0056] in, This represents the severity index of the i-th injury (similarly, (Index representing the severity of the j-th injury). Let be the independent weighting coefficients for the i-th damage. Let i be the interaction coefficient between damage i and j. Based on the distance between damages The spatial decay function.

[0057] Step S5 is specifically designed to address the limitations of traditional testing methods that isolate localized damage. Multiple damages that are considered minor in individual assessments can potentially trigger catastrophic systemic risks due to spatial coupling and mechanical interactions. Therefore, this step introduces a multi-damage coupling analysis module to quantify the synergistic and cumulative effects between damages from a structurally holistic perspective. This ensures that risk assessment is no longer limited to a single point but rather reflects the overall safety status of the structure more realistically and comprehensively, avoiding misjudgments caused by neglecting the destructive power of damage combinations.

[0058] In one embodiment, such as Figure 4 As shown, a digital inspection method for the quality of highway construction projects also includes:

[0059] Step S6: Construct and continuously maintain a multi-damage interaction case database to support the continuous optimization of the multi-damage coupling analysis module. This database stores damage combination instances captured in each assessment, corresponding actual monitoring data, and subsequently observed structural responses (such as changes in performance degradation rate or repair effects). Based on this database, data-driven methods (such as statistical regression or machine learning) are periodically used for analysis to correlate the actual long-term impact of different damage combinations with the current interaction coefficients used by the multi-damage coupling analysis module. The predicted results are compared and verified, and then the interaction coefficients are calibrated and updated. .

[0060] Step S6 is to resolve the interaction coefficients upon which the multi-damage coupling analysis in step S5 depends. The lack of real-world data calibration is a problem. By building a continuously growing case database and systematically accumulating actual cases of damage interactions and real structural responses, the coupling analysis model can move from theoretical assumptions to being driven by real data. By continuously comparing predictions with actual results, the interaction coefficients can be dynamically calibrated. This ensures that the analytical conclusions continuously approach engineering reality as experience accumulates.

[0061] In one embodiment, such as Figure 5 As shown, a digital inspection method for the quality of highway construction projects also includes:

[0062] Step S7: Continuously compare the deviation between the predicted mechanical data (such as the predicted strain value) of the digital twin model and the actual monitoring data. When the deviation exceeds the threshold, the parameter calibration process is automatically triggered. The latest monitoring data is used to fine-tune the material constitutive parameters, boundary conditions or load assumptions in the digital twin model through optimization algorithms, so that the simulation behavior of the digital twin model is synchronized with the behavior drift of the physical entity due to material aging and changes in constraint conditions during long-term service.

[0063] Step S7 aims to address the "model drift" problem that inevitably occurs in digital twin models during long-term service. Even with the most accurate initial model, aging of engineering materials and minor changes in boundary conditions can cause the behavior of the physical entity to gradually deviate from the model's original predictions. Step S7 establishes an automatic comparison mechanism between monitoring data and model prediction data, along with a reverse parameter tuning mechanism, enabling the digital twin model to self-update. It can sensitively detect this deviation and automatically fine-tune its intrinsic parameters, ensuring that the model remains synchronized with the constantly changing physical entity, thereby maintaining its long-term accuracy and authority as the basis for prediction and decision-making.

[0064] In one embodiment, such as Figure 6 As shown, a digital inspection system for the quality of highway construction projects includes:

[0065] Data acquisition module 1 is used to collect full life cycle performance data (stress, strain, displacement, cracks, temperature and humidity, corrosion, etc.) of key parts (tunnel lining, bridge bearings, etc.) of engineering facilities (such as tunnels and bridges) and perform cleaning, alignment and feature engineering to obtain spatiotemporally consistent full life cycle performance data.

[0066] Model building module 2 is used to build a BIM model corresponding to the real engineering facility, control the integration of spatiotemporally consistent full life cycle performance data of the BIM model, and create a digital twin model that reflects the real engineering facility. The digital twin model can not only present the current state of the structure in three dimensions, but also dynamically simulate the long-term performance evolution process of the structure under real loads and environmental effects by integrating mechanical and material degradation algorithms.

[0067] Damage assessment and prediction module 3 is used to embed machine learning algorithms into the digital twin model to mine massive amounts of historical monitoring data and real-time streaming data. The machine learning algorithm establishes a quantitative mapping relationship between data features (such as strain combinations under specific modes and the rate of change of crack geometric parameters) and structural damage states (such as microcrack initiation, local yielding, and initial corrosion of steel bars) through training. Based on real-time data input, it automatically identifies and quantifies the type, location, and severity of current damage to obtain assessment results. On this basis, the machine learning algorithm further combines the current damage state, continuously input environmental load data, and the embedded material time-varying degradation model to dynamically deduce the degradation trajectory of the structure's performance indicators (such as bearing capacity and stiffness) within F (through recursive or iterative calculation). Based on preset safety thresholds or failure criteria, it calculates the probability and time of reaching the critical state and finally outputs the degradation trend and quantitative risk level of the structure's long-term performance to obtain prediction results.

[0068] Taking a tunnel as an example, after the tunnel is completed, fiber optic strain sensors, crack gauges, piezometers, and temperature and humidity sensors are installed at key locations such as the lining arch, arch waist, and sidewalls. Data acquisition module 1 continuously collects data 24 hours a day through these data acquisition devices. The raw data is cleaned by the platform to remove outliers caused by temporary equipment failures, and the timestamps of all sensors are aligned with spatial coordinates to finally generate a spatiotemporally consistent tunnel lifecycle performance data.

[0069] Based on the tunnel's design BIM model, a high-fidelity digital twin model is created by integrating the tunnel's full life-cycle performance data with the geological and hydrological information of the tunnel's location. This digital twin model can not only display the tunnel's current deformation and crack distribution in three dimensions, but also simulate the stress redistribution and performance degradation of the lining structure under the combined effects of vehicle cyclic loads, mountain bias pressure, and groundwater erosion through built-in rock mechanics and concrete creep algorithms.

[0070] Machine learning algorithms in a digital twin model analyzed historical data and found that when the strain at the arch waist increases slightly and is accompanied by vibrations of a specific spectrum, it often indicates that the local concrete is entering the fatigue stage. One day, the system matched this pattern to real-time data streams and automatically assessed that "micro-crack initiation" had occurred at the arch waist, with a damage level of mild. Simultaneously, the algorithm, considering the current high humidity and traffic flow inside the tunnel, predicted that without intervention, the crack would propagate within the next 8 years and could potentially lead to lining collapse.

[0071] In one embodiment, such as Figure 7 As shown, a digital inspection system for the quality of highway construction projects also includes:

[0072] The intelligent decision feedback module 4 is used to establish an intelligent decision-making and closed-loop feedback mechanism. It automatically generates maintenance, repair or reinforcement decisions, including specific measures, priority ranking and implementation timing, based on the obtained evaluation results and prediction results through a preset expert rule base and optimization algorithm. The decisions guide the preventive maintenance plan and emergency response during the operation period. The effect data of the decision implementation is collected in real time and fed back to the digital twin model to continuously verify and correct the prediction accuracy of the digital twin model.

[0073] Based on the early warning from the damage assessment and prediction module 3, the decision-making mechanism automatically generated a recommendation: "Within the next three months, conduct non-destructive testing on the arch section (the specific location is known from the data acquisition equipment) for confirmation, and include it in the next quarter's preventative maintenance plan for surface sealing treatment." After the maintenance team implemented the sealing treatment, the strain data at that location stabilized. This maintenance effectiveness data was fed back to the system, confirming the effectiveness of the prediction and decision-making, and reinforcing the handling rules for this type of damage pattern.

[0074] In one embodiment, such as Figure 8 As shown, a digital inspection system for the quality of highway construction projects also includes:

[0075] The comprehensive loss calculation module 5 is used to introduce a multi-damage coupling analysis module based on the obtained single damage assessment or prediction results. This module addresses the superposition effect caused by the interaction of damages of different locations and types. The multi-damage coupling analysis module quantifies the synergistic effects between damages (e.g., adjacent microcracks may merge into macro-cracks, or local yielding may lead to stress redistribution and exacerbate corrosion in other areas) and the overall severity of damage superposition through structural topology and mechanical transfer models. Calculated using the following formula:

[0076] ;

[0077] in, This represents the severity index of the i-th injury. Let be the independent weighting coefficients for the i-th damage. Let i be the interaction coefficient between damage i and j. Based on the distance between damages The spatial decay function.

[0078] Taking a section of the tunnel as an example, the system's independent assessment revealed a longitudinal crack (damage) 2 meters long and 0.3 millimeters wide in the arch. (Severity index: 0.4); localized water seepage in the lining on the left arch waist led to concrete saturation (damage). The severity index is 0.3. If assessed individually, both would be considered mild. The multi-damage coupling analysis module was activated, and based on the structural topology, it was determined that the two damages were only 3 meters apart. =3), and the cracks provide channels for water seepage, which in turn exacerbates the freeze-thaw or dissolution effects at the crack tips. Therefore, an interaction coefficient is set. The value is 0.8. Substitute this value into the formula to calculate the overall damage degree. The independent weights are... , All are set to 1, spatial decay function The value is 0.9. The calculation yields... = 1*0.4+1*0.3+0.8*(0.4*0.3)*0.9 = 0.4 + 0.3 + 0.0864 = 0.7864. This value is significantly higher than any independent damage value. If the system's set moderate risk threshold is 0.6, the system will raise the overall risk level of this section and issue a warning that the combined effect of cracks and water seepage may accelerate lining spalling.

[0079] In one embodiment, such as Figure 9As shown, a digital inspection system for the quality of highway construction projects also includes:

[0080] The interaction coefficient calibration module 6 is used to build and continuously maintain a multi-damage interaction case database to support the continuous optimization of the multi-damage coupling analysis module. This database stores damage combination instances captured in each assessment, corresponding actual monitoring data, and subsequently observed structural responses (such as changes in performance degradation rate or repair effects). Based on this database, data-driven methods (such as statistical regression or machine learning) are used periodically to analyze the actual long-term effects of different damage combinations and the current interaction coefficients used by the multi-damage coupling analysis module. The predicted results are compared and verified, and then the interaction coefficients are calibrated and updated. .

[0081] The aforementioned case of coupled cracks in the arch crown and water seepage in the arch waist was fully recorded in the database, including initial damage characteristics, environmental data (temperature, humidity), and monitoring data for the following year (showing crack width expanding at a rate of 0.05 mm per month, far exceeding the typical rate for a single crack). One year later, the system underwent periodic coefficient calibration. It retrieved all similar "crack-seepage" coupled cases (15 groups in total) from the database, using the actual observed performance degradation rates (such as crack propagation rate and concrete strength loss rate) as the true values, and applied the current default interaction coefficients. The degradation rate predicted by the digital twin model (with a coefficient of 0.8) was used as the predicted value for regression comparison. Analysis revealed that current digital twin models generally underestimate the degradation rate, with prediction errors reaching 20%. Therefore, the system automatically incorporates the interaction coefficients of such damage combinations using an optimization algorithm that minimizes the prediction error. The calibration was updated from 0.8 to 0.95. From this point forward, the digital twin model will use more accurate coefficients when performing risk assessments on new cases containing such combinations of damage using the multi-damage coupling analysis module. Using a value of 0.95, the prediction results will be closer to the actual risk.

[0082] In one embodiment, such as Figure 10 As shown, a digital inspection system for the quality of highway construction projects also includes:

[0083] Model fine-tuning module 7 is used to continuously compare the deviation between the predicted mechanical data (such as the predicted strain value) of the digital twin model and the actual monitoring data. When the deviation exceeds the threshold, the parameter calibration process is automatically triggered. The latest monitoring data is used to fine-tune the material constitutive parameters, boundary conditions or load assumptions in the digital twin model through optimization algorithms, so that the simulation behavior of the digital twin model is kept in sync with the behavior drift of the physical entity caused by material aging and changes in constraint conditions during long-term service.

[0084] Five years after the tunnel began operation, continuous comparisons revealed that the actual deformation of the lining in many areas was generally slightly higher than the initial predictions of the digital twin model. This indicated that the actual soil and rock creep effect was stronger than the original design assumptions. Consequently, the parameter calibration process was automatically triggered, fine-tuning the surrounding rock pressure parameters and the concrete creep coefficient in reverse. This resynchronized the simulated behavior of the digital twin model with the actual wear and tear of the tunnel, ensuring the reliability of subsequent long-term predictions.

[0085] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0086] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0087] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0088] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0089] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for digital detection of highway construction engineering quality, characterized in that, The digital inspection method for the quality of highway construction projects includes the following steps: Collect full life cycle performance data of key components of engineering facilities and perform cleaning, alignment and feature engineering to obtain spatiotemporally consistent full life cycle performance data; Construct a BIM model corresponding to the real engineering facility, control the integration of spatiotemporally consistent full life cycle performance data of the BIM model, and create a digital twin model that reflects the real engineering facility. The digital twin model can not only present the current state of the structure in three dimensions, but also dynamically simulate the long-term performance evolution process of the structure under real loads and environmental effects by integrating mechanical and material degradation algorithms. In the digital twin model, machine learning algorithms are embedded to mine massive amounts of historical monitoring data and real-time streaming data. The machine learning algorithms establish a quantitative mapping relationship between data features and structural damage states through training, thereby automatically identifying and quantifying the type, location, and severity of current damage based on real-time data input, and obtaining assessment results. On this basis, the machine learning algorithms further combine the current damage state, continuously input environmental load data, and embedded material time-varying degradation models to dynamically predict the degradation trajectory of the structure's performance indicators over a future set period of time, and calculate the probability and time of reaching the critical state based on preset safety thresholds or failure criteria, and finally output the degradation trend and quantitative risk level of the structure's long-term performance, obtaining prediction results. On the basis of obtaining single damage assessment results or prediction results, a multi-damage coupling analysis module is introduced to solve the superposition effect of different positions and different types of damages due to interaction, the multi-damage coupling analysis module quantifies the synergistic effect between damages through a structure topological relationship and a mechanical transmission model, and comprehensively determines the severity of damage superposition is calculated from the following equation: ； wherein, represents the severity index of the ith lesion, is an independent weight coefficient for the ith lesion, is an interaction coefficient for lesions i and j, is a spatial decay function based on the distance between lesions ; A multi-damage interaction case database is constructed and continuously maintained to support the continuous optimization of the multi-damage coupling analysis module, the database stores the damage combination instances captured in each assessment, the corresponding actual monitoring data, and the subsequent observed structural responses, based on the database, the actual long-term impact of different damage combinations is compared and verified with the prediction results of the current multi-damage coupling analysis module using the current interaction coefficient , and then the interaction coefficient is calibrated and updated.

2. The expressway construction engineering quality digital detection method according to claim 1, characterized in that, Also includes: An intelligent decision-making and closed-loop feedback mechanism is established. The obtained evaluation results and prediction results are automatically generated into maintenance, repair or reinforcement decisions, including specific measures, priority ranking and implementation timing, through a preset expert rule base and optimization algorithm. The decisions guide the preventive maintenance plan and emergency response during the operation period. The effect data of the decision implementation is collected in real time and fed back to the digital twin model to continuously verify and correct the prediction accuracy of the digital twin model.

3. The highway construction engineering quality digital detection method according to claim 1 or 2, characterized in that, Also includes: The deviation between the predicted mechanical data of the digital twin model and the actual monitoring data is continuously compared. When the deviation exceeds the threshold, the parameter calibration process is automatically triggered. The latest monitoring data is used to fine-tune the material constitutive parameters, boundary conditions or load assumptions in the digital twin model through optimization algorithms, so that the simulated behavior of the digital twin model is synchronized with the behavior drift of the physical entity caused by material aging and changes in constraint conditions during long-term service.

4. A highway construction engineering quality digital detection system, characterized in that, include: The data acquisition module is used to collect full life cycle performance data of key parts of engineering facilities and perform cleaning, alignment and feature engineering to obtain spatiotemporally consistent full life cycle performance data. The model building module is used to build a BIM model corresponding to the real engineering facility, control the integration of spatiotemporally consistent full life cycle performance data of the BIM model, and create a digital twin model that reflects the real engineering facility. The digital twin model can not only present the current state of the structure in three dimensions, but also dynamically simulate the long-term performance evolution process of the structure under real loads and environmental effects by integrating mechanics and material degradation algorithms. The damage assessment and prediction module is used to embed machine learning algorithms into the digital twin model to mine massive amounts of historical monitoring data and real-time streaming data. The machine learning algorithm establishes a quantitative mapping relationship between data features and structural damage state through training, thereby automatically identifying and quantifying the type, location, and severity of current damage based on real-time data input to obtain assessment results. On this basis, the machine learning algorithm further combines the current damage state, continuously input environmental load data, and embedded material time-varying degradation model to dynamically deduce the degradation trajectory of the structure's performance indicators within a set future period, and calculates the probability and time of reaching the critical state based on preset safety thresholds or failure criteria, finally outputting the degradation trend and quantitative risk level of the structure's long-term performance to obtain prediction results. The comprehensive loss calculation module is used to introduce a multi-damage coupling analysis module on the basis of obtaining a single damage assessment result or a prediction result, solve superposition effects caused by interaction of different positions and different types of damages, and quantify cooperative influences between damages and comprehensive severity of damage superposition through a structure topological relationship and a mechanical transmission model. This is calculated by the following formula: ； wherein, represents the severity index of the ith lesion, is an independent weight coefficient for the ith lesion, is an interaction coefficient for lesions i and j, is a spatial decay function based on the distance between lesions ; An interaction coefficient calibration module for building and continuously maintaining a multi-damage interaction case database to support continuous optimization of the multi-damage coupling analysis module, the database stores instances of damage combinations captured in each assessment, corresponding actual monitoring data, and subsequently observed structural responses, based on the database, periodically using data-driven methods for analysis, comparing actual long-term effects of different damage combinations with predictions of the current multi-damage coupling analysis module using the current interaction coefficients , and thereby calibrating and updating the interaction coefficients .

5. The highway construction engineering quality digital detection system according to claim 4, characterized in that, Also includes: The intelligent decision feedback module is used to establish an intelligent decision-making and closed-loop feedback mechanism. It automatically generates maintenance, repair or reinforcement decisions, including specific measures, priority ranking and implementation timing, based on the obtained evaluation results and prediction results through a preset expert rule base and optimization algorithm. The decisions guide the preventive maintenance plan and emergency response during the operation period. The effect data of the decision implementation is collected in real time and fed back to the digital twin model to continuously verify and correct the prediction accuracy of the digital twin model.

6. The highway construction engineering quality digital detection system according to claim 4 or 5, characterized in that, Also includes: The model fine-tuning module is used to continuously compare the deviation between the predicted mechanical data of the digital twin model and the actual monitoring data. When the deviation exceeds the threshold, the parameter calibration process is automatically triggered. The latest monitoring data is used to fine-tune the material constitutive parameters, boundary conditions or load assumptions in the digital twin model through optimization algorithms, so that the simulated behavior of the digital twin model is kept in sync with the behavior drift of the physical entity caused by material aging and changes in constraint conditions during long-term service.

Images