Durability detection system of double-mixed fiber concrete for cross-river and cross-sea shield tunnel

Abstract

The present application relates to the technical field of concrete detection, and discloses a durability detection system for double-mixed fiber concrete of a river-crossing and sea-crossing shield tunnel. The durability data acquisition unit of the system acquires multi-source monitoring data under the service environment of segment concrete, including material performance, environmental erosion and structural response parameters; the feature fusion processing unit integrates multi-source data according to time sequence, extracts performance degradation features and constructs a multi-dimensional durability feature space; the damage mode recognition unit establishes a damage state classification model based on the feature space, and identifies the current damage development stage; the critical state prediction unit calculates the performance degradation rate according to the damage change trajectory, and predicts the time node of reaching the critical state; and the detection scheme generation unit automatically matches the optimal detection scheme according to the prediction result, and outputs an execution instruction including the detection position, method and cycle. The system can comprehensively grasp the durability state of concrete, and improve the scientificity and effectiveness of detection.

Description

Technical Field

[0001] This invention relates to the field of concrete testing technology, specifically a durability testing system for double-fiber-blended concrete in cross-river and cross-sea shield tunnels. Background Technology

[0002] As a crucial component of cross-water transportation infrastructure, shield tunnels operating in environments characterized by high humidity, high salinity, and frequent temperature fluctuations are particularly challenging. The concrete segments of these tunnels are constantly exposed to this complex and corrosive environment, making them susceptible to durability damage such as chloride ion penetration, steel reinforcement corrosion, and the propagation of superficial cracks. This ultimately impacts the overall safety and service life of the tunnel structure. While fiber-reinforced concrete, with its excellent crack resistance and impermeability, is widely used in the construction of shield tunnel segments for cross-river and cross-sea tunnels, even this type of concrete experiences gradual degradation in durability due to environmental erosion over long-term service. Failure to accurately and promptly assess the concrete's durability can lead to overlooked structural damage, ultimately resulting in safety accidents or increased maintenance costs.

[0003] The current methods for testing the durability of concrete in shield tunnels primarily rely on periodic manual inspections. Inspectors must enter the tunnel and use handheld devices to sample and inspect the surface condition of the concrete segments and the extent of steel reinforcement corrosion. This method has significant limitations. First, manual sampling cannot achieve comprehensive coverage of tunnel segments, easily leading to blind spots and causing some areas with localized deterioration to go undetected. Second, manual inspection depends on the operator's experience and expertise, making the results susceptible to human error and compromising data accuracy and consistency. Furthermore, traditional methods cannot achieve real-time acquisition and continuous monitoring of concrete durability data, only obtaining discrete data at a specific point in time. This fails to reflect the dynamic degradation process of concrete performance over time and cannot provide comprehensive data support for subsequent damage prediction and maintenance decisions.

[0004] While some existing detection systems attempt to incorporate sensors for data acquisition, most can only process single types of monitoring data, such as collecting only chloride ion concentration data within concrete or monitoring only steel corrosion potential data. This fails to effectively integrate multi-source monitoring data. Since concrete durability degradation is the result of multiple factors, including material properties, environmental erosion, and structural response, single-type data cannot comprehensively reflect the actual durability state of concrete. This leads to low accuracy in damage assessment models built upon this data, making it difficult to accurately identify the damage development stage of concrete and to effectively predict the timeline when concrete reaches its critical durability state. Furthermore, existing systems often employ fixed detection patterns, failing to dynamically adjust detection locations, methods, and cycles based on the actual durability state and degradation trend of the concrete. This results in unreasonable allocation of detection resources, with some areas being over-tested while critical areas are under-tested, impacting the efficiency and effectiveness of the detection work. Summary of the Invention

[0005] The purpose of this invention is to provide a durability testing system for double-fiber-coated concrete in cross-river and cross-sea shield tunnels, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a durability testing system for double-fiber-coated concrete in cross-river and cross-sea shield tunnels, the system comprising:

[0007] The durability data acquisition unit is used to acquire multi-source monitoring data of the concrete of the shield tunnel segment under service environment. The multi-source monitoring data includes material performance parameters, environmental erosion parameters and structural response parameters.

[0008] The feature fusion processing unit is used to align and integrate multi-source monitoring data according to time series, extract the performance degradation characteristics of concrete at different erosion stages, and construct a multi-dimensional durability feature space.

[0009] The damage pattern recognition unit establishes a damage state classification model based on a multidimensional durability feature space to identify the current damage development stage of the concrete.

[0010] The critical state prediction unit calculates the performance degradation rate based on the change trajectory of the damage development stage and predicts the time node when concrete reaches the critical durability state.

[0011] The detection scheme generation unit automatically matches the optimal detection scheme based on the critical state prediction results and outputs execution instructions including detection location, detection method and detection cycle.

[0012] Preferably, the durability data acquisition unit acquires multi-source monitoring data in the following specific manner:

[0013] Data on chloride ion concentration distribution inside concrete and steel corrosion potential are collected using an embedded sensor array.

[0014] The morphological characteristics and distribution density parameters of apparent cracks in concrete were obtained using surface scanning equipment.

[0015] Environmental monitoring devices were used to record temperature and humidity variation curves and contact frequency data of corrosive media inside the tunnel.

[0016] Preferably, the specific steps for the feature fusion processing unit to construct the multidimensional durability feature space are as follows:

[0017] Chloride ion concentration distribution data and steel corrosion potential data are matched by gridding according to spatial coordinates to generate a material performance degradation characteristic matrix.

[0018] Topological structure analysis was performed on the morphological characteristic parameters of the apparent cracks to extract the crack network connectivity and propagation direction characteristics.

[0019] A time-domain correlation analysis was performed between the temperature and humidity change curves and the contact frequency data of the corrosive medium to establish a characteristic vector of environmental erosion intensity.

[0020] The material performance degradation feature matrix, crack network features, and environmental erosion intensity feature vector are mapped to a unified time coordinate system to form a durability feature space that includes spatial and temporal dimensions.

[0021] Preferably, the specific method for the damage pattern recognition unit to establish a damage state classification model is as follows:

[0022] Four damage development stages are divided in the durability feature space, corresponding to the initial stabilization period, local degradation period, accelerated degradation period and critical failure period, respectively.

[0023] Density clustering algorithm is used to identify clustered regions of data points in the feature space, and typical patterns of each damage stage are determined based on the combination of feature parameters of the center point of the clustered region.

[0024] A discrimination function based on Mahalanobis distance is established to calculate the distance between the current monitoring data and each typical pattern in real time, and the damage stage corresponding to the minimum distance value is taken as the identification result.

[0025] Preferably, the specific process by which the critical state prediction unit calculates the performance degradation rate is as follows:

[0026] Extract the trajectory of characteristic parameter changes corresponding to the current stage of damage development, and calculate the local rate of change of each characteristic parameter using a sliding time window;

[0027] Establish a mapping relationship between the rate of change of characteristic parameters and the remaining service time, and obtain a comprehensive degradation rate index through weighted fusion;

[0028] The predicted remaining time for the current stage is adjusted based on the duration of development of the same damage stage in historical data.

[0029] Preferably, the specific logic of the detection scheme generation unit matching the optimal detection scheme is as follows:

[0030] When the predicted remaining time is greater than a preset threshold, a conventional inspection scheme based on non-destructive testing is generated, and the inspection cycle is set to a fixed proportion of the predicted remaining time.

[0031] When the predicted remaining time is less than a preset threshold, a special testing plan including core sampling and laboratory analysis is generated, shortening the testing cycle to half that of the conventional plan.

[0032] For tunnel segments in the accelerated degradation phase, crack propagation monitoring and steel corrosion rate monitoring should be added to the testing plan.

[0033] Preferably, the system further includes a data quality verification unit, the workflow of which is as follows:

[0034] Real-time monitoring of the integrity and consistency of data collected by each sensor, and spatial interpolation of missing data using data from adjacent sensors;

[0035] When the difference in monitoring data from different sensors at the same location exceeds the allowable range, a redundant measurement mechanism is activated to verify the data.

[0036] Abnormal data points are marked and their credibility weights are calculated. During the feature fusion processing stage, the proportion of data participating in the calculation is adjusted according to the weight values.

[0037] Preferably, the data quality verification unit further includes a drift compensation module, the compensation method of which is as follows:

[0038] Regularly collect benchmark test data of standard concrete test blocks and establish a deviation curve between sensor measurements and benchmark values;

[0039] The drift compensation coefficient of each sensor is calculated based on the trend of the deviation curve, and the original monitoring data is automatically corrected during the data processing stage.

[0040] When the drift compensation coefficient exceeds the critical threshold, a sensor calibration reminder signal is generated and the data weight of the sensor is adjusted.

[0041] Preferably, the system further includes a detection result feedback unit, the operation mechanism of which is as follows:

[0042] The on-site detection results are compared and analyzed with the system prediction results to calculate the prediction accuracy index.

[0043] The discrimination threshold and weight parameters of the critical state prediction unit of the damage state classification model are dynamically adjusted based on the accuracy index.

[0044] When the deviation of three consecutive detection results exceeds the allowable range, the model retraining process is triggered and the feature space partitioning criteria are updated.

[0045] Preferably, the detection result feedback unit further includes a scheme optimization module, the optimization strategy of which is:

[0046] Record the actual implementation effect of each testing plan and establish an evaluation index system for testing efficiency and accuracy;

[0047] If the improvement in diagnostic accuracy by a newly added test item is less than the preset value, the test item will be removed from the subsequent plan.

[0048] Optimize the spatial allocation scheme of testing resources based on the importance level and accessibility of the pipeline area.

[0049] Compared with the prior art, the beneficial effects of the present invention are:

[0050] This durability testing system for cross-river and cross-sea shield tunnels using double-fiber-coated concrete, through the establishment of durability data acquisition units, can comprehensively acquire multi-source monitoring data of shield tunnel segment concrete under service conditions. This data covers material performance parameters, environmental erosion parameters, and structural response parameters, breaking through the limitations of traditional testing methods that can only acquire single-type or discrete data. It achieves comprehensive and multi-dimensional acquisition of concrete durability-related data, allowing staff to understand the initial state and changes of concrete from multiple perspectives, providing a rich and comprehensive data foundation for subsequent data analysis and condition assessment.

[0051] The feature fusion processing unit aligns and integrates multi-source monitoring data according to time series, extracts performance degradation features of concrete at different erosion stages, and constructs a multi-dimensional durability feature space. This effectively solves the problem of scattered multi-source data and difficulty in collaborative analysis in existing detection systems. Time series alignment ensures the consistency of data from different sources in the time dimension, avoiding analysis errors caused by data time misalignment. The extraction of performance degradation features and the construction of the multi-dimensional feature space can transform scattered and seemingly unrelated data into feature information with clear physical meaning, intuitively presenting the performance change law of concrete at different erosion stages. This allows staff to clearly grasp the dynamic process of concrete from its initial state to gradual degradation, providing a structured and systematic feature basis for subsequent damage pattern identification.

[0052] The damage pattern recognition unit establishes a damage state classification model based on a multidimensional durability feature space, which can accurately identify the current damage development stage of concrete. Compared with traditional manual judgment or single-data judgment methods, this classification model makes full use of the rich information in the multidimensional feature space. Through comprehensive analysis of feature parameters, it can accurately distinguish different damage stages of concrete, avoiding misjudgments or omissions caused by incomplete information. This allows staff to grasp the degree of concrete damage in a timely and accurate manner, providing a reliable status basis for subsequent targeted treatment measures.

[0053] The critical state prediction unit calculates the performance degradation rate based on the trajectory of damage development stages, predicting the time point when concrete reaches its critical durability state. This changes the traditional detection method's passive approach to detecting damage, enabling proactive prediction of concrete durability. By calculating the performance degradation rate and predicting critical time points, staff can understand in advance when concrete may reach a critical state requiring emergency treatment, allowing sufficient time to develop response plans and avoid structural safety risks caused by sudden damage. It also provides a scientific reference for scheduling maintenance work, facilitating the rational planning of maintenance cycles and resource allocation.

[0054] The detection scheme generation unit automatically matches the optimal detection scheme based on the critical state prediction results, outputting execution instructions including detection location, detection method, and detection cycle, thus realizing personalized and dynamic adjustment of the detection scheme. This unit can specifically determine the locations requiring key detection based on the actual durability state of the concrete and the critical state prediction results, select the most suitable detection method for the current state, and rationally set the detection cycle, avoiding the problems of wasted detection resources or insufficient detection in traditional fixed detection modes. By optimizing the detection location, key damaged areas can be ensured to receive focused attention; by selecting appropriate detection methods, the accuracy of detection data and the efficiency of detection work can be improved; by adjusting the detection cycle, unnecessary detection times can be reduced while ensuring timely understanding of the concrete state, lowering detection costs, and comprehensively improving the scientific nature and effectiveness of the detection work, providing strong support for the long-term stable service of double-fiber-modified concrete structures in cross-river and cross-sea shield tunnels. Attached Figure Description

[0055] Figure 1 This is a timing diagram of the durability testing system for double-fiber-blended concrete in cross-river and cross-sea shield tunnels as described in this invention.

[0056] Figure 2 A flowchart for acquiring multi-source monitoring data for the durability data acquisition unit;

[0057] Figure 3 A flowchart for establishing a damage state classification model for the damage pattern recognition unit. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please see Figure 1 This invention provides a durability testing system for double-fiber-coated concrete in cross-river and cross-sea shield tunnels. The system includes dynamic assessment and prediction of concrete durability through multi-source data fusion and machine learning algorithms. The core of the system consists of a durability data acquisition unit, a feature fusion processing unit, a damage pattern recognition unit, a critical state prediction unit, and a detection scheme generation unit, all connected sequentially to form a closed-loop processing flow. The durability data acquisition unit is responsible for collecting material performance parameters, environmental erosion parameters, and structural response parameters in real time from the service environment of the shield tunnel segment concrete. These parameters include, but are not limited to, chloride ion concentration, rebar potential, crack morphology, and temperature and humidity data. The feature fusion processing unit aligns and integrates the collected multi-source monitoring data according to time series, extracts the performance degradation characteristics of concrete at different erosion stages, and constructs a multi-dimensional durability feature space containing spatiotemporal dimensions. The damage pattern recognition unit uses this feature space to establish a damage state classification model, identifying the current damage development stage of the concrete, such as the initial stable period or the accelerated degradation period. The critical state prediction unit calculates the performance degradation rate based on the change trajectory of the damage development stage and predicts the time node when the concrete reaches the critical durability state. The detection scheme generation unit automatically matches the optimal detection scheme based on the prediction results and outputs execution instructions including detection location, detection method and detection cycle, thereby realizing fully automated processing from data acquisition to decision output.

[0060] Example 1: See Figure 2The implementation of the durability data acquisition unit relies on the deployment of a multi-layered, multi-type sensor network, which is pre-planned and embedded in key parts of the shield tunnel segment concrete. Embedded sensor arrays are integrated into the concrete during the segment prefabrication stage. Their placement strategy is determined based on fluid dynamics simulations and past erosion experience, focusing on vulnerable areas such as segment joints, stress concentration areas, and the backwater surface. The chloride ion sensor in the array uses the Ag / AgCl electrode principle to continuously monitor the concentration gradient of free chloride ions in the concrete pore fluid. The data is stored in the form of a two-dimensional grid coordinate system, with each grid point associated with depth information. The rebar corrosion potential sensor is connected to the rebar network through a pre-embedded reference electrode to measure changes in the rebar half-cell potential. Its data has a one-to-one spatial correspondence with chloride ion concentration data, and the acquisition timestamps are strictly synchronized to ensure data consistency. Surface scanning equipment includes a high-resolution linear array camera and a laser scanner, which periodically move on a pre-set track within the tunnel to acquire panoramic images of the inner wall of the segments.

[0061] After preprocessing, the image data is used to identify and quantify apparent cracks using computer vision algorithms. Morphological parameters include the maximum width, average width, total length, and tortuosity of the cracks, while the distribution density parameter is obtained by calculating the total crack length per unit area. The environmental monitoring device consists of temperature and humidity sensors and chemical medium sensors. The temperature and humidity sensors are installed on the surface of the tunnel segments and at different depths inside, recording dynamic changes in temperature and relative humidity. Their sampling frequency is configured according to the differences between the internal and external environments of the tunnel. The chemical medium sensor is used to detect the contact frequency of corrosive media, such as monitoring carbon dioxide concentration using an electrochemical sensor or monitoring sulfate ion concentration using an ion-selective electrode. The contact frequency is quantified by counting the number of times the medium concentration exceeds a threshold per unit time. All these multi-source monitoring data are aggregated to the central processor of the data acquisition unit via wired or wireless transmission networks. The data packets include precise timestamps and spatial location tags to prepare for subsequent time-series alignment.

[0062] The core task of the feature fusion processing unit is to integrate monitoring data from different sources and in different formats into a unified analysis framework with spatiotemporal dimensions. Data processing begins with time series alignment. Due to potential differences in sampling periods and start-up times among various sensors, the system uses interpolation algorithms to resample all data sequences onto a unified time axis, for example, aligning all data to one data point per minute. Spatial alignment relies on the precise coordinate mapping relationship established during the data point deployment phase, mapping the physical location of each sensor to a virtual three-dimensional mesh model. Based on spatiotemporal alignment, for material performance parameters, chloride ion concentration distribution data and steel corrosion potential data are meshed and matched according to the same spatial coordinates. Within each mesh cell, the chloride ion concentration value and the steel corrosion potential value are combined into a data pair. All data pairs from all mesh cells together constitute a feature matrix reflecting material performance degradation. This matrix not only contains numerical information but also implicitly contains the spatial correlation between ion migration and electrochemical corrosion. For apparent crack data, topological analysis is a key step. The algorithm treats the identified cracks as a network graph, where the intersections and endpoints of cracks constitute nodes, and crack segments constitute edges. By calculating parameters such as the number of connected components, node degree distribution, and average length of the crack segment, connectivity features describing the overall structure of the crack network are extracted.

[0063] By comparing crack images at consecutive time points, the algorithm tracks the movement trajectory of each crack tip, thereby extracting the crack propagation direction features, which can be represented by a principal direction vector. For environmental erosion parameters, the temporal correlation analysis of temperature and humidity variation curves and contact frequency data of the eroding medium uses a sliding window method to calculate the cross-correlation function, seeking the temporal coupling relationship between drastic temperature and humidity fluctuations and peak concentrations of the eroding medium, thus establishing a feature vector that comprehensively characterizes the intensity of environmental erosion. This vector can identify the period of most severe environmental erosion. Finally, the material performance degradation feature matrix, crack network features, and environmental erosion intensity feature vector are jointly mapped into a unified multidimensional durability feature space. This feature space can be understood as a high-dimensional database, where each data point represents the comprehensive state of concrete at a specific time and spatial location. The dimensions include physicochemical parameters, structural morphology parameters, and environmental history parameters, with the time dimension serving as the main thread running through all parameters, allowing the complete tracing and quantitative analysis of the concrete performance degradation process. The multidimensional durability feature space output by the feature fusion processing unit is not a simple dataset, but an ordered structure capable of depicting the entire process of concrete from microscopic material degradation to macroscopic structural response. The data flow process is automated, but each processing step includes a data quality verification step. For example, after spatiotemporal alignment, the reasonableness of the interpolated data is judged, and after feature extraction, the significance of the features is evaluated. These internal verification mechanisms ensure the credibility of the fused features.

[0064] Example 2: See Figure 3 The implementation of the damage pattern recognition unit begins with an in-depth analysis of the multidimensional durability feature space, constructed by the feature fusion processing unit, which contains comprehensive information on concrete performance degradation. The core task of the unit is to establish a classification model in this high-dimensional space, capable of automatically categorizing the concrete state at any given time into a preset damage development stage. The implementation process first defines four representative damage development stages in the feature space based on extensive historical engineering data, laboratory accelerated aging test data, and theoretical models: the initial stable period, the localized deterioration period, the accelerated degradation period, and the critical failure period. Each stage is not simply defined by a single threshold, but rather by a specific range of values ​​for a set of feature parameters and the patterns of their interrelationships. For example, the initial stable period might be characterized by chloride ion concentration below a critical value, zero crack network connectivity, and low fluctuations in the environmental erosion intensity vector; while the accelerated degradation period might correspond to chloride ion concentration reaching a critical threshold on the steel reinforcement surface, active crack expansion forming connected paths, and a significant negative shift in the steel reinforcement corrosion potential.

[0065] After establishing the theoretical framework for stage division, a density clustering algorithm is used to perform unsupervised learning on accumulated historical and real-time monitoring data to discover the natural clustering of data points in the feature space. This algorithm can adapt to the high dimensionality of the feature space and the complexity of data distribution. It does not pre-determine the shape and number of clusters, but rather identifies dense regions based on the density distribution of the data points themselves. The algorithm scans the entire feature space, marking data points with a sufficient number of neighbors within a specified radius as core points, and grouping interconnected core points and their neighbors into the same cluster. These clusters represent specific "clouds" of different damage stages in the feature space. The center point of each cluster is obtained by calculating the average of the characteristic parameters of all data points within the cluster. The combination of characteristic parameters at this center point is considered the "typical pattern" or "prototype" of that damage stage. For example, the center point of a cluster might contain a median chloride ion concentration, an average crack length, and a typical corrosion potential value; this combination quantifies the typical state of a "localized deterioration period." Based on the above clustering results, the next step is to establish a discriminant function capable of real-time classification. Mahalanobis distance is chosen as the basis for the discriminant function because it considers the correlation between feature parameters, overcoming the inaccuracy of Euclidean distance when the parameters have different scales and are correlated. The system pre-calculates and stores the feature vectors and their covariance matrices representing four typical patterns of four damage stages. When new monitoring data is fused to form a new feature vector, the discriminant function calculates the Mahalanobis distance from this new vector to the feature vectors of the four typical patterns. The calculation of Mahalanobis distance is essentially a weighted distance metric, with the weights determined by the inverse of the covariance matrix of the feature parameters. This compresses the distance in directions of strong correlation and stretches it in directions of strong independence, thus more accurately reflecting the true "similarity" between data points.

[0066] The four calculated distance values ​​represent the degree of difference between the current state and four typical damage states. The identification logic follows the "nearest neighbor" principle, selecting the damage stage with the smallest Mahalanobis distance as the current identification result. This identification process is dynamic and continuous. The system obtains the latest feature vector from the feature fusion processing unit at fixed time intervals (e.g., hourly or daily), performs a classification judgment, and generates a continuous time series label on the health status of the concrete structure. This time series not only indicates the current stage but also reveals the progress of damage development through the transitions between stages. The damage state classification model is not static; it has a built-in update mechanism. After the system has run for a period of time and accumulated enough new monitoring data, especially when the detection result feedback unit provides label data that has been validated in the field, the system will initiate the model retraining process. The retraining process may include fine-tuning the cluster centers or recalculating the covariance matrix based on new data to make the discriminant function more adaptable to the specific characteristics of the actual tunnel environment. The successful operation of this unit depends on the rationality of the feature space construction, the robustness of the clustering algorithm, and the scientific nature of the distance metric.

[0067] Example 3: The implementation of the critical state prediction unit is based on the current damage development stage output by the damage pattern recognition unit. The core task of this unit is to predict the time when concrete reaches the critical state by quantifying the performance degradation dynamics. Its operation begins with extracting the trajectory of feature parameter changes associated with the current identification stage from the multidimensional durability feature space. These parameters include, but are not limited to, chloride ion diffusion coefficient, crack propagation rate, and steel corrosion current density. When extracting the trajectory, the system backtracks historical data within a specific time period, the length of which matches the typical duration of the current damage stage to ensure the completeness and representativeness of the analyzed trajectory. The sliding time window method is used to calculate the local change rate of each feature parameter. The window size is dynamically adjusted according to the data acquisition frequency and parameter fluctuation characteristics. For example, a larger window can be used for slowly changing chloride ion concentration to smooth random fluctuations, while a smaller window is used for rapidly responding crack propagation data to capture instantaneous changes. The method for calculating the local change rate is based on linear regression analysis of the data points within the window, obtaining the slope of the parameter value changing with time within that time interval, which is taken as the degradation rate of the parameter within this window.

[0068] After obtaining the local change rate sequences of each characteristic parameter, it is necessary to establish a mapping relationship between these change rates and the remaining service time of concrete. This relationship is obtained by training a prediction model through the analysis of a large amount of historical case data. The model considers the differences in the impact of the change rate of each parameter on overall durability at different damage stages. For example, in the initial stabilization period, the chloride ion penetration rate may be the dominant factor, while in the critical failure period, the change in crack connectivity may be more critical. To obtain a comprehensive degradation rate index, the system performs a weighted fusion of the change rates of each parameter. The weight coefficients are determined through an expert knowledge base and statistical analysis of historical data, reflecting the relative importance of each parameter under specific conditions. The weighted fusion process can be represented as a linear combination, and its expression is:

[0069]

[0070] Where: symbol This represents the calculated overall degradation rate index; a higher value indicates faster performance deterioration. Indicates the total number of characteristic parameters considered; symbol Representing the The weight coefficients of each feature parameter satisfy the following conditions: ;symbol Representing the The local rate of change of each feature parameter within the current sliding window. This comprehensive index condenses multi-dimensional degradation information into a single, interpretable quantifiable value for subsequent time prediction.

[0071] When predicting the remaining time, the system will use the current overall degradation rate. The remaining time is initially estimated by comparing it with a preset critical state threshold. However, the initial estimate contains uncertainties, thus requiring a correction mechanism. This correction process relies on a historical database storing the actual time taken for concrete to progress from its current state to the critical state at the same damage stage in similar past tunnel projects. The system retrieves cases from historical data that are most similar to the current combination of feature parameters, calculates the average duration of these cases at that stage, and adjusts for differences in environmental conditions, thereby performing a Bayesian update on the initial prediction to obtain a more reliable corrected remaining time. The entire prediction process is iterative. As new monitoring data continuously enters, the sliding window moves forward, the local rate of change is recalculated, the comprehensive degradation rate index is updated, and the predicted remaining time is dynamically adjusted accordingly, enabling the prediction model to learn and adapt online.

[0072] The inspection plan generation unit automatically triggers the corresponding inspection strategy decision logic based on the predicted remaining time value output by the critical state prediction unit. The core basis for the decision is comparing the predicted remaining time with one or more preset time thresholds, which are comprehensively set based on engineering specifications, risk assessment, and economic benefit analysis. When the predicted remaining time is greater than a higher preset threshold (e.g., five years), the system determines that the structure is in a relatively safe state and therefore generates a conventional inspection plan primarily using non-destructive testing methods. The conventional inspection plan focuses on macroscopic, rapid, and comprehensive inspections, and inspection items may include visual inspection of the entire tunnel area, tapping to check for hollow areas, and radar scanning of the thickness of the reinforcing steel protective layer. The inspection cycle is set proportionally to the predicted remaining time, for example, set as a fixed percentage of the remaining time, such as 20%. This means that if the predicted remaining time is ten years, the inspection cycle is set to once every two years. This setting allows the inspection frequency to adaptively increase gradually as the structural condition deteriorates.

[0073] When the predicted remaining time decreases to below the higher threshold but still above a lower threshold (e.g., two years), the system may maintain the conventional detection scheme but will appropriately shorten the detection cycle, for example, by adjusting the cycle to a smaller percentage of the remaining time to increase monitoring density. Once the predicted remaining time falls below the lower threshold, the system determines that the structure has entered a high-risk period and immediately switches to a specialized detection scheme. The specialized detection scheme includes more refined and invasive detection methods. The core items are core sampling and detailed physicochemical analysis in the laboratory, such as determining the concrete compressive strength, chloride ion content profile, and microstructure observation. The cycle of specialized detection is significantly shortened, usually set to half or less of the conventional scheme cycle to ensure close tracking of the degradation process. In addition, the scheme generation logic also deeply integrates damage mode information. In particular, when the damage mode recognition unit indicates that it is currently in an accelerated degradation phase, the generated scheme will force the addition of targeted monitoring items, such as installing deformation gauges at key crack locations to continuously monitor crack propagation rates, or deploying linear polarized resistance sensor arrays to directly measure the instantaneous corrosion rate of the reinforcing steel.

[0074] The output of the detection scheme generation unit is a series of executable instructions that clearly specify the location of the detection operation (high-risk grid areas identified based on feature space), a list of recommended detection methods, and a detailed implementation cycle. The scheme generation process is not rigidly executed; it incorporates resource optimization algorithms that consider tunnel operation constraints, such as scheduling detection operations during off-peak traffic periods and prioritizing resource allocation to the segments with the highest predicted risk and structural importance. The entire process from prediction to scheme generation forms a closed loop. The effectiveness of the detection scheme implementation is recorded and evaluated by the detection result feedback unit, used to continuously optimize the prediction model and decision thresholds, enabling the system to self-improve. The collaborative work of critical state prediction and detection scheme generation achieves seamless integration from condition monitoring to maintenance decision-making, providing precise action guidelines for the tunnel's preventative maintenance.

[0075] Example 4: The implementation of the data quality verification unit spans the entire process from raw data acquisition to feature fusion. Its operation can be illustrated using a hypothetical tunnel segment monitoring scenario. This segment area is equipped with four chloride ion sensors (C1 to C4) and four rebar potential sensors (P1 to P4). The unit continuously monitors the integrity and internal consistency of the data stream. For example, during a data acquisition cycle, the system detects a data packet loss in sensor C3, resulting in a data gap. At this time, the integrity verification module initiates a spatial interpolation procedure. This procedure, based on the principle of spatial correlation, uses the readings of adjacent sensors C2 and C4 at the same time to estimate the value at the location of C3. The interpolation algorithm may employ an inverse distance weighting method, assuming that sensors closer to each other have higher correlation readings and thus are given greater weight, thereby generating an estimated value for the location of C3. This estimated value is marked as "interpolated" and stored separately from the original data.

[0076] Data consistency verification is performed after interpolation. The system compares the logical relationships between different sensor data from the same or adjacent monitoring points. For example, at grid coordinate G-07, the chloride ion sensor C1 reading shows a concentration of 0.15% (percentage of cement mass). According to historical models, the corresponding rebar potential sensor P1 reading at this concentration level should theoretically be between -250mV and -350mV (relative to the Cu / CuSO4 electrode). However, the actual received P1 data is -120mV. This value deviates significantly from the expected range in relation to the chloride ion concentration, exceeding the preset allowable threshold. At this point, the consistency verification module determines that the data pair is abnormal and immediately activates the redundant measurement mechanism. Redundant measurement may be performed by activating a backup sensor in the area (if available) for retesting, or by generating a command to request maintenance personnel to go to location G-07 and perform on-site manual verification measurement using a portable potentiometer. The manually verified potential value is -320mV, which is significantly different from the initial P1 sensor reading of -120mV.

[0077] Based on the verification results, the system marks and processes the original data points. The initial P1 reading of -120mV is marked as "suspicious anomaly," while the manually verified value of -320mV is marked as "verified." The system calculates the confidence weight of each anomalous data point. The weight of the initial anomalous data is significantly reduced, for example, set to 0.2, while the weight of the verified data remains at 1.0. In the subsequent feature fusion processing stage, when the data at this time point is used to calculate the feature vector, these weight values ​​will participate in the calculation. For example, when calculating the average potential, the contribution of anomalous data to the final result is greatly reduced due to its low weight. This weight-based data processing method avoids a few anomalies interfering with the overall judgment and enhances the robustness of the system.

[0078] Table 1: Sensor Data Consistency Verification and Marking

[0079]

[0080] The data quality verification unit also integrates an important drift compensation module, which addresses the inevitable performance degradation or calibration misalignment issues that occur during long-term sensor operation. The module periodically (e.g., quarterly) initiates a calibration process, measuring a batch of standard test blocks placed in the same environment, with the same mix proportions, curing conditions, and environment as the tunnel segment concrete. The baseline physicochemical parameters of these test blocks are known, such as their chloride ion content, accurately measured by laboratory titration. During calibration, monitoring sensors installed near or integrated with the test blocks synchronously read the measurements. The drift compensation module collects the baseline and measured values ​​from all sensors over a period, establishing a "measured value - baseline value" deviation curve for each sensor. For an ideal sensor, its deviation curve should fluctuate slightly around zero. However, a chloride ion sensor in actual operation may show a continuous upward or downward drift trend in its deviation curve. The module fits this deviation curve using linear regression analysis to calculate the average drift of the sensor in the current period and its trend, i.e., the drift compensation coefficient. In subsequent daily data processing, each raw reading of the sensor is automatically corrected by subtracting this compensation coefficient before proceeding to later stages such as feature fusion. For example, a chloride ion sensor calculated to have a positive drift of +0.02% will have a corrected value of 0.08% when its reading is 0.10%. When the drift compensation coefficient of a sensor continues to increase and exceeds a set critical threshold (e.g., the drift reaches 5% of full scale), the drift compensation module will determine that the sensor's performance has severely degraded. At this point, the module will not only continue to perform data correction but also generate a high-level sensor calibration reminder signal to the system administrator.

[0081] Within the data quality verification unit, the overall reliability weight of the sensor is systematically reduced, decreasing its influence in the fusion calculation until maintenance personnel complete on-site calibration or replacement of the sensor and reset its weight and compensation coefficients. This drift compensation mechanism effectively maintains the accuracy and comparability of long-term monitoring data, providing a reliable time-series basis for durability assessment. The entire data quality verification unit acts like a sophisticated filter, screening and correcting the data flowing into the system layer by layer. It first ensures the integrity of the data, then verifies the inherent consistency and rationality of the data, and finally combats measurement errors caused by time through drift compensation. Through a series of combined strategies including real-time monitoring, redundancy verification, weight allocation, and drift compensation, this unit provides high-quality, reliable data raw materials for the entire durability testing system.

[0082] Example 5: The implementation of the detection result feedback unit constitutes a closed loop of system self-learning and optimization. Its operation begins with the comparative analysis of on-site detection results and system prediction results. This unit has a dedicated database to store prediction records generated by the system each time and results verified by subsequent actual detection actions. For example, the system may predict that segment A of tunnel X ring is in a "local deterioration period" and predicts a remaining service life of eight years. After the maintenance team conducts core sampling and laboratory analysis of the area according to the instructions generated by the detection plan, the feedback unit will collect the actual detection report. The report may indicate that the chloride ion concentration of the concrete in the area is close to the critical value, the steel bars show slight corrosion, and the actual condition is assessed as the late stage of the "local deterioration period," with experts estimating a remaining service life of approximately seven years. The system compares the predicted "local deterioration period" with the measured "local deterioration period" to determine that the stage prediction is correct; at the same time, it calculates the relative error between the predicted remaining time (eight years) and the actual assessment time (seven years), which is approximately 14%. This error value will be recorded as a data point for an accuracy index.

[0083] Accuracy metric calculations are periodic, typically accumulating a certain number of detection feedbacks (e.g., ten) to calculate an average accuracy or error distribution. Based on this statistical result, the feedback unit dynamically adjusts the parameters of the upstream model. If the system is found to be consistently optimistic in its predictions of the "accelerated degradation phase" (i.e., the predicted remaining time is always longer than the actual time), the feedback unit may trigger an adjustment to the "accelerated degradation phase" discrimination threshold in the damage state classification model. For example, if a crack width greater than 0.3 mm was originally considered one of the indicators of the "accelerated degradation phase," this threshold might now be revised to 0.25 mm, allowing the model to identify this stage earlier. Simultaneously, the weighting parameters used in the critical state prediction unit to calculate the overall degradation rate may also be adjusted. If the actual importance of a certain feature parameter (such as crack propagation rate) is found to be higher than initially set, its weight coefficient in the weighted fusion will be appropriately increased.

[0084] A key triggering mechanism is continuous deviation monitoring. Suppose the system's three consecutive predictions for a tunnel segment deviate significantly from the actual detection results—for example, the predicted damage stages differ from the actual stages, or the remaining time errors all exceed the allowable range (e.g., 25%). This systematic deviation indicates that the current model may no longer be suitable for the current environmental conditions or material state of the tunnel segment. In this case, the feedback unit will not merely fine-tune the parameters but will trigger a model retraining process. This process uses all recently accumulated new data with measured labels (including monitoring data and corresponding detection results) as a new training set, potentially employing incremental or batch learning methods to re-execute the feature space partitioning and damage pattern clustering processes. This means that the typical pattern feature vectors representing each damage stage and the classification boundaries will be updated to better adapt the model to the latest evolution of the structural state.

[0085] The solution optimization module embedded in the detection result feedback unit focuses on improving the efficiency and effectiveness of the detection operation itself. This module records detailed metadata for each executed detection plan, including the types of detection methods used, the time spent, the equipment and personnel costs invested, and the role the detection ultimately played in confirming or correcting the system's judgment. Based on these records, the module establishes an evaluation index system, with core indicators including detection efficiency (such as average time per detection, cost per unit length of tunnel detection) and detection accuracy (such as defect detection rate, false positive rate). The solution optimization module performs cost-benefit analysis based on this data. For example, suppose in a tunnel detection plan during a "local deterioration period," the standard items include full-section laser scanning and percussion inspection, while a new item is "infrared thermal imaging detection." The solution optimization module analyzes historical data to compare the differences in diagnostic accuracy between detection plans that include and do not include infrared thermal imaging. If the analysis finds that adding infrared thermal imaging only improves diagnostic accuracy by less than 1%, but increases costs by 15%, and that this method does not provide unique, irreplaceable information at this stage of damage, then the optimization module may decide to automatically remove infrared thermal imaging from subsequent detection schemes for similar conditions, and concentrate resources on more effective detection methods.

[0086] The optimization module also considers the importance level and on-site accessibility of tunnel segments. Importance level is determined based on the segment's load-bearing role in the tunnel structure (e.g., main load-bearing segments, connecting segments) and the critical equipment attached to it (e.g., cables, pipelines). Accessibility involves factors such as inspection window, traffic organization difficulty, and safety risks. The module uses an optimization algorithm to prioritize the allocation of limited inspection resources to areas with high importance, high predicted risk, and good accessibility. For example, for a critical load-bearing segment under a pedestrian tunnel, even if its predicted risk is the same as a non-load-bearing segment in a ventilation shaft, the system will allocate more frequent and refined inspection resources. Conversely, for a segment with low predicted risk located in a difficult maintenance area, the system may appropriately extend its inspection cycle or adopt remote monitoring as the primary method. Through this continuous feedback, comparison, adjustment, and optimization, the inspection result feedback unit transforms the entire durability inspection system from a static tool into an intelligent entity that learns from practice and evolves over time.

[0087] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0088] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A durability testing system for double-fiber-coated concrete in cross-river and cross-sea shield tunnels, characterized in that, include: The durability data acquisition unit is used to acquire multi-source monitoring data of the concrete of the shield tunnel segment under service environment. The multi-source monitoring data includes material performance parameters, environmental erosion parameters and structural response parameters. The feature fusion processing unit is used to align and integrate multi-source monitoring data according to time series, extract the performance degradation characteristics of concrete at different erosion stages, and construct a multi-dimensional durability feature space. The damage pattern recognition unit establishes a damage state classification model based on a multidimensional durability feature space to identify the current damage development stage of the concrete. The critical state prediction unit calculates the performance degradation rate based on the change trajectory of the damage development stage and predicts the time node when concrete reaches the critical durability state. The detection scheme generation unit automatically matches the optimal detection scheme based on the critical state prediction results and outputs execution instructions including detection location, detection method and detection cycle; The specific method for the damage pattern recognition unit to establish a damage state classification model is as follows: Four damage development stages are divided in the durability feature space, corresponding to the initial stabilization period, local degradation period, accelerated degradation period and critical failure period, respectively. Density clustering algorithm is used to identify clustered regions of data points in the feature space, and typical patterns of each damage stage are determined based on the combination of feature parameters of the center point of the clustered region. A discrimination function based on Mahalanobis distance is established to calculate the distance between the current monitoring data and each typical pattern in real time, and the damage stage corresponding to the minimum distance value is taken as the identification result.

2. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 1, characterized in that, The specific method by which the durability data acquisition unit acquires multi-source monitoring data is as follows: Data on chloride ion concentration distribution inside concrete and steel corrosion potential are collected using an embedded sensor array. The morphological characteristics and distribution density parameters of apparent cracks in concrete were obtained using surface scanning equipment. Environmental monitoring devices were used to record temperature and humidity variation curves and contact frequency data of corrosive media inside the tunnel.

3. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 2, is characterized in that, The specific steps for the feature fusion processing unit to construct the multidimensional durable feature space are as follows: Chloride ion concentration distribution data and steel corrosion potential data are matched by gridding according to spatial coordinates to generate a material performance degradation characteristic matrix. Topological structure analysis was performed on the morphological characteristic parameters of the apparent cracks to extract the crack network connectivity and propagation direction characteristics. A time-domain correlation analysis was performed between the temperature and humidity change curves and the contact frequency data of the corrosive medium to establish a characteristic vector of environmental erosion intensity. The material performance degradation feature matrix, crack network features, and environmental erosion intensity feature vector are mapped to a unified time coordinate system to form a durability feature space that includes spatial and temporal dimensions.

4. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 1, characterized in that, The specific process by which the critical state prediction unit calculates the performance degradation rate is as follows: Extract the trajectory of characteristic parameter changes corresponding to the current stage of damage development, and calculate the local rate of change of each characteristic parameter using a sliding time window; Establish a mapping relationship between the rate of change of characteristic parameters and the remaining service time, and obtain a comprehensive degradation rate index through weighted fusion; The predicted remaining time for the current stage is adjusted based on the duration of development of the same damage stage in historical data.

5. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 4, characterized in that, The specific logic of the detection scheme generation unit matching the optimal detection scheme is as follows: When the predicted remaining time is greater than a preset threshold, a conventional inspection scheme based on non-destructive testing is generated, and the inspection cycle is set to a fixed proportion of the predicted remaining time. When the predicted remaining time is less than a preset threshold, a special testing plan including core sampling and laboratory analysis is generated, shortening the testing cycle to half that of the conventional plan. For tunnel segments in the accelerated degradation phase, crack propagation monitoring and steel corrosion rate monitoring should be added to the testing plan.

6. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 1, characterized in that, It also includes a data quality verification unit, whose workflow is as follows: Real-time monitoring of the integrity and consistency of data collected by each sensor, and spatial interpolation of missing data using data from adjacent sensors; When the difference in monitoring data from different sensors at the same location exceeds the allowable range, a redundant measurement mechanism is activated to verify the data. Abnormal data points are marked and their credibility weights are calculated. During the feature fusion processing stage, the proportion of data participating in the calculation is adjusted according to the weight values.

7. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 6, characterized in that, The data quality verification unit also includes a drift compensation module, the compensation method of which is as follows: Regularly collect benchmark test data of standard concrete test blocks and establish a deviation curve between sensor measurements and benchmark values; The drift compensation coefficient of each sensor is calculated based on the trend of the deviation curve, and the original monitoring data is automatically corrected during the data processing stage. When the drift compensation coefficient exceeds the critical threshold, a sensor calibration reminder signal is generated and the data weight of the sensor is adjusted.

8. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 1, characterized in that, It also includes a test result feedback unit, whose operating mechanism is as follows: The on-site detection results are compared and analyzed with the system prediction results to calculate the prediction accuracy index. The discrimination threshold and weight parameters of the critical state prediction unit of the damage state classification model are dynamically adjusted based on the accuracy index. When the deviation of three consecutive detection results exceeds the allowable range, the model retraining process is triggered and the feature space partitioning criteria are updated.

9. The durability testing system for cross-river and cross-sea shield tunnels with double-fiber-coated concrete as described in claim 8, characterized in that, The detection result feedback unit also includes a scheme optimization module, whose optimization strategy is as follows: Record the actual implementation effect of each testing plan and establish an evaluation index system for testing efficiency and accuracy; If the improvement in diagnostic accuracy by a newly added test item is less than the preset value, the test item will be removed from the subsequent plan. Optimize the spatial allocation scheme of testing resources based on the importance level and accessibility of the pipeline area.

Images