A highway subgrade structure health monitoring method and system

Abstract

The application provides a highway subgrade structure health monitoring method and system, the method comprising: collecting strain response data under the action of multiple vehicle loads in a reference observation period, and extracting envelope lines and statistical characteristic parameters of each vehicle; constructing a reference envelope line and a reference characteristic distribution; the reference envelope line is the mean value of the envelope lines of multiple vehicles, and the reference characteristic distribution is the statistical distribution of the statistical characteristic parameters of multiple vehicles; in the operation period, real-time strain response data under the action of a current vehicle load is collected, and real-time envelope lines and real-time statistical characteristic parameters of the current vehicle are extracted; the morphological similarity index of the real-time envelope line and the reference envelope line and the difference measurement index of the real-time statistical characteristic parameters and the reference characteristic distribution are determined; when the morphological similarity index is less than a first threshold value and / or the difference measurement index is greater than a second threshold value, it is determined that the highway subgrade has a structural anomaly. The application improves the accuracy and robustness of highway subgrade structure anomaly monitoring.

Description

Technical Field

[0001] This invention relates to the field of highway structure monitoring technology, specifically to a method and system for monitoring the health of highway subgrade structures. Background Technology

[0002] With the advancement of the "Transportation Powerhouse" strategy, my country's expressway mileage has exceeded 177,000 kilometers, with annual maintenance expenditures exceeding 32 billion yuan. Comprehensive acquisition of highway structural information and the realization of scientific maintenance management have become urgent industry needs. Currently, sensors used for highway infrastructure monitoring are mainly divided into two categories. One category is point-type electromagnetic sensors, such as resistance strain gauges and piezoelectric sensors. These sensors can only acquire discrete point data, making it difficult to achieve continuous monitoring of the entire highway cross-section. They are also susceptible to electromagnetic interference and lack long-term stability in harsh road engineering environments such as humidity and strong interference. The other category is distributed fiber optic sensors, such as monitoring systems based on Brillouin scattering. Although they can achieve continuous measurement, they suffer from low strain accuracy and low sampling frequency (usually less than 1Hz), making it difficult to capture the dynamic response under vehicle loads and failing to meet the needs of real-time dynamic monitoring of highway infrastructure.

[0003] In recent years, sensing technology based on weak reflection grating arrays has combined the advantages of high precision of point-type fiber optic gratings with continuous measurement through distributed sensing. It allows for the fabrication of thousands of weak reflection grating sensors on a single fiber, achieving high-density, high-precision, and high-frequency quasi-distributed measurements, making it particularly suitable for large-scale, long-distance highway subgrade monitoring. However, existing monitoring methods based on this technology typically analyze only a single dimension of response indicators (such as peak strain), failing to fully utilize the rich response information acquired by the grating array. This makes it difficult to accurately distinguish between structural degradation and interference factors such as environmental factors and load differences, resulting in high false alarm rates and insufficient early damage identification capabilities.

[0004] Therefore, there is an urgent need to provide a method and system for monitoring the health of highway subgrade structures, which can make full use of high-density strain response data and accurately identify early structural degradation. Summary of the Invention

[0005] In view of this, it is necessary to provide a method and system for monitoring the health of highway subgrade structures to solve the technical problems in the existing technology that fail to make full use of the rich response information obtained by the grating array, resulting in a high false alarm rate and insufficient early damage identification capability.

[0006] To address the aforementioned technical problems, in a first aspect, the present invention provides a method for monitoring the health of highway subgrade structures, comprising: During the baseline observation period, strain response data under multiple vehicle loads are collected based on optical fiber sensing cables laid in the roadbed, and the envelope and statistical characteristic parameters of each vehicle are extracted based on the strain response data. Construct a baseline envelope and a baseline feature distribution; the baseline envelope is the mean of the envelopes of multiple vehicles, and the baseline feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles; During operation, real-time strain response data under the current vehicle load is collected based on the optical fiber sensing cable laid in the roadbed, and the real-time envelope and real-time statistical characteristic parameters of the current vehicle are extracted based on the real-time strain response data. Determine the morphological similarity index between the real-time envelope and the baseline envelope, as well as the difference measurement index between the real-time statistical feature parameters and the baseline feature distribution; When the morphological similarity index is less than the first threshold and / or the difference measurement index is greater than the second threshold, it is determined that the roadbed has a structural anomaly.

[0007] In one possible implementation, the statistical characteristic parameters include amplitude characteristics, energy characteristics, and gradient characteristics. The amplitude characteristics include peak strain and peak-to-valley difference, the energy characteristics include strain integral and equivalent energy, and the gradient characteristics include the strain rate of change along the mileage direction.

[0008] In one possible implementation, the morphological similarity index is the correlation coefficient between the real-time envelope and the reference envelope, and the difference metric is the KL divergence or optimal transmission distance between the real-time statistical feature parameters and the reference feature distribution.

[0009] In one possible implementation, a baseline feature distribution is constructed, including: Principal component analysis is used to reduce the dimensionality of the statistical feature parameters to obtain the principal component feature parameters. The statistical distribution of the principal component feature parameters of multiple vehicles is used as the benchmark feature distribution.

[0010] In one possible implementation, constructing the baseline feature distribution also includes: Multi-vehicle events are classified into multiple load condition categories based on clustering methods; The statistical distributions of the principal component characteristic parameters under each load case category are constructed respectively, serving as the baseline characteristic distributions. The baseline characteristic distributions include multiple baseline characteristic sub-distributions corresponding to the multiple load case categories.

[0011] In one possible implementation, determining the difference metric between the real-time statistical feature parameters and the baseline feature distribution includes: The real-time load condition category is determined based on the real-time statistical feature parameters, and the target benchmark feature sub-distribution among the multiple benchmark feature sub-distributions is determined based on the real-time load condition category. Determine the difference metric between the real-time statistical feature parameters and the target baseline feature sub-distribution.

[0012] In one possible implementation, the method further includes: During the operation period, multiple real-time strain response data are acquired within a preset time period. If all of the real-time strain response data indicate that the roadbed has a structural abnormality, then the roadbed is determined to have a structural abnormality.

[0013] In one possible implementation, the method further includes: Determine the abrupt change point of the morphological similarity index or the difference measurement index along the mileage direction of the roadbed. Obtain the grating measurement point number in the optical fiber sensing cable corresponding to the mutation point, and determine the location of the anomaly based on the mapping relationship between the grating measurement point number and the spatial location of the highway subgrade.

[0014] In one possible implementation, the method further includes: The anomaly level is determined based on the location of the anomaly, the first ratio of the morphological similarity index to the first threshold, and the second ratio of the difference measurement index to the second threshold.

[0015] Secondly, the present invention also provides a highway subgrade structure health monitoring system, comprising: The reference data acquisition unit is used to collect strain response data under multiple vehicle loads based on the optical fiber sensing cable laid in the roadbed during the reference observation period, and to extract the envelope and statistical characteristic parameters of each vehicle based on the strain response data. A benchmark construction unit is used to construct a benchmark envelope and a benchmark feature distribution; the benchmark envelope is the mean of the envelopes of multiple vehicles, and the benchmark feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles. The real-time data acquisition unit is used to collect real-time strain response data under the current vehicle load based on the optical fiber sensing cable laid in the roadbed during the operation period, and to extract the real-time envelope and real-time statistical characteristic parameters of the current vehicle based on the real-time strain response data. The parameter comparison unit is used to determine the morphological similarity index between the real-time envelope and the reference envelope, as well as the difference measurement index between the real-time statistical feature parameters and the reference feature distribution. An anomaly determination unit is used to determine that the roadbed has a structural anomaly when the morphological similarity index is less than a first threshold and / or the difference measurement index is greater than a second threshold.

[0016] The beneficial effects of this invention are as follows: The highway subgrade structure health monitoring method provided by this invention determines highway subgrade structural anomalies based on two dimensions: the morphological similarity index between the real-time envelope and the baseline envelope, and the difference measurement index between real-time statistical characteristic parameters and the baseline characteristic distribution. By comparing the morphological similarity between the real-time envelope and the baseline envelope, macroscopic changes in the structure can be captured from the perspective of overall response morphology. By comparing the difference measurement index between the real-time statistical characteristic parameters and the baseline characteristic distribution, subtle structural degradation can be perceived from the perspective of statistical distribution. The two dimensions complement and verify each other, avoiding over-reliance on a single indicator and significantly improving the accuracy and robustness of anomaly identification.

[0017] Furthermore, in this invention, the baseline envelope is the mean of the envelopes of multiple vehicles, and the baseline feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles. Both the baseline envelope and the baseline feature distribution take into account the differences in loads of different vehicles, effectively eliminating the accidental influence caused by the differences in loads of a single vehicle, and further improving the accuracy and robustness of highway subgrade structural health monitoring.

[0018] Furthermore, this invention, through a two-dimensional comparison of envelope morphology and statistical characteristic distribution, can capture statistical distribution shifts or envelope morphology distortions from strain response data in the early stages of minor degradation within the structure. This enables early diagnosis of hidden structural defects, truly transforming maintenance from reactive repair to preventative maintenance and significantly reducing the overall life-cycle maintenance cost. Simultaneously, this invention fully utilizes the high density, high precision, and high frequency response characteristics of grating array sensing technology to transform raw strain time history data into a structured expression of envelope morphology and statistical characteristic parameters. This eliminates the monitoring blind spots of traditional point sensors, greatly improving the completeness and usability of monitoring data. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A schematic flowchart of an embodiment of the highway subgrade structure health monitoring method provided by the present invention; Figure 2 This is a schematic diagram of an embodiment of the process for constructing a baseline feature distribution provided by the present invention; Figure 3 This is a schematic diagram of another embodiment of the construction of a baseline feature distribution provided by the present invention; Figure 4 A schematic diagram of an embodiment of the present invention for determining a difference measurement index; Figure 5 A schematic flowchart of an embodiment of the present invention for determining the location of an anomaly; Figure 6 This is a schematic diagram of an embodiment of the highway subgrade structure health monitoring system provided by the present invention. Detailed Implementation

[0021] 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 a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0022] It should be understood that the illustrative drawings are not drawn to scale. The flowcharts used in this invention illustrate operations implemented according to some embodiments of the invention. It should be understood that the operations in the flowcharts may be implemented out of order, and steps without logical contextual relationships may be reversed or performed simultaneously. Furthermore, those skilled in the art, guided by the content of this invention, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor systems and / or microcontroller systems.

[0023] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0024] This invention provides a method and system for monitoring the health of highway subgrade structures, which will be described below.

[0025] Figure 1 This is a schematic flowchart of an embodiment of the highway subgrade structure health monitoring method provided by the present invention, as shown below. Figure 1As shown, the methods for monitoring the health of highway subgrade structures include: S101. During the baseline observation period, strain response data under multiple vehicle loads are collected based on optical fiber sensing cables laid in the roadbed, and the envelope and statistical characteristic parameters of each vehicle are extracted based on the strain response data.

[0026] Specifically, the optical fiber sensing cable in this embodiment of the invention is a low-reflection optical fiber sensing cable, which can fabricate up to tens of thousands of low-reflection grating sensors on a single optical fiber, achieving high-density, high-precision, and high-frequency quasi-distributed measurement. Its high multiplexing capability is particularly suitable for large-scale, long-distance highway subgrade monitoring, and it is easy to construct a three-dimensional sensing network and monitoring system for road structures, providing full-section, high-precision data support for highway structure health status assessment and early warning.

[0027] Among them, the fiber optic sensing cable is a direct-buried strain sensing cable with a central tube structure and an internal fixed-point method, which can apply prestress to the sensing fiber to realize continuous large gauge length strain monitoring function.

[0028] To improve the quality of strain response data, in some embodiments of the present invention, the strain response data can be preprocessed, including noise reduction filtering, outlier removal, and temperature compensation.

[0029] S102. Construct the baseline envelope and baseline feature distribution; the baseline envelope is the mean of the envelopes of multiple vehicles, and the baseline feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles.

[0030] Specifically, the fiber optic sensing cable includes multiple grating measurement points deployed along the mileage direction of the highway subgrade. These grating measurement points serve as strain response acquisition points. For each vehicle load, the peak strain response of each measurement point is extracted and connected along the mileage direction to form the vehicle's envelope. During the baseline observation period, for each measurement point, the arithmetic mean of the peak strain responses of multiple vehicles at that point is calculated as the baseline strain value for that point. The baseline strain values ​​of all measurement points are then connected along the mileage direction to form the baseline envelope. The baseline envelope characterizes the typical morphological response of the structure to vehicle loads under healthy conditions.

[0031] For each vehicle under load, statistical characteristic parameters of its strain response data are extracted to form sample points for that vehicle in the feature space. During the baseline observation period, all vehicle sample points are considered as observations from the same population, and their statistical distributions in the feature space are constructed, including the mean vector, covariance matrix, or probability density function, as the baseline characteristic distribution. The baseline characteristic distribution characterizes the overall statistical regularity of the structure's response characteristics under healthy conditions.

[0032] S103. During the operation period, real-time strain response data under the current vehicle load is collected based on the optical fiber sensing cable laid in the roadbed, and the real-time envelope and real-time statistical characteristic parameters of the current vehicle are extracted based on the real-time strain response data. S104. Determine the morphological similarity index between the real-time envelope and the baseline envelope, as well as the difference measurement index between the real-time statistical characteristic parameters and the baseline characteristic distribution; S105. When the morphological similarity index is less than the first threshold and / or the difference measurement index is greater than the second threshold, it is determined that the roadbed has structural anomalies.

[0033] It should be understood that the first and second thresholds can be dynamically set according to the actual application scenario or historical strain response data, which will not be elaborated here.

[0034] It should be understood that the highway subgrade structure health monitoring method in this embodiment of the invention can be implemented in any device based on the highway subgrade structure health monitoring method, such as electronic devices such as highway subgrade structure monitoring equipment or highway management equipment. Specifically, the highway subgrade structure health monitoring method is stored in the aforementioned device as a pre-programmed program. When the device is started, the program is invoked, and the highway subgrade structure health monitoring method is implemented.

[0035] Compared with existing technologies, the highway subgrade structure health monitoring method provided in this invention determines highway subgrade structural anomalies based on two dimensions: the morphological similarity index between the real-time envelope and the baseline envelope, and the difference measurement index between real-time statistical feature parameters and the baseline feature distribution. By comparing the morphological similarity between the real-time envelope and the baseline envelope, macroscopic changes in the structure can be captured from the perspective of overall response morphology. By comparing the difference measurement index between the real-time statistical feature parameters and the baseline feature distribution, subtle structural degradation can be perceived from the perspective of statistical distribution. The two dimensions complement and verify each other, avoiding over-reliance on a single indicator and significantly improving the accuracy and robustness of anomaly identification.

[0036] Furthermore, in this embodiment of the invention, the reference envelope is the mean of the envelopes of multiple vehicles, and the reference feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles. Both the reference envelope and the reference feature distribution take into account the differences in loads of different vehicles, effectively eliminating the accidental influence caused by the differences in loads of a single vehicle, and further improving the accuracy and robustness of highway subgrade structure health monitoring.

[0037] Furthermore, by comparing the envelope morphology and statistical characteristic distribution in a two-dimensional manner, this invention can capture statistical distribution shifts or envelope morphology distortions from strain response data at the early stages of minor degradation within the structure. This enables early diagnosis of hidden structural defects, truly transforming maintenance from reactive repair to preventative maintenance and significantly reducing the overall lifecycle maintenance cost. Simultaneously, this invention fully utilizes the high density, high precision, and high frequency response characteristics of grating array sensing technology to transform raw strain time history data into a structured expression of envelope morphology and statistical characteristic parameters. This eliminates the monitoring blind spots of traditional point sensors, greatly improving the completeness and usability of the monitoring data.

[0038] To improve the completeness and comprehensiveness of statistical parameters and enhance the accuracy and robustness of anomaly detection, in some embodiments of the present invention, statistical characteristic parameters include amplitude characteristics, energy characteristics, and gradient characteristics. Amplitude characteristics include peak strain and peak-to-valley difference, energy characteristics include strain integral value and equivalent energy, and gradient characteristics include strain change rate along the mileage direction.

[0039] In a specific embodiment of the present invention, the morphological similarity index is the correlation coefficient between the real-time envelope and the reference envelope, and the difference measurement index is the KL (Kullback-Leibler) divergence or optimal transmission (Wasserstein) distance between the real-time statistical feature parameters and the reference feature distribution.

[0040] Specifically, the correlation coefficient can be the Pearson correlation coefficient, which assesses the consistency of response from the perspective of overall morphology.

[0041] Although extracting multi-dimensional statistical parameters such as amplitude features, energy features, and gradient features can comprehensively characterize structural response characteristics, these parameters often have information redundancy and strong correlation. If the differences of all parameters are directly measured, on the one hand, redundant information will be introduced and the computational complexity will be increased; on the other hand, noise features may interfere with the anomaly judgment results and reduce the accuracy of monitoring.

[0042] To solve the above-mentioned technical problems, in some embodiments of the present invention, such as Figure 2 As shown, the baseline feature distribution is constructed, including: S201. Dimensionality reduction of statistical characteristic parameters is performed based on principal component analysis (PCA) to obtain principal component characteristic parameters.

[0043] Principal component analysis is a well-established method and will not be elaborated upon here.

[0044] S202. The statistical distribution of the principal component characteristic parameters of multiple vehicles is used as the benchmark characteristic distribution.

[0045] In constructing the baseline feature distribution, this invention introduces principal component analysis (PCA) to reduce the dimensionality of statistical feature parameters. By extracting principal components whose cumulative contribution rate exceeds a set threshold as principal component feature parameters, dimensionality reduction and decorrelation of the feature space are achieved, effectively eliminating redundant information between the original features. Based on this, the baseline feature distribution is constructed using the principal component feature parameters. This not only reduces the computational complexity of subsequent difference measurements but, more importantly, enhances the robustness and generalization ability of the baseline feature distribution by eliminating noisy features that contribute little to the structural state characterization. This provides a more reliable reference benchmark for subsequent anomaly detection, thereby improving the accuracy and robustness of anomaly identification.

[0046] Considering the significant differences in vehicle loads during actual road operation, vehicles with different speeds, axle types, and loads will exhibit significantly different strain response characteristics even when acting on the same healthy structure. Concentrating vehicle response characteristics from all operating conditions into a single baseline feature distribution would result in an excessively broad and complex distribution. Consequently, newly collected vehicle data, even if structurally healthy, might be misjudged as abnormal due to a mismatch between the operating conditions and the baseline distribution. Conversely, setting excessively wide thresholds to reduce false alarms would weaken the sensitivity to actual structural degradation.

[0047] Based on the above-mentioned technical problems, in some embodiments of the present invention, such as Figure 3 As shown, constructing the baseline feature distribution also includes: S301. Based on clustering methods, multi-vehicle events are divided into multiple load condition categories.

[0048] Clustering methods can include K-Means clustering, Gaussian mixture model clustering, or density-based spatial clustering algorithms.

[0049] In specific embodiments of the present invention, the load condition categories may include heavy load low speed, heavy load high speed, light load low speed, light load high speed, etc.

[0050] S302. Construct the statistical distribution of principal component characteristic parameters for each load case category as the baseline characteristic distribution. The baseline characteristic distribution includes multiple baseline characteristic sub-distributions corresponding to multiple load case categories.

[0051] This invention, based on feature dimensionality reduction, further introduces a clustering method to perform cluster analysis on multi-vehicle events in the principal component feature space, automatically identifying and classifying them into multiple load case categories. On this basis, statistical distributions of principal component feature parameters are constructed for each load case category, forming multiple benchmark feature sub-distributions. This effectively eliminates the influence of different load cases on the feature distribution, achieving accurate comparisons under similar load cases. While significantly reducing the false alarm rate, it maintains high sensitivity to early structural degradation, greatly improving the accuracy and reliability of anomaly detection.

[0052] Based on the constructed multiple benchmark feature sub-distributions, in some embodiments of the present invention, such as Figure 4 As shown, the determination of the difference measure index between the real-time statistical characteristic parameters and the benchmark characteristic distribution in step S104 includes: S401. Determine the real-time load condition category based on real-time statistical characteristic parameters, and determine the target benchmark characteristic sub-distribution among multiple benchmark characteristic sub-distributions based on the real-time load condition category.

[0053] The specific process of determining the real-time load condition category based on real-time statistical feature parameters is as follows: project the real-time statistical feature parameters onto the principal component space obtained during the baseline observation period to obtain the principal component feature vector of the vehicle; then, calculate the distance (such as Euclidean distance or Mahalanobis distance) from the principal component feature vector to the cluster center of each load condition category, and determine the load condition category corresponding to the nearest cluster center as the real-time load condition category of the vehicle.

[0054] S402. Determine the difference measurement index between the real-time statistical feature parameters and the target baseline feature sub-distribution.

[0055] This invention, through first identifying the load condition category of a real-time vehicle and then measuring the difference between its statistical characteristic parameters and the corresponding baseline characteristic sub-distribution, achieves a precise anomaly determination mechanism based on similar categories. This avoids false alarms caused by mismatched load conditions while maintaining high sensitivity to early structural degradation. Compared to traditional methods that treat all load conditions the same, this significantly improves the accuracy and reliability of anomaly determination, providing robust technical support for long-term health monitoring of highway subgrade structures in complex traffic environments.

[0056] In actual highway operation environments, occasional disturbances are common, such as temporary obstruction by road obstacles or the passage of temporarily overloaded vehicles. These factors may cause abnormal characteristics in the strain response data under a single vehicle load, but the structure itself has not undergone substantial degradation or damage. If an abnormal alarm is issued based solely on a single judgment, it will lead to false alarms, which will not only increase the verification burden on maintenance personnel, but also reduce the reliability of monitoring and early warning results.

[0057] To effectively filter out occasional interference factors, in some embodiments of the present invention, the method for monitoring the health of highway subgrade structures further includes: During the operation period, multiple real-time strain response data are acquired within a preset time period. If all real-time strain response data indicate that the roadbed has a structural anomaly, then the roadbed is determined to have a structural anomaly.

[0058] The preset duration can be set or adjusted according to the actual application scenario, and no specific limit is set here.

[0059] This invention, based on single-instance anomaly detection, further introduces a time-duration confirmation mechanism. Specifically, a structural anomaly is only confirmed when real-time strain response data under multiple consecutive vehicle loads within a preset time period are all determined to indicate a structural anomaly. This mechanism fully utilizes the persistent nature of structural degradation. That is, once actual structural damage occurs, its impact on the response to subsequent vehicle loads will continue, while occasional interference factors typically only affect a single or a few vehicle responses and lack temporal continuity. Through this mechanism, this invention effectively eliminates false alarms caused by occasional interference, significantly improving the reliability and engineering practical value of structural anomaly monitoring, and providing a more credible early warning basis for maintenance decisions.

[0060] Maintenance personnel need to know the specific location of the abnormality in order to carry out targeted maintenance. Based on this, in some embodiments of the present invention, such as... Figure 5 As shown, methods for monitoring the health of highway subgrade structures also include: S501. Determine the abrupt change points of morphological similarity indicators or difference measurement indicators along the mileage direction of the highway subgrade.

[0061] Specifically: along the mileage direction of the highway subgrade, the morphological similarity index or difference measurement index corresponding to each grating measurement point is arranged according to the spatial location of the measurement point to form an index sequence along the mileage direction; the sliding window statistical method (such as calculating the mean and standard deviation of the index within the window) or the mutation detection algorithm (such as Bayesian change point detection) is used to identify the measurement points in the index sequence where the index changes significantly as mutation points.

[0062] S502. Obtain the grating measurement point number in the optical fiber sensing cable corresponding to the mutation point, and determine the location of the anomaly based on the mapping relationship between the grating measurement point number and the spatial location of the highway subgrade.

[0063] To improve the accuracy of anomaly location determination, in a specific embodiment of this invention, fiber optic sensing cables are distributed in a three-dimensional manner within the highway subgrade. Specifically, this includes embedding fiber optic sensing cables along the depth direction of the highway subgrade in each structural layer, such as the soil base layer, graded crushed stone layer, cement-stabilized crushed stone layer, asphalt-stabilized crushed stone layer (ATB layer), and asphalt surface layer. Simultaneously, sensing cables are also deployed along the transverse direction of the highway at wheel track locations in different lanes (e.g., driving lane, overtaking lane, emergency lane). Through this three-dimensional deployment scheme, when a structural anomaly is detected, the anomaly is focused on a specific structural layer (e.g., graded layer or water-stabilized layer) and a specific lane (e.g., driving lane or overtaking lane), providing maintenance personnel with refined three-dimensional spatial positioning data, significantly improving the targeting and efficiency of maintenance and repair work.

[0064] To achieve differentiated maintenance, in some embodiments of the present invention, the method for monitoring the health of highway subgrade structures further includes: The anomaly level is determined based on the location of the anomaly, the first ratio of the morphological similarity index to the first threshold, and the second ratio of the difference measurement index to the second threshold.

[0065] Specifically, the first ratio characterizes the degree to which the morphological similarity index deviates from the first threshold, reflecting the severity of morphological distortion between the real-time envelope and the baseline envelope; the second ratio characterizes the degree to which the difference measurement index deviates from the second threshold, reflecting the severity of the shift of the real-time statistical feature parameters relative to the baseline feature distribution; and the anomaly location provides spatial information about the defect, including the mileage of the anomaly section, the structural layers involved, and the lateral lane distribution. These three parameters, from different dimensions, constitute the basis for the quantitative assessment of anomaly severity: the first and second ratios measure the degree of structural degradation, while the anomaly location measures the spatial scale of the defect's impact. For example, if the anomaly location involves multiple continuous structural layers or covers multiple lanes, and both the first and second ratios significantly exceed the thresholds, it indicates that the structure has undergone significant and widespread degradation; conversely, if the anomaly is limited to a single layer or point with a small degree of deviation, it may be in the early stage of localized degradation.

[0066] In a specific embodiment of the present invention, the anomaly levels include mild concern, moderate warning, and severe alarm.

[0067] This invention classifies anomalies into levels, allowing maintenance management departments to allocate maintenance resources rationally based on priority. For example, minor anomalies can be included in a regular observation list, moderate anomalies can be scheduled for special inspections, and severe anomalies require immediate repair and handling. This mechanism ensures that limited maintenance funds are prioritized for the highest-risk road sections, significantly improving the scientific and economical nature of maintenance decisions.

[0068] In summary, the highway subgrade structure health monitoring method proposed in this invention has the following advantages: 1. It enables three-dimensional, high-resolution, continuous, and real-time perception of the internal state of the subgrade. Through distributed measurement using a grating array, it eliminates the blind spots of traditional point sensors, capturing and presenting the spatiotemporal distribution and evolution of strain and temperature fields of each structural layer of the subgrade under load and environmental influences in a comprehensive and detailed manner, greatly improving the completeness and accuracy of monitoring data. 2. It achieves multi-physics, large-scale distributed measurement using only a small number of optical cables, significantly reducing the types of sensors, the number of cables, and connection nodes, lowering the complexity of construction and implementation costs, and providing an efficient and economical solution for full-coverage, full-lifecycle health monitoring on long linear infrastructure such as highways. 3. It can detect early degradation of the internal structure before visible damage (such as voids or settlement) appears on the pavement, truly achieving preventative maintenance and significantly reducing maintenance costs. 4. It provides precise mileage and lateral location of defects, providing a basis for accurate repair, avoiding blind large-scale excavation, and saving maintenance time and costs. 5. The inherent electromagnetic interference resistance and corrosion resistance of fiber optic sensing technology, combined with optimized deployment processes, ensure that this method can work stably for a long time in typical harsh environments of road engineering such as humidity and strong interference, without performance degradation. It can effectively overcome the inherent defects of traditional electrical sensors and meet the monitoring needs of the entire life cycle of road infrastructure.

[0069] On the other hand, embodiments of the present invention also provide a highway subgrade structure health monitoring system, such as... Figure 6 As shown, the highway subgrade structure health monitoring system 600 includes: The reference data acquisition unit 601 is used to collect strain response data under multiple vehicle loads based on the optical fiber sensing cable laid in the roadbed during the reference observation period, and to extract the envelope and statistical characteristic parameters of each vehicle based on the strain response data. The baseline construction unit 602 is used to construct the baseline envelope and the baseline feature distribution; the baseline envelope is the mean of the envelopes of multiple vehicles, and the baseline feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles. The real-time data acquisition unit 603 is used to collect real-time strain response data under the current vehicle load based on the optical fiber sensing cable laid in the roadbed during the operation period, and to extract the real-time envelope and real-time statistical characteristic parameters of the current vehicle based on the real-time strain response data. The parameter comparison unit 604 is used to determine the morphological similarity index between the real-time envelope and the reference envelope, as well as the difference measurement index between the real-time statistical characteristic parameters and the reference characteristic distribution. The anomaly determination unit 605 is used to determine that a structural anomaly has occurred in the roadbed when the morphological similarity index is less than a first threshold and / or the difference measurement index is greater than a second threshold.

[0070] The highway subgrade structure health monitoring system 600 provided in the above embodiments can realize the technical solutions described in the above embodiments of the highway subgrade structure health monitoring method. The specific implementation principles of each module or unit can be found in the corresponding content in the above embodiments of the highway subgrade structure health monitoring method, and will not be repeated here.

[0071] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0072] The present invention provides a detailed description of a method and system for monitoring the health of highway subgrade structures. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for monitoring the health of highway subgrade structures, characterized in that, include: During the baseline observation period, strain response data under multiple vehicle loads are collected based on optical fiber sensing cables laid in the roadbed, and the envelope and statistical characteristic parameters of each vehicle are extracted based on the strain response data. Construct the baseline envelope and baseline feature distribution; The baseline envelope is the mean of the envelopes of multiple vehicles, and the baseline feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles. During operation, real-time strain response data under the current vehicle load is collected based on the optical fiber sensing cable laid in the roadbed, and the real-time envelope and real-time statistical characteristic parameters of the current vehicle are extracted based on the real-time strain response data. Determine the morphological similarity index between the real-time envelope and the baseline envelope, as well as the difference measurement index between the real-time statistical feature parameters and the baseline feature distribution; When the morphological similarity index is less than the first threshold and / or the difference measurement index is greater than the second threshold, it is determined that the roadbed has a structural anomaly.

2. The method for monitoring the health of highway subgrade structures according to claim 1, characterized in that, The statistical characteristic parameters include amplitude characteristics, energy characteristics, and gradient characteristics. The amplitude characteristics include peak strain and peak-to-valley difference. The energy characteristics include strain integral value and equivalent energy. The gradient characteristics include strain change rate along the mileage direction.

3. The method for monitoring the health of highway subgrade structures according to claim 1, characterized in that, The morphological similarity index is the correlation coefficient between the real-time envelope and the reference envelope, and the difference metric is the KL divergence or optimal transmission distance between the real-time statistical feature parameters and the reference feature distribution.

4. The method for monitoring the health of highway subgrade structures according to claim 1, characterized in that, Constructing a baseline feature distribution includes: Principal component analysis is used to reduce the dimensionality of the statistical feature parameters to obtain the principal component feature parameters. The statistical distribution of the principal component feature parameters of multiple vehicles is used as the benchmark feature distribution.

5. The method for monitoring the health of highway subgrade structures according to claim 4, characterized in that, Constructing a baseline feature distribution also includes: Multi-vehicle events are classified into multiple load condition categories based on clustering methods; The statistical distributions of the principal component characteristic parameters under each load case category are constructed respectively, serving as the baseline characteristic distributions. The baseline characteristic distributions include multiple baseline characteristic sub-distributions corresponding to the multiple load case categories.

6. The method for monitoring the health of highway subgrade structures according to claim 5, characterized in that, Determining the difference metric between the real-time statistical feature parameters and the baseline feature distribution includes: The real-time load condition category is determined based on the real-time statistical feature parameters, and the target benchmark feature sub-distribution among the multiple benchmark feature sub-distributions is determined based on the real-time load condition category. Determine the difference metric between the real-time statistical feature parameters and the target baseline feature sub-distribution.

7. The method for monitoring the health of highway subgrade structures according to claim 1, characterized in that, The method further includes: During the operation period, multiple real-time strain response data are acquired within a preset time period. If all of the real-time strain response data indicate that the roadbed has a structural abnormality, then the roadbed is determined to have a structural abnormality.

8. The method for monitoring the health of highway subgrade structures according to claim 1, characterized in that, The method further includes: Determine the abrupt change point of the morphological similarity index or the difference measurement index along the mileage direction of the roadbed. Obtain the grating measurement point number in the optical fiber sensing cable corresponding to the mutation point, and determine the location of the anomaly based on the mapping relationship between the grating measurement point number and the spatial location of the highway subgrade.

9. The method for monitoring the health of highway subgrade structures according to claim 8, characterized in that, The method further includes: The anomaly level is determined based on the location of the anomaly, the first ratio of the morphological similarity index to the first threshold, and the second ratio of the difference measurement index to the second threshold.

10. A highway subgrade structure health monitoring system, characterized in that, include: The reference data acquisition unit is used to collect strain response data under multiple vehicle loads based on the optical fiber sensing cable laid in the roadbed during the reference observation period, and to extract the envelope and statistical characteristic parameters of each vehicle based on the strain response data. Benchmark building units are used to construct the benchmark envelope and benchmark feature distribution; The baseline envelope is the mean of the envelopes of multiple vehicles, and the baseline feature distribution is the statistical distribution of the statistical feature parameters of multiple vehicles. The real-time data acquisition unit is used to collect real-time strain response data under the current vehicle load based on the optical fiber sensing cable laid in the roadbed during the operation period, and to extract the real-time envelope and real-time statistical characteristic parameters of the current vehicle based on the real-time strain response data. The parameter comparison unit is used to determine the morphological similarity index between the real-time envelope and the reference envelope, as well as the difference measurement index between the real-time statistical feature parameters and the reference feature distribution. An anomaly determination unit is used to determine that the roadbed has a structural anomaly when the morphological similarity index is less than a first threshold and / or the difference measurement index is greater than a second threshold.

Images