Soil roadbed deformation monitoring method and monitoring system based on distributed optical fiber

Abstract

The present invention provides a soil roadbed deformation monitoring method and system based on distributed optical fiber. This method acquires a distributed optical fiber sensing data set of a target soil roadbed and, based on the deformation intensity distribution of the optical fiber strain data sequence, divides the target soil roadbed into multiple monitoring areas with different deformation characteristics. The optical fiber strain data sequence and the temperature compensation data sequence of each monitoring area are coupled and analyzed to generate a multi-dimensional deformation feature set for the monitoring area. Based on the spatiotemporal matching results of the multi-dimensional deformation feature set with preset deformation threshold conditions, deformation warning information for the target soil roadbed is generated. Based on the deformation warning information, a soil roadbed deformation optimization strategy is generated. This soil roadbed deformation optimization strategy is then fed back to the roadbed maintenance system to trigger reinforcement response operations. This method can effectively enhance the active control capability for maintaining the health of soil roadbeds.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a soil roadbed deformation monitoring method and monitoring system based on distributed optical fiber. Background Art

[0002] In the field of civil engineering health monitoring, soil roadbed deformation monitoring technology assesses structural stability by collecting and analyzing physical deformation data of the roadbed structure in real time. Existing technologies typically use a network of optical fiber strain sensors to collect deformation data. Based on a preset spatial segmentation rule, the roadbed is divided into fixed-length monitoring zones. Deformation displacement indicators are extracted from each zone and statically compared with a uniformly set deformation threshold. Warning signals are triggered when the threshold is exceeded. However, this approach fails to account for strain measurement drift caused by ambient temperature fluctuations, resulting in systematic errors in the identification of deformation intensity distribution. Fixed-length zone divisions fail to accurately capture the natural boundaries of actual deformation gradients, creating blind spots in monitoring critical deformation transmission paths. Relying on simple deformation displacement determination models makes it difficult to distinguish between the mechanisms of internal stress accumulation and external environmental disturbances, resulting in insufficient early warning sensitivity. Furthermore, static threshold settings are prone to false positives and negatives in complex geological conditions. The lack of automated integration between warning information and engineering reinforcement strategies leads to delayed maintenance responses. How to achieve accurate and reliable monitoring and active control of soil roadbed deformation has become a technical problem that needs to be solved urgently in the field of infrastructure safety operation and maintenance. Summary of the Invention

[0003] The present invention provides a soil roadbed deformation monitoring method and monitoring system based on distributed optical fiber.

[0004] In a first aspect, an embodiment of the present invention provides a soil roadbed deformation monitoring method based on distributed optical fiber, comprising: obtaining a distributed optical fiber sensing data set of a target soil roadbed, the distributed optical fiber sensing data set comprising a plurality of spatiotemporally continuous optical fiber strain data sequences and temperature compensation data sequences; dividing the target soil roadbed into a plurality of monitoring areas with different deformation characteristics according to the deformation intensity distribution of the optical fiber strain data sequence; performing coupling analysis processing on the optical fiber strain data sequence and the temperature compensation data sequence of each monitoring area to generate a multi-dimensional deformation feature set of the monitoring area; the multi-dimensional deformation feature set comprises a deformation rate feature, a stress accumulation correlation feature and an environmental disturbance response feature; generating deformation warning information of the target soil roadbed according to the spatiotemporal matching result of the multi-dimensional deformation feature set and a preset deformation threshold condition; generating a soil roadbed deformation optimization strategy based on the deformation warning information, and feeding back the soil roadbed deformation optimization strategy to a roadbed maintenance system to trigger a reinforcement response operation.

[0005] In a second aspect, an embodiment of the present invention provides a monitoring system, comprising: a memory storing a computer program; and a processor for loading the computer program to implement the soil roadbed deformation monitoring method based on distributed optical fiber as described above.

[0006] The distributed optical fiber-based soil roadbed deformation monitoring method provided by this invention overcomes the limitations of traditional monitoring methods that rely on single-parameter static threshold determination by acquiring spatiotemporally continuous optical fiber strain data sequences and temperature-compensated data sequences of the target soil roadbed and performing dynamic coupled analysis. By physically correlating the optical fiber strain data with the temperature-compensated data, strain measurement distortion caused by ambient temperature fluctuations is effectively eliminated, significantly improving the accuracy of deformation intensity distribution identification. Dynamically partitioning monitoring areas based on differential deformation characteristics based on the deformation intensity distribution adaptively captures the true boundaries of the mechanical transmission paths within the soil roadbed, achieving a precise mapping of monitoring area divisions with physical deformation characteristics. By constructing a multidimensional deformation feature set encompassing deformation rate characteristics, stress accumulation correlation characteristics, and environmental disturbance response characteristics, the coupling relationship between deformation evolution mechanisms and external environmental disturbances is revealed within a joint time-frequency domain analysis framework, forming a comprehensive evaluation index with early warning capabilities. By employing a spatiotemporal matching rule with preset deformation threshold conditions, the threshold is dynamically adjusted according to the geological characteristics and evolution stage of different monitoring areas, significantly reducing the false alarm rate under complex geological conditions while improving the timeliness of early warnings. Ultimately, through the automated mapping mechanism of deformation warning information and reinforcement strategies, the multi-dimensional deformation feature analysis results are directly converted into executable engineering optimization plans, achieving a full-link response from deformation monitoring to engineering intervention, and effectively improving the active control capability of maintaining the health status of the soil roadbed. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 This is a flow chart of a soil roadbed deformation monitoring method based on distributed optical fiber provided by an embodiment of the present invention.

[0008] Figure 2 It is a schematic diagram of the composition of a monitoring system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0009] See also Figure 1 , Figure 1 A flowchart of a method for monitoring soil roadbed deformation based on distributed optical fiber provided in an embodiment of the present invention. The method can be executed by a monitoring system and includes the following steps:

[0010] Step S100: obtaining a distributed optical fiber sensing data set of a target soil roadbed, wherein the distributed optical fiber sensing data set includes a plurality of temporally and spatially continuous optical fiber strain data sequences and temperature compensation data sequences.

[0011] Distributed fiber optic sensing data sets are data sets related to the target soil roadbed acquired using distributed fiber optic sensing technology. The fiber optic strain data series is a chronological sequence of strain data generated by the fiber optic strain due to soil roadbed deformation at different time and spatial locations, reflecting the soil roadbed deformation under different temporal and spatial conditions. The temperature compensation data series is collected to eliminate the effects of temperature changes on fiber optic strain measurements. Temperature changes cause the fiber to expand and contract, generating false strain signals, necessitating temperature compensation.

[0012] The process of acquiring distributed fiber optic sensing data from a target soil roadbed can be accomplished by pre-installing distributed fiber optic sensors within the target soil roadbed. For example, distributed fiber optic sensors can be installed at predetermined intervals within the soil roadbed of a highway. These sensors can utilize distributed fiber optic sensing technology based on Raman scattering or Brillouin scattering. During data acquisition, a fiber optic sensor demodulator detects and demodulates scattered light from the optical fiber to obtain fiber strain and temperature data. The fiber strain data is arranged in chronological order of acquisition to form a fiber strain data sequence; the temperature data is similarly arranged in chronological order to form a temperature compensation data sequence.

[0013] Step S200: Divide the target soil roadbed into a plurality of monitoring areas with different deformation characteristics according to the deformation intensity distribution of the optical fiber strain data sequence.

[0014] The deformation intensity distribution of the fiber optic strain data sequence represents the distribution of the deformation intensity of the target soil roadbed as reflected by the fiber optic strain data at different locations and times. Deformation characteristics are the various features exhibited by the soil roadbed during its deformation process, such as deformation rate, amplitude, and stress accumulation. Monitoring areas with differential deformation characteristics are defined as areas within the target soil roadbed where different deformation characteristics vary due to factors such as geological conditions and load conditions. These areas with different deformation characteristics are then separated to form monitoring areas.

[0015] The process of dividing monitoring areas based on the deformation intensity distribution of the optical fiber strain data sequence requires in-depth analysis of the optical fiber strain data sequence. For example, by statistically analyzing the collected optical fiber strain data sequence, the deformation intensity values ​​at different locations are calculated. Then, the target soil roadbed is divided into different areas based on the magnitude and distribution of these deformation intensity values. Areas with large and drastic deformation intensity can be classified as high-deformation monitoring areas, while areas with small and relatively stable deformation intensity can be classified as low-deformation monitoring areas.

[0016] As an implementation manner, step S200 may specifically include the following steps S210 to S240:

[0017] Step S210: performing segmented gradient calculation on the optical fiber strain data sequence, and extracting the maximum deformation gradient value and average deformation fluctuation amplitude of each optical fiber segment.

[0018] Segmented gradient calculation involves segmenting the optical fiber strain data sequence into segments of predetermined lengths or time intervals, then calculating the gradient for each segment. A gradient represents the rate of change of a function at a specific point. In the optical fiber strain data sequence, the gradient reflects the rate of change of deformation of the subgrade within the region corresponding to that fiber segment. The maximum deformation gradient is the maximum value of the gradient within each segment of the optical fiber strain data sequence, reflecting the most dramatic deformation change within the subgrade region corresponding to that fiber segment. The average deformation fluctuation amplitude is the average value of the fluctuation amplitude of the strain data relative to its average value within each segment of the optical fiber strain data sequence, reflecting the degree of deformation stability within the subgrade region corresponding to that fiber segment. Segmented gradient calculation of the optical fiber strain data sequence can be performed using numerical differentiation. For example, the optical fiber strain data sequence can be segmented into 10-meter segments, and the gradient of each segment is calculated using the central difference method. By calculating the gradient for each segment, the maximum value is found, which is the maximum deformation gradient value for that fiber segment. At the same time, the fluctuation amplitude of the strain value in each segment of data relative to its average value is calculated, and then the average value is calculated to obtain the average deformation fluctuation amplitude of the optical fiber segment.

[0019] Step S220: assigning a deformation intensity level to each optical fiber segment based on the ratio of the maximum deformation gradient value to the average deformation fluctuation amplitude.

[0020] The deformation intensity grade classifies fiber segments based on the deformation intensity of the subgrade, indicating the degree of deformation in the subgrade region corresponding to each fiber segment. The ratio of the maximum deformation gradient to the average deformation fluctuation amplitude reflects the combined severity and stability of the subgrade deformation within the region corresponding to the fiber segment.

[0021] A threshold classification method can be used to assign a deformation intensity level to each fiber segment based on the ratio of the maximum deformation gradient value to the average deformation fluctuation amplitude. For example, several thresholds can be pre-set, such as a ratio less than 1 for a low deformation intensity level, a ratio between 1 and 3 for a medium deformation intensity level, and a ratio greater than 3 for a high deformation intensity level. For each fiber segment, the ratio of its maximum deformation gradient value to the average deformation fluctuation amplitude is calculated. Then, based on the comparison of this ratio with the preset threshold, the corresponding deformation intensity level is assigned to the fiber segment. This allows the fiber segments in the target soil roadbed to be classified according to deformation intensity, facilitating subsequent analysis and processing.

[0022] Step S230: clustering continuously distributed optical fiber segments with the same or adjacent deformation intensity levels into initial candidate monitoring areas, and adjusting the area boundaries of the initial candidate monitoring areas according to the deformation intensity level fluctuation range of the initial candidate monitoring areas.

[0023] Continuous distribution refers to spatially adjacent fiber segments with no gaps. Clustering is the process of grouping objects with similar characteristics. In this case, continuously distributed fiber segments with the same or adjacent deformation intensity levels are grouped together to form the initial candidate monitoring area. The deformation intensity level fluctuation range refers to the range of variation in the deformation intensity level of each fiber segment within the initial candidate monitoring area. Adjusting the area boundary involves modifying the boundaries of the initial candidate monitoring area based on the deformation intensity level fluctuation range to more accurately reflect the actual deformation of the soil roadbed.

[0024] The process of clustering continuously distributed fiber segments with the same or adjacent deformation intensity levels into initial candidate monitoring areas can use a clustering algorithm based on adjacency relationships. For example, starting from a certain fiber segment, check the deformation intensity levels of its adjacent fiber segments. If the deformation intensity levels of the adjacent fiber segments are the same or adjacent, add them to the current cluster until there are no adjacent fiber segments that meet the conditions. In this way, an initial candidate monitoring area is formed. Then, calculate the fluctuation range of the deformation intensity level of each fiber segment in the initial candidate monitoring area. If the fluctuation range is large, it means that the deformation situation in the area is relatively complex and the area boundary needs to be adjusted. The adjustment method can be to divide out the fiber segments in the boundary part with large fluctuations, or to merge adjacent areas with similar deformation conditions to improve the accuracy of the monitoring area.

[0025] Step S240: performing region merging or segmentation processing on the adjusted initial candidate monitoring areas according to a preset region division density requirement to generate a final region division result including multiple monitoring areas.

[0026] The preset regional division density requirement is a pre-set requirement for the number, size, and distribution of monitoring areas based on actual monitoring needs and the characteristics of the soil roadbed. Regional merging involves combining adjacent monitoring areas with similar deformation characteristics into a larger monitoring area to reduce the number of monitoring areas and improve monitoring efficiency. Regional segmentation involves dividing a larger monitoring area into multiple smaller monitoring areas according to preset rules to meet regional division density requirements and improve monitoring accuracy. Rule-based methods can be used to merge or segment the adjusted initial candidate monitoring areas based on the preset regional division density requirements. Segmentation can involve selecting an appropriate dividing line based on the distribution of deformation intensity within the area, dividing it into two monitoring areas with an area not exceeding 100 square meters. If the preset regional division density requirement is for the number of monitoring areas to be no more than 10, and the adjusted initial candidate monitoring areas have 15, adjacent areas with similar deformation characteristics need to be merged. This merging can be accomplished by fusing the boundaries of adjacent areas to form a larger monitoring area. By merging or segmenting regions, a final regional division result containing multiple monitoring areas is generated. This result can meet the preset regional division density requirements and more accurately reflect the deformation of the soil roadbed. The differential deformation characteristics are characterized by the median and standard deviation of the deformation intensity level of each monitoring area.

[0027] The median of the deformation intensity level is the value found in the middle of the deformation intensity levels of the fiber segments within the monitoring area, arranged in ascending order. This value reflects the median level of deformation intensity within the monitoring area. The standard deviation measures the degree to which a set of data deviates from its mean. In this case, it represents the degree to which the deformation intensity levels of the fiber segments within the monitoring area deviate from their median. The median and standard deviation together characterize differential deformation characteristics because the median reflects the overall deformation intensity level within the monitoring area, while the standard deviation reflects the degree of dispersion within the monitoring area.

[0028] Step S300: Couple the optical fiber strain data sequence and the temperature compensation data sequence of each monitoring area and perform analysis and processing to generate a multi-dimensional deformation feature set of the monitoring area; the multi-dimensional deformation feature set includes deformation rate features, stress accumulation correlation features, and environmental disturbance response features.

[0029] Coupled analysis is the process of jointly analyzing the relationship between the fiber-optic strain data series and the temperature-compensated data series by comprehensively considering the relationship between them. Because temperature changes can affect fiber-optic strain measurements, it is necessary to couple the temperature-compensated data series with the fiber-optic strain data series to eliminate temperature interference and more accurately obtain soil roadbed deformation information. The multi-dimensional deformation feature set describes the soil roadbed deformation in the monitoring area from different perspectives, including deformation rate features, stress accumulation correlation features, and environmental disturbance response features. The deformation rate feature represents the rate of change of soil roadbed deformation over time within the monitoring area, reflecting the speed of soil roadbed deformation. The stress accumulation correlation feature represents the correlation between stress accumulation and deformation within the monitoring area, reflecting the impact of stress accumulation on deformation during long-term stress loading. The environmental disturbance response feature represents the response of the soil roadbed within the monitoring area to changes in environmental factors (such as temperature, humidity, and rainfall), reflecting the impact of environmental factors on soil roadbed deformation.

[0030] As an implementation manner, step S300 may specifically include the following steps S310 to S350:

[0031] Step S310: performing noise filtering and baseline calibration on the optical fiber strain data sequence to obtain a denoised standardized strain data sequence.

[0032] Noise filtering is the process of removing noise signals from the fiber optic strain data sequence. Noise signals can be caused by factors such as measurement errors in the fiber optic sensor and external interference, which can affect the accurate assessment of soil roadbed deformation. Baseline calibration adjusts the baseline of the fiber optic strain data sequence to an appropriate level to eliminate deviations caused by factors such as the sensor's initial state and installation errors. The resulting standardized strain data sequence after noise filtering and baseline calibration, with noise removed and the baseline adjusted, more accurately reflects the actual soil roadbed deformation.

[0033] The process of performing noise filtering and baseline calibration on the optical fiber strain data sequence is, for example: For noise filtering, a digital filtering algorithm, such as a sliding average filtering algorithm or a median filtering algorithm, can be used. Taking the sliding average filtering algorithm as an example, the basic idea of ​​this algorithm is to take the average of a set number of data points before and after each data point in the optical fiber strain data sequence as the filtered value of that point. For example, for an optical fiber strain data sequence of length N, a sliding window of length M is used. For the i-th data point, its filtered value is the average of that point and the (M-1) / 2 data points before and after it. For baseline calibration, a linear regression method can be used to linearly fit the optical fiber strain data over a period of time to obtain the slope and intercept of the baseline. The baseline is then subtracted from the data sequence to obtain the calibrated strain data sequence. Through noise filtering and baseline calibration, a denoised and standardized strain data sequence is obtained.

[0034] Step S320: constructing a temperature strain compensation coefficient matrix according to the temperature variation trend of the temperature compensation data sequence, and performing temperature drift correction on the standardized strain data sequence using the temperature strain compensation coefficient matrix to generate a corrected target strain data sequence.

[0035] The temperature change trend is the temporal temperature variation in the temperature-compensated data sequence, including trends such as temperature increases, decreases, and fluctuations. The temperature strain compensation coefficient matrix is ​​used to describe the impact of temperature changes on optical fiber strain measurement. The elements of the matrix represent the compensation coefficients required for the optical fiber strain data under different temperature variations. Temperature drift correction utilizes the temperature strain compensation coefficient matrix to adjust the standardized strain data sequence to eliminate strain measurement errors caused by temperature changes. The corrected target strain data sequence is the optical fiber strain data sequence that has been corrected for temperature drift, eliminating the effects of temperature. This sequence more accurately reflects the actual deformation of the soil roadbed. The process of constructing the temperature strain compensation coefficient matrix, for example, involves analyzing the temperature-compensated data sequence and dividing the temperature variation into different intervals, such as a temperature-rising interval, a temperature-decreasing interval, and a temperature-stable interval. Then, within each temperature interval, the relationship between temperature change and optical fiber strain change is determined through experimental or theoretical analysis to obtain the temperature strain compensation coefficient for that interval. These compensation coefficients are arranged in order of the temperature intervals to form the temperature strain compensation coefficient matrix.

[0036] Step S330: performing a joint time-frequency domain analysis on the target strain data sequence to extract the time-domain deformation accumulation and frequency-domain disturbance response intensity of the monitoring area.

[0037] Joint time-frequency domain analysis involves analyzing the target strain data series simultaneously in both the time and frequency domains. While time-domain analysis focuses on how the data changes over time, frequency-domain analysis focuses on the distribution of different frequency components within the data. Time-domain deformation accumulation measures the cumulative deformation of the subgrade within the monitoring area over a set period of time, reflecting the overall deformation of the subgrade over that period. Frequency-domain disturbance response intensity measures the response strength of the subgrade within the monitoring area to disturbance signals of different frequencies, reflecting its dynamic characteristics at different frequencies.

[0038] The process of performing a joint time-domain and frequency-domain analysis on the target strain data series to extract the time-domain deformation accumulation and frequency-domain disturbance response intensity of the monitored area is as follows: For time-domain analysis, the target strain data series can be integrated to obtain the time-domain deformation accumulation. For frequency-domain analysis, a fast Fourier transform (FFT) algorithm can be used to convert the target strain data series from the time domain to the frequency domain to obtain its spectrum. The amplitudes of the different frequency components in the spectrum are then analyzed to determine the frequency-domain disturbance response intensity.

[0039] Step S340: Calculate the stress accumulation correlation characteristic value of the monitoring area according to the nonlinear correlation relationship between the time domain deformation accumulation amount and the frequency domain disturbance response intensity.

[0040] Specifically, first, through experiments or theoretical analysis, a nonlinear function model is established between the time domain deformation accumulation and the frequency domain disturbance response intensity. For example, a polynomial regression model or a neural network model can be used. Taking the polynomial regression model as an example, let the time domain deformation accumulation be C, the frequency domain disturbance response intensity be F, and the stress accumulation correlation characteristic value be S, then the following polynomial regression model can be established: S = a0 + a1C + a2F + a3C 2 +a4F 2 +a5CF, where a0, a1, a2, a3, a4, and a5 are the coefficients of the model, which can be obtained by fitting the experimental data. Then, the calculated time-domain deformation accumulation and frequency-domain disturbance response intensity are substituted into the model to calculate the stress accumulation correlation characteristic value of the monitoring area.

[0041] Step S350: The time domain deformation accumulation, the frequency domain disturbance response intensity and the stress accumulation correlation characteristic value are integrated to generate a multi-dimensional deformation feature set of the monitoring area.

[0042] Specifically, first, normalize the time-domain deformation accumulation, frequency-domain disturbance response intensity, and stress accumulation correlation eigenvalues ​​to unify their numerical ranges to an appropriate interval, such as [0, 1]. This normalization can be performed using linear normalization methods, such as the maximum-minimum normalization method. Then, the normalized time-domain deformation accumulation, frequency-domain disturbance response intensity, and stress accumulation correlation eigenvalues ​​are combined to form a vector, which is the multidimensional deformation feature set of the monitored area.

[0043] Step S400: generating deformation warning information of the target soil roadbed according to the spatiotemporal matching result of the multi-dimensional deformation feature set and the preset deformation threshold condition.

[0044] The multi-dimensional deformation feature set is a feature set that describes the deformation of the soil roadbed in the monitoring area from different perspectives, including deformation rate characteristics, stress accumulation correlation characteristics, and environmental disturbance response characteristics. The preset deformation threshold conditions are pre-set threshold conditions for soil roadbed deformation based on factors such as the design requirements and safety standards of the soil roadbed, including thresholds for deformation size, rate, and accumulation. The spatiotemporal matching result is the result of comparing and matching the multi-dimensional deformation feature set with the preset deformation threshold conditions in time and space, reflecting whether the deformation in the monitoring area exceeds the preset threshold conditions. Deformation warning information is warning information about the deformation of the target soil roadbed generated based on the spatiotemporal matching results, including warning trigger conditions, risk area location identification, and recommended response time.

[0045] As an implementation manner, step S400 may specifically include the following steps S410 to S450:

[0046] Step S410: Obtain a historical deformation feature library of the monitored area, where the historical deformation feature library contains multi-dimensional deformation feature data of multiple historical deformation cases and corresponding deformation evolution results.

[0047] The historical deformation feature database for the monitored area stores information on past deformation cases within the monitored area. Historical deformation cases are soil roadbed deformation events that have occurred within the monitored area. Multi-dimensional deformation feature data describes the soil roadbed deformation in these historical deformation cases from different perspectives, including deformation rate characteristics, stress accumulation correlation characteristics, and environmental disturbance response characteristics. Deformation evolution results describe the development process and final outcome of soil roadbed deformation in these historical deformation cases, such as whether severe conditions such as collapse and subsidence occurred.

[0048] The process of obtaining a historical deformation feature library for the monitored area can be achieved using the following method. First, historical monitoring data for the monitored area is collected and organized. This historical monitoring data can include soil roadbed deformation data collected over the past few years or even decades using distributed fiber optic sensors and other equipment. Next, this historical monitoring data is analyzed and processed to extract multi-dimensional deformation feature data. For example, using the method described in the previous step, coupled analysis and processing is performed on the historical fiber optic strain data series and the temperature compensation data series to extract deformation rate characteristics, stress accumulation correlation characteristics, and environmental disturbance response characteristics. Simultaneously, the deformation evolution results for each historical deformation case are recorded, such as whether reinforcement measures were implemented and the effectiveness of the reinforcement. Finally, this multi-dimensional deformation feature data and the corresponding deformation evolution results are stored in a database to form a historical deformation feature library for the monitored area. Obtaining this historical deformation feature library provides a reference and comparison basis for subsequent early warning analysis.

[0049] Step S420: performing similarity matching between the multi-dimensional deformation feature set and feature data in the historical deformation feature library to determine target matching cases between the current deformation pattern of the monitoring area and the historical deformation cases.

[0050] Similarity matching involves comparing a multidimensional deformation feature set with feature data in a historical deformation feature library to identify the most similar feature data. The current deformation pattern is the current deformation of the subgrade in the monitored area, described by a multidimensional deformation feature set. The target matching case is the historical deformation case in the historical deformation feature library that is most similar to the current deformation pattern in the monitored area.

[0051] The process of similarity matching the multidimensional deformation feature set with the feature data in the historical deformation feature library to determine the target matching case between the current deformation pattern of the monitored area and the historical deformation case is, for example, as follows: First, a similarity measurement method is selected, such as Euclidean distance or cosine similarity. Then, the multidimensional deformation feature set is compared with the feature data of each historical deformation case in the historical deformation feature library, and a similarity measurement value is calculated between them. Finally, the historical deformation case with the smallest similarity measurement value is selected as the target matching case. Determining the target matching case through similarity matching can provide a reference for subsequent deformation development trend prediction.

[0052] Step S430: predicting the deformation development trend of the monitoring area based on the deformation evolution results of the target matching case, and generating a preliminary warning level based on the degree of deviation between the deformation development trend and the preset deformation threshold condition.

[0053] The deformation trend represents the expected change in subgrade deformation within the monitored area over a period of time, including changes in deformation magnitude, rate, and cumulative volume. The deviation degree represents the difference between the deformation trend and the preset deformation threshold, reflecting whether the deformation in the monitored area has exceeded a safe range. The preliminary warning level, determined based on the degree of deviation between the deformation trend and the preset deformation threshold, is a preliminary warning level for the monitored area, typically low, medium, or high.

[0054] The process of predicting the deformation trend of the monitored area based on the deformation evolution results of the target matching case and generating a preliminary warning level based on the degree of deviation between the deformation trend and the preset deformation threshold condition is as follows: First, analyze the deformation evolution results of the target matching case to understand the development process and patterns of soil roadbed deformation in this case. For example, observe the changes in deformation rate over time and the growth trend of stress accumulation in this case. Then, based on the similarity between the current deformation pattern of the monitored area and the target matching case, assume that the deformation trend of the monitored area is similar to that of the target matching case, and thus predict the deformation trend of the monitored area. Next, compare the predicted deformation trend with the preset deformation threshold condition and calculate the degree of deviation between them. For example, for deformation rate, calculate the difference between the predicted deformation rate and the preset deformation rate threshold; for stress accumulation, calculate the difference between the predicted stress accumulation and the preset stress accumulation threshold. Finally, determine the preliminary warning level based on the degree of deviation. For example, if the degree of deviation is small, the preliminary warning level is low; if the degree of deviation is medium, the preliminary warning level is medium; if the degree of deviation is large, the preliminary warning level is high.

[0055] Step S440: Based on the spatial distribution correlation of the preliminary warning levels of adjacent monitoring areas, the preliminary warning levels are subjected to spatiotemporal correction processing to generate the final warning level of the monitoring area.

[0056] Spatial correlation refers to the relationship between the preliminary warning levels of adjacent monitoring areas. Since soil roadbed deformation is spatially continuous, the deformation of adjacent monitoring areas often influences each other, so their preliminary warning levels may also be correlated. Spatiotemporal correction is the process of adjusting and revising the preliminary warning levels by comprehensively considering both temporal and spatial factors. The final warning level is the final warning level for the monitoring area after spatiotemporal correction, which more accurately reflects the actual deformation risk in the monitoring area.

[0057] The process of performing spatiotemporal correction on the preliminary warning levels based on the spatial distribution correlation of the preliminary warning levels of adjacent monitoring areas to generate the final warning level for the monitoring area is, for example, as follows: First, the spatial topological relationship between adjacent monitoring areas is established to determine the adjacent monitoring areas of each monitoring area. For example, for the distribution of monitoring areas on a two-dimensional plane, the adjacent monitoring areas can be determined using a four-neighborhood or eight-neighborhood method. Then, the spatial distribution of the preliminary warning levels of the adjacent monitoring areas is analyzed, and the correlation between them is calculated. For example, the correlation between the preliminary warning levels of adjacent monitoring areas can be calculated using methods such as the Pearson correlation coefficient. If the correlation between the preliminary warning levels of adjacent monitoring areas is high, it indicates that their deformations may have a significant impact on each other, and the preliminary warning level of the current monitoring area needs to be adjusted. This adjustment can be performed based on the average or weighted average of the preliminary warning levels of the adjacent monitoring areas. For example, if the preliminary warning levels of the adjacent monitoring areas are all high, while the preliminary warning level of the current monitoring area is low, the preliminary warning level of the current monitoring area can be appropriately increased. Finally, after spatiotemporal correction, the final warning level of the monitoring area is obtained. Spatiotemporal correction can improve the accuracy of the warning level and reduce false positives and false negatives.

[0058] Step S450: Based on the spatial superposition results of the final warning levels of all monitoring areas, generate the overall deformation warning information of the target soil roadbed; the deformation warning information includes warning trigger conditions, risk area location identifiers and recommended response time.

[0059] The spatial overlay results are the result of spatially overlaying and comprehensively analyzing the final warning levels of all monitored areas, reflecting the overall deformation risk distribution of the target soil roadbed. The overall deformation warning information is generated based on the spatial overlay results and includes warning trigger conditions, risk area location identifiers, and recommended response time limits. Warning trigger conditions are those that trigger a deformation warning, such as a deformation rate exceeding a set threshold or accumulated stress reaching a certain level. Risk area location identifiers clearly indicate which areas of the target soil roadbed are at risk of deformation, such as the location and extent of the area. The recommended response time limit is the time frame within which measures are recommended to be taken to address the deformation risk.

[0060] The process of generating the overall deformation warning information of the target soil roadbed is, for example, to spatially superimpose the final warning levels of all monitored areas to form a warning level distribution map. In this distribution map, different colors or symbols can be used to represent different warning levels, intuitively showing the overall deformation risk distribution of the target soil roadbed. According to the warning level distribution map, the warning trigger conditions are determined. For example, if the warning level of a certain area reaches a high level, it is considered that the area has triggered a warning. The warning trigger condition can be that the deformation rate of the area exceeds a preset high threshold. Then, based on the warning level distribution map, a risk area location identifier is generated. Technologies such as geographic information systems (GIS) can be used to mark the location and scope of the risk area on the map to form a clear risk area location identifier. Finally, based on the warning level and the situation of the risk area, the recommended response time is determined.

[0061] Step S500: generating a soil roadbed deformation optimization strategy based on the deformation warning information, and feeding back the soil roadbed deformation optimization strategy to the roadbed maintenance system to trigger a reinforcement response operation.

[0062] The soil roadbed deformation optimization strategy is a strategy developed based on deformation warning information to optimize soil roadbed deformation and reduce deformation risks. It includes reinforcement material configuration parameters, construction schedule planning, and resource allocation ratios. The roadbed maintenance system is a system for maintaining and managing soil roadbeds. This system receives soil roadbed deformation optimization strategies and triggers corresponding reinforcement response operations based on these strategies. Reinforcement response operations are reinforcement measures implemented in response to soil roadbed deformation risks, such as laying reinforcement materials and performing soil improvements.

[0063] As an implementation method, step S500, generating a soil roadbed deformation optimization strategy based on deformation warning information, may specifically include the following steps S510 to S540:

[0064] Step S510: Determine at least one high-risk monitoring area in the target soil roadbed that needs to be processed first based on the risk area location identifier.

[0065] Risk area positioning markers clearly indicate which areas within the target soil roadbed are at risk of deformation, such as their location and extent. High-risk monitoring areas are those within the target soil roadbed with higher warning levels and greater deformation risks. These areas require priority treatment to avoid serious deformation accidents.

[0066] The process of determining at least one high-risk monitoring area in the target soil roadbed that needs to be treated as a priority based on the risk area positioning identifier is, for example: first, according to the risk area positioning identifier, the location and range information of each monitoring area in the target soil roadbed is obtained. Then, combined with the warning level in the deformation warning information, the monitoring areas with high warning levels are screened out. For example, if the deformation warning information clearly indicates that the warning level of a certain area is high, then the area is determined as a high-risk monitoring area. If there are multiple monitoring areas with high warning levels, the order of priority treatment can be further determined based on factors such as the importance of these areas and the urgency of the deformation risk. For example, for a high-risk monitoring area located below a major traffic artery, it is necessary to give priority to it because it may have a serious impact on traffic once it is deformed. Through the above steps, at least one high-risk monitoring area in the target soil roadbed that needs to be treated as a priority is determined, providing a clear target for subsequent processing.

[0067] Step S520: Acquire geological structure data and historical maintenance records of the high-risk monitoring area, and construct a deformation response model of the high-risk monitoring area.

[0068] Geological structural data describes the underground geological structure and geotechnical properties of high-risk monitoring areas, including soil stratification data and geotechnical parameters. Historical maintenance records include records of past maintenance and reinforcement efforts in high-risk monitoring areas, including maintenance time, maintenance measures, and effectiveness. The deformation response model describes the deformation response of high-risk monitoring areas under different loads. This model can predict deformation in high-risk monitoring areas based on input load data.

[0069] The process for obtaining geological structural data and historical maintenance records for a high-risk monitoring area and constructing a deformation response model for that area, for example, involves the following: First, geological structural data for the high-risk monitoring area is obtained through geological exploration and geotechnical testing. For example, soil stratification data can be obtained through drilling and geophysical exploration, and geotechnical parameters can be obtained through indoor geotechnical testing. Furthermore, relevant archival materials are consulted to obtain historical maintenance records for the high-risk monitoring area. Then, based on the geological structural data and historical maintenance records, an appropriate model structure and parameters are selected to construct a deformation response model. For example, a finite element model or discrete element model can be used. For example, a finite element model divides the high-risk monitoring area into multiple finite elements and solves the deformation of the entire area by establishing mechanical equilibrium equations for each element. During model construction, boundary conditions and material parameters must be determined based on the geological structural data and historical maintenance records. Finally, the constructed deformation response model is verified and calibrated to ensure its accuracy and reliability. Through these steps, a deformation response model for the high-risk monitoring area is constructed, providing a foundation for subsequent reinforcement solution simulations.

[0070] As an embodiment, the process of constructing the deformation response model includes the following steps S521 to S524:

[0071] Step S521: Collect soil stratification data, moisture content distribution data and load history data in the high-risk monitoring area.

[0072] Soil stratification data describes the stratification of underground soil layers within high-risk monitoring areas, including soil thickness and rock and soil types. Moisture content distribution data describes the distribution of moisture content within underground soil layers within high-risk monitoring areas, as moisture content has a significant impact on the mechanical properties of rock and soil. Load history data is data on past loads experienced within high-risk monitoring areas, including load magnitude, duration, and frequency.

[0073] The process of collecting soil stratification data, moisture distribution data, and load history data in high-risk monitoring areas is as follows: For soil stratification data, drilling, geophysical exploration, and other methods can be used for collection. The drilling method is to obtain samples of underground soil layers by drilling holes, and then analyze the samples to determine information such as the thickness of the soil layers and the rock and soil type. The geophysical exploration method uses geophysical principles to infer the stratification of underground soil layers by measuring changes in the underground physical field. For example, resistivity method, seismic wave method, etc. can be used. For moisture distribution data, drying method, capacitance method, etc. can be used for collection.

[0074] Step S522: constructing a layered geomechanical model based on the soil layer data, and correcting the interlayer shear strength parameters of the layered geomechanical model using the moisture content distribution data.

[0075] The layered geomechanical model, constructed based on soil layer data, is used to describe the mechanical properties of subsurface soil in high-risk monitoring areas. This model divides the subsurface into multiple layers, each with distinct mechanical properties. The interlayer shear strength parameter, which describes the shear strength between adjacent soil layers, has a significant impact on soil stability. Moisture content has a significant impact on the mechanical properties of rock and soil, particularly the interlayer shear strength parameter. Therefore, moisture content distribution data is required to modify the interlayer shear strength parameter in the layered geomechanical model.

[0076] The process of constructing a layered geomechanical model based on soil layer stratification data and correcting the interlayer shear strength parameters of the layered geomechanical model using moisture content distribution data is as follows: First, based on the soil layer stratification data, the underground soil layer in the high-risk monitoring area is divided into multiple layers, each with different rock and soil types and thicknesses. Then, an appropriate mechanical model and parameters are selected for each layer to construct the layered geomechanical model. For example, the Mohr-Coulomb model or the Duncan-Zhang model can be used. Taking the Mohr-Coulomb model as an example, this model describes the mechanical properties of rock and soil through the angle of internal friction and cohesion. Next, based on the moisture content distribution data, the effect of moisture content on the interlayer shear strength parameters is analyzed. The relationship between moisture content and interlayer shear strength parameters can be established through experiments or empirical formulas. For example, for a certain rock and soil type, the interlayer shear strength at different moisture contents can be measured through indoor direct shear tests, and then a relationship curve between moisture content and interlayer shear strength parameters can be fitted. Finally, the interlayer shear strength parameters of the layered geomechanical model are corrected based on the moisture content distribution data and the established relationship curve. Through the above steps, a more accurate layered geomechanical model is constructed, providing a basis for subsequent calculations.

[0077] Step S523: Input the load history data into the revised layered geomechanical model to calculate the theoretical deformation response curve of the high-risk monitoring area under different load scenarios.

[0078] Historical load data is data related to the load conditions experienced by the high-risk monitoring area in the past, including load magnitude, duration, and frequency. The theoretical deformation response curve is a curve showing the theoretical deformation of the high-risk monitoring area over time under different load scenarios, reflecting the deformation response of the high-risk monitoring area under different load conditions.

[0079] The process of inputting load history data into the revised hierarchical geomechanical model to calculate the theoretical deformation response curves for the high-risk monitoring area under different load scenarios is as follows: First, different load scenarios are determined based on the load history data. For example, the load history data can be classified according to factors such as load magnitude and duration to obtain different load scenarios. Then, the load data for each load scenario is input into the revised hierarchical geomechanical model. In the model, the mechanical equilibrium equation is solved for each time step based on the mechanical properties of the rock and soil and the boundary conditions to obtain the deformation of the high-risk monitoring area at that time step. By calculating each time step, the deformation of the high-risk monitoring area during the entire load process is obtained. Finally, the calculated deformation scenarios are arranged in chronological order to plot the theoretical deformation response curve. For example, for a load process lasting T, the theoretical deformation response curve is plotted with the deformation value at each time step as the ordinate and time as the abscissa. Through these steps, the theoretical deformation response curves for the high-risk monitoring area under different load scenarios are calculated, providing a reference for subsequent model calibration.

[0080] Step S524: By comparing the theoretical deformation response curve with the actual deformation monitoring data of the high-risk monitoring area, the parameter error of the layered geomechanical model is calibrated to obtain a calibrated deformation response model; wherein, the parameter error calibration process adopts an iterative feedback mechanism until the matching degree between the theoretical deformation response curve and the actual deformation monitoring data reaches a preset threshold.

[0081] Actual deformation monitoring data refers to the actual deformation data of high-risk monitoring areas collected through distributed fiber optic sensors and other equipment, reflecting the true deformation of high-risk monitoring areas under actual loads. Parameter error refers to the difference between the actual and theoretical values ​​of parameters in the layered geomechanical model. This difference can lead to deviations between the theoretical deformation response curve and the actual deformation monitoring data. The iterative feedback mechanism is a process that gradually improves the matching degree between the theoretical deformation response curve and the actual deformation monitoring data by continuously adjusting the parameters of the layered geomechanical model. The preset threshold is a pre-set threshold for the matching degree between the theoretical deformation response curve and the actual deformation monitoring data. When the matching degree reaches this threshold, the model parameter error is considered to have been effectively calibrated.

[0082] By comparing the theoretical deformation response curve with actual deformation monitoring data from high-risk monitoring areas, the parameter errors of the layered geomechanical model are calibrated to obtain the calibrated deformation response model. For example, the calculated theoretical deformation response curve is compared with the actual deformation monitoring data to calculate the degree of match between them. The degree of match can be measured using metrics such as the correlation coefficient and mean square error. For example, the correlation coefficient is used to measure the degree of match; the closer the correlation coefficient is to 1, the higher the degree of match between the theoretical deformation response curve and the actual deformation monitoring data. Then, based on the calculated degree of match, it is determined whether the parameters of the layered geomechanical model need to be adjusted. If the degree of match does not reach a preset threshold, the model parameters need to be adjusted. Trial-and-error methods, optimization algorithms, and other methods can be used for adjustment. For example, in the trial-and-error method, the model parameters are adjusted based on experience or theoretical analysis. The theoretical deformation response curve is then recalculated and compared again with the actual deformation monitoring data to calculate the degree of match. This process is repeated until the degree of match reaches a preset threshold. Finally, when the matching degree reaches the preset threshold, it is considered that the parameter error of the model has been effectively calibrated, and the layered geomechanical model obtained at this time is the calibrated deformation response model.

[0083] Step S530: Based on the deformation response model, simulate the deformation suppression effects of different reinforcement schemes on the high-risk monitoring area, and select the optimal reinforcement scheme based on the simulation results.

[0084] The reinforcement scheme is a method used to reinforce high-risk monitoring areas, including the type of reinforcement material, layout density, and structural form. The deformation suppression effect is the degree to which deformation in the high-risk monitoring area is suppressed after adopting different reinforcement schemes, reflecting the effectiveness of the reinforcement scheme. The optimal reinforcement scheme is the one with the best deformation suppression effect among all simulated reinforcement schemes.

[0085] As an embodiment, step S530 simulates the deformation suppression effect of different reinforcement schemes on the high-risk monitoring area, which may specifically include the following steps S531 to S534:

[0086] Step S531: Obtain a set of candidate reinforcement solutions, where the candidate reinforcement solution set includes a variety of combinations of reinforcement material types, layout densities, and structural forms.

[0087] The candidate reinforcement scheme set consists of a collection of different reinforcement schemes, each composed of different combinations of reinforcement material types, deployment densities, and structural forms. Reinforcement material types refer to the type of material used to reinforce the high-risk monitoring area, such as geogrids, cement mixing piles, and reinforced concrete piles. Deployment density refers to the distribution density of the reinforcement material within the high-risk monitoring area, such as the spacing of geogrids and the number of cement mixing piles. Structural form refers to the arrangement and connection method of the reinforcement materials, such as the laying direction of geogrids and the arrangement of cement mixing piles.

[0088] The process for obtaining a set of candidate reinforcement solutions, for example, is as follows: First, based on factors such as the geological structure data, deformation conditions, and load conditions of the high-risk monitoring area, potentially applicable reinforcement material types are determined. For example, if the soil in the high-risk monitoring area is soft, rigid reinforcement materials such as cement-mixed piles and reinforced concrete piles can be selected. If the soil is relatively stable but its tensile strength needs to be increased, flexible reinforcement materials such as geogrids can be selected. Then, for each reinforcement material type, different deployment densities and structural forms are considered to create a variety of reinforcement solutions. For example, for geogrids, different spacings (such as 0.5 meters, 1 meter, and 1.5 meters) and deployment directions (such as horizontal, vertical, and diagonal) can be considered to create a variety of reinforcement solutions. Finally, all these reinforcement solutions are summarized to form a set of candidate reinforcement solutions. This set of candidate reinforcement solutions provides multiple options for subsequent simulation and screening.

[0089] Step S532: Input the parameters of each candidate reinforcement scheme into the deformation response model to calculate the predicted deformation and stress distribution uniformity index of the high-risk monitoring area after reinforcement.

[0090] The parameters of the candidate reinforcement schemes include the type of reinforcement material, layout density, structural form, and other specific parameters for each candidate reinforcement scheme. The predicted deformation is the predicted deformation of the high-risk monitoring area under different load scenarios after applying a candidate reinforcement scheme, reflecting the reinforcement scheme's effectiveness in suppressing deformation in the high-risk monitoring area. The stress distribution uniformity index is the uniformity of the stress distribution within the high-risk monitoring area after applying a candidate reinforcement scheme, reflecting the reinforcement scheme's effectiveness in improving the stress distribution in the high-risk monitoring area.

[0091] The process of inputting the parameters of each candidate reinforcement scheme into a deformation response model to calculate the predicted deformation and stress distribution uniformity index of the high-risk monitoring area after reinforcement is, for example, as follows: First, the parameters of each candidate reinforcement scheme are input into a calibrated deformation response model. Within the model, the boundary conditions and material parameters are adjusted accordingly based on the mechanical properties and layout of the reinforcement materials. Next, different load scenarios are input and the model's mechanical equilibrium equations are solved to obtain the deformation and stress distribution of the high-risk monitoring area at each time step after reinforcement. For the predicted deformation, indicators such as the maximum and average deformation of the high-risk monitoring area after reinforcement can be calculated. For the stress distribution uniformity index, indicators such as standard deviation and coefficient of variation can be used. For example, the standard deviation of stress values ​​at different locations within the high-risk monitoring area can be calculated. A smaller standard deviation indicates a more uniform stress distribution. Through these steps, the predicted deformation and stress distribution uniformity index of the high-risk monitoring area after reinforcement are calculated, providing a basis for subsequent scoring and screening.

[0092] Step S533: assigning a comprehensive suppression score to each candidate reinforcement scheme based on the difference between the predicted deformation and the preset safe deformation threshold, and the deviation between the stress distribution uniformity index and the ideal distribution.

[0093] The preset safety deformation threshold is a safety threshold for deformation in high-risk monitoring areas that is pre-set based on factors such as the design requirements and safety standards of the high-risk monitoring area. When the predicted deformation exceeds this threshold, it is considered that there is a safety hazard in the high-risk monitoring area. Ideal distribution is an ideal state of stress distribution, such as a state of uniform stress distribution. The deviation is the degree of difference between the stress distribution uniformity index and the ideal distribution, reflecting the degree of unevenness of the stress distribution. The comprehensive suppression score is a comprehensive score assigned to each candidate reinforcement scheme based on the difference between the predicted deformation and the preset safety deformation threshold, as well as the deviation between the stress distribution uniformity index and the ideal distribution. It reflects the comprehensive suppression effect of the reinforcement scheme on the deformation and stress distribution in the high-risk monitoring area.

[0094] Based on the difference between the predicted deformation and the preset safe deformation threshold, as well as the deviation between the stress distribution uniformity index and the ideal distribution, a comprehensive suppression score is assigned to each candidate reinforcement scheme. For example, the following process is used: First, the difference between the predicted deformation and the preset safe deformation threshold is calculated. If the predicted deformation is less than the preset safe deformation threshold, the difference is positive, indicating that the reinforcement scheme has a suppressive effect on deformation. If the predicted deformation is greater than the preset safe deformation threshold, the difference is negative, indicating that the reinforcement scheme may not meet safety requirements. Next, the deviation between the stress distribution uniformity index and the ideal distribution is calculated. For example, if the stress distribution uniformity index is measured using standard deviation, the deviation can be calculated as the difference between the standard deviation and the ideal standard deviation (e.g., a standard deviation of 0). Next, the difference and deviation are normalized to a suitable range, such as [0, 1]. Linear normalization can be used for normalization. Finally, a weighted fusion method is used to assign a comprehensive suppression score to each candidate reinforcement scheme based on the normalized values ​​of the difference and deviation.

[0095] Step S534: Sort all candidate reinforcement schemes based on the comprehensive suppression score, and select the candidate reinforcement scheme with the highest score as the optimal reinforcement scheme; wherein the comprehensive suppression score is obtained by weighted fusion of the normalized value of the difference and the deviation, and the weight coefficient is configured according to the warning level of the high-risk monitoring area.

[0096] Sorting is the process of arranging all candidate reinforcement schemes according to their comprehensive suppression scores. The optimal reinforcement scheme is the one with the highest comprehensive suppression score among all candidate reinforcement schemes, and this scheme has the best comprehensive suppression effect on deformation and stress distribution in the high-risk monitoring area. Weighted fusion is the process of multiplying the normalized values ​​of the difference and deviation by the corresponding weight coefficients, and then adding them to obtain the comprehensive suppression score. The weight coefficient is used to adjust the proportion of the difference and deviation in the comprehensive suppression score and is configured according to the warning level of the high-risk monitoring area.

[0097] The process of ranking all candidate reinforcement schemes based on their comprehensive inhibition scores and selecting the candidate with the highest score as the optimal reinforcement scheme is as follows: First, weighting coefficients for the difference and deviation are configured based on the warning level of the high-risk monitoring area. For example, for high-risk monitoring areas with a higher warning level, the difference between the predicted deformation and the preset safe deformation threshold is of greater importance, so the weighting coefficient for the difference can be set higher. For high-risk monitoring areas with a lower warning level, the weighting coefficient for the difference can be appropriately reduced, while the weighting coefficient for the deviation can be increased. Then, based on the configured weighting coefficients, a weighted fusion method is used to calculate the comprehensive inhibition score for each candidate reinforcement scheme. Next, all candidate reinforcement schemes are sorted from high to low according to their comprehensive inhibition scores. This can be achieved using sorting algorithms such as bubble sort and quick sort. For example, using bubble sort, the comprehensive inhibition scores of two adjacent candidate reinforcement schemes are compared. If they are in the wrong order, they are swapped, and this process is repeated until the entire sequence is in order. Finally, the candidate reinforcement scheme with the highest score after sorting is selected as the optimal reinforcement scheme. This optimal reinforcement scheme can suppress deformation and improve stress distribution to the greatest extent while taking into account the warning level of high-risk monitoring areas.

[0098] Step S540: Generate a soil roadbed deformation optimization strategy including reinforcement material configuration parameters, construction sequence planning, and resource allocation ratio based on the recommended response time and the optimal reinforcement plan.

[0099] The recommended response time is the timeframe for taking measures to address deformation risks, as recommended by deformation warning information. Reinforcement material configuration parameters are specific settings for reinforcement materials, including the preferred sequence of material types, the gradient of unit area usage, and interlayer layout rules. Construction sequence planning is the time scheduling and sequence planning for the reinforcement construction process, including the division of construction sub-areas, the determination of construction time windows, parallel execution strategies, and emergency adjustment mechanisms. The resource allocation ratio is the proportion of human, material, and financial resources allocated based on the deformation warning level of the high-risk monitoring area and the resource requirements of the remaining monitoring areas.

[0100] As an implementation method, the process of generating the above construction timing plan may include the following steps S54A1 to S54A5:

[0101] Step S54A1: Determine the supply cycle and transportation route of the required materials and equipment based on the reinforcement material configuration parameters.

[0102] Reinforcement material configuration parameters specify the type and quantity of materials required for reinforcement. The supply cycle is the time required from the procurement of materials and equipment to their arrival at the construction site, influenced by factors such as the supplier's production capacity, transportation distance, and transportation method. The transportation route is the route taken by which materials and equipment are transported from the supplier to the construction site.

[0103] The process for determining the supply cycle and transportation routes for required materials and equipment based on reinforcement material configuration parameters is as follows. First, determine the specific materials and equipment to be procured based on the material types and quantities specified in the reinforcement material configuration parameters. For example, if the reinforcement material configuration parameters specify the use of geogrids and cement mixing piles, then a list of the required geogrid specifications and quantities, as well as a list of the equipment required for cement mixing pile construction, is required. Next, contact multiple suppliers to understand their production capacity and delivery timelines. For some commonly used reinforcement materials and equipment, there may be multiple suppliers available. By communicating with these suppliers, you can obtain information on the supply cycles of different suppliers. Also, consider the impact of transportation distance and mode of transport on the supply cycle. If the supplier is far from the construction site, it may be necessary to select appropriate transportation methods, such as road or rail, to ensure the timely arrival of materials and equipment. When determining the transportation route, factors such as road conditions, traffic volume, and transportation costs should be comprehensively considered. Tools such as geographic information systems (GIS) can be used to analyze the advantages and disadvantages of different transportation routes and select the optimal one. For example, routes with good road conditions and low traffic volume can be selected to reduce transportation time and costs.

[0104] Step S54A2: Based on the distribution density and traffic accessibility of the high-risk monitoring area, multiple construction sub-areas are divided and construction priorities are assigned.

[0105] The distribution density of high-risk monitoring areas refers to how densely they are distributed within the target soil roadbed. Accessibility refers to the ease with which construction vehicles and equipment can reach the high-risk monitoring areas, influenced by factors such as road conditions and topography. Construction sub-areas are defined as smaller areas divided into high-risk monitoring areas to facilitate construction management. Construction priority is the order in which construction work is assigned to each construction sub-area.

[0106] The process for dividing high-risk monitoring areas into multiple construction sub-areas and assigning construction priorities is as follows. First, using tools such as geographic information systems (GIS), the distribution of high-risk monitoring areas is analyzed. Traffic maps and topographic data are combined to understand the accessibility of each area. Then, based on the distribution density and accessibility, cluster analysis and other methods are used to divide the high-risk monitoring areas into multiple construction sub-areas. For example, if the high-risk monitoring areas are concentrated and have convenient transportation, they can be divided into one or several larger construction sub-areas. If they are more dispersed and have significantly varying traffic conditions, they should be divided into multiple smaller construction sub-areas. When dividing construction sub-areas, consideration should be given to construction convenience and efficiency, ensuring that construction tasks within each sub-area are relatively independent. Next, construction priorities are assigned based on factors such as the importance of each construction sub-area, its degree of deformation risk, and its accessibility. Construction should be prioritized for construction sub-areas located below major traffic arteries, with higher deformation risk and better accessibility. Construction can be postponed for construction sub-areas with lower deformation risk and inconvenient transportation.

[0107] Step S54A3: Based on the resource allocation ratio and supply cycle, a construction time window and human resource scheduling plan are configured for each construction sub-area.

[0108] The resource allocation ratio determines the amount of human, material, and financial resources available to each construction sub-area. The supply cycle determines the time it takes for materials and equipment to arrive at the construction site. The construction time window is the time frame within which construction can take place in each construction sub-area, taking into account factors such as weather and the surrounding environment. The human resource scheduling plan arranges construction personnel for different construction sub-areas and during different construction times.

[0109] The process for configuring a construction time window and human resource scheduling plan for each construction sub-area, combining resource allocation ratios and supply cycles, is as follows. First, based on the resource allocation ratio, the available human, material, and financial resources for each construction sub-area are determined. For example, construction personnel, reinforcement materials, and construction equipment are allocated to a construction sub-area based on the resource allocation ratio. Then, considering the supply cycle, the arrival time of materials and equipment at the construction site is determined. Based on this time, combined with factors such as weather and the surrounding environment, a construction time window is determined for each construction sub-area. For example, if materials and equipment are expected to arrive within a certain time period, and the weather and surrounding environment allow construction during that time period, then this time period can be designated as the construction time window. Next, a human resource scheduling plan is developed based on the construction time window and the workload of the construction task. For construction sub-areas with a large workload and a shorter construction time window, the number of construction personnel should be appropriately increased and assigned to different construction phases to ensure timely completion of the construction task. Furthermore, the skill level and work experience of construction personnel should be considered and their appropriate assignments should be made to different construction positions.

[0110] Step S54A4: Optimize the parallel execution strategy of the human resource scheduling plan based on the overlap of construction time windows and equipment reuse requirements.

[0111] The overlap of construction time windows refers to whether the construction time windows of different construction sub-areas overlap. Equipment reuse requirements refer to the need to use some shared equipment during construction in different construction sub-areas. The parallel execution strategy refers to how to arrange construction personnel to work in parallel in different construction sub-areas in the human resource scheduling plan to improve construction efficiency.

[0112] The process for optimizing the parallel execution strategy of the human resource scheduling plan based on the overlap of construction time windows and the need for equipment reuse is as follows. First, analyze the overlap of construction time windows. If the construction time windows of multiple construction sub-areas overlap, parallel construction can be considered. Next, consider the need for equipment reuse. For shared equipment, such as cranes and mixers, the order of their use in different construction sub-areas should be rationally arranged to avoid equipment idleness and conflicts. For example, if two construction sub-areas require the use of a crane during the same time period, the crane can be assigned to which sub-area first based on the urgency of the construction task and the workload. Next, adjust the human resource scheduling plan based on the overlap of construction time windows and the need for equipment reuse. For construction sub-areas with parallel construction, rationally allocate construction personnel to ensure that each construction sub-area has sufficient manpower for construction. Furthermore, establish an effective communication mechanism to ensure timely communication and collaboration between construction personnel. For example, a site dispatcher can be assigned to coordinate the construction progress and equipment usage of different construction sub-areas.

[0113] Step S54A5: Generate a construction timing plan including construction sub-areas, construction time windows, parallel execution strategies, and emergency adjustment mechanisms.

[0114] The construction sub-area clarifies the specific scope of construction, the construction time window stipulates the construction time of each construction sub-area, the parallel execution strategy improves construction efficiency, and the emergency adjustment mechanism provides a guarantee for dealing with emergencies during the construction process.

[0115] The process for generating a construction schedule that includes construction sub-areas, construction time windows, parallel execution strategies, and an emergency adjustment mechanism is as follows. First, the divided construction sub-areas, determined construction time windows, and optimized parallel execution strategies are integrated to form a preliminary construction schedule. Next, an emergency adjustment mechanism is developed. This emergency adjustment mechanism should account for various possible emergencies, such as severe weather, equipment failure, and casualties. For each emergency, a corresponding response plan should be developed. For example, in the event of severe weather, construction should be suspended, and measures should be taken to protect the completed construction section and construction equipment. If equipment fails, maintenance personnel should be promptly dispatched for repairs, or backup equipment should be deployed. Furthermore, an emergency response process should be established to ensure that emergency measures can be rapidly implemented in the event of an emergency. Finally, the construction schedule is reviewed and refined to ensure its rationality and feasibility. During the review process, factors such as construction safety, quality, and cost should be considered. Through these steps, a construction schedule that includes construction sub-areas, construction time windows, parallel execution strategies, and an emergency adjustment mechanism is generated.

[0116] As an embodiment, the above process of generating the reinforcement material configuration parameters may include the following steps S54B1 to S54B6:

[0117] Step S54B1: Determine a set of candidate reinforcement material types based on the material type requirements of the optimal reinforcement solution.

[0118] The optimal reinforcement solution specifies the type requirements of the required reinforcement materials, and the candidate reinforcement material type set is the set of all possible reinforcement material types that meet these requirements.

[0119] The process of determining the set of candidate reinforcement material types based on the material type requirements of the optimal reinforcement solution is as follows. First, carefully analyze the specific requirements for the reinforcement material type in the optimal reinforcement solution, including the mechanical properties, chemical properties, durability and other aspects of the material. For example, if the optimal reinforcement solution requires the reinforcement material to have high tensile strength and good flexibility, then materials that meet these requirements can be screened out from common reinforcement materials. Common reinforcement materials include geogrids, geotextiles, cement mixing piles, reinforced concrete piles, etc. Then, by consulting relevant information, consulting experts, and conducting market research, etc., reinforcement materials that meet the material type requirements of the optimal reinforcement solution are collected. For each reinforcement material, it is necessary to understand its performance characteristics, scope of application, price and other information. Finally, all the collected reinforcement material types that meet the requirements are summarized to form a set of candidate reinforcement material types.

[0120] Step S54B2: Perform historical performance parameter analysis on each material in the candidate reinforcement material type set, and extract a set of key performance parameters associated with the stress distribution uniformity index of the high-risk monitoring area.

[0121] Historical performance parameter analysis involves collecting, organizing, and analyzing performance data from previous engineering applications for each candidate reinforcement material type. The stress uniformity index measures the uniformity of stress distribution within high-risk monitoring areas. The key performance parameter set is a collection of reinforcement material performance parameters that are closely related to the stress uniformity index.

[0122] The process of analyzing and processing the historical performance parameters of each material in the set of candidate reinforcement material types and extracting a set of key performance parameters is as follows. First, the performance data of each candidate reinforcement material in previous engineering applications is collected, including parameters such as the material's strength, elastic modulus, Poisson's ratio, and durability. These data can be obtained from channels such as engineering reports, academic papers, and laboratory test reports. Then, the collected performance data is sorted and analyzed to establish a performance database for each material. Next, the influencing factors of the stress distribution uniformity index in the high-risk monitoring area are analyzed, and the performance parameters of the reinforcement material associated with this index are determined. For example, the elastic modulus and Poisson's ratio of the material will affect the deformation of the material when subjected to stress, and thus affect the stress distribution uniformity in the high-risk monitoring area. Therefore, the elastic modulus and Poisson's ratio can be used as key performance parameters. Finally, the key performance parameters associated with the stress distribution uniformity index are extracted from the performance database of each material to form a set of key performance parameters.

[0123] Step S54B3: Based on the mapping relationship between the key performance parameter set and the predicted deformation variable of the deformation response model, dynamic load simulation processing is performed on each candidate reinforcement material to generate simulated performance data of different materials in the high-risk monitoring area.

[0124] The mapping relationship is a functional or statistical relationship between a set of key performance parameters and the predicted deformation variables of the deformation response model. Dynamic load simulation involves computer simulation of the performance of each candidate reinforcement material under dynamic loads in the high-risk monitoring area. Simulated performance data, such as deformation and stress distribution, is obtained through dynamic load simulation for each candidate reinforcement material in the high-risk monitoring area.

[0125] Based on the mapping relationship between the key performance parameter set and the predicted deformation variable of the deformation response model, dynamic load simulation is performed on each candidate reinforcement material to generate simulated performance data as follows. First, a mapping relationship is established between the key performance parameter set and the predicted deformation variable of the deformation response model. This mapping relationship can be established through methods such as experimental data fitting and machine learning algorithms. For example, a multivariate linear regression algorithm is used to establish a regression equation with the key performance parameters as independent variables and the predicted deformation variable of the deformation response model as the dependent variable. Then, dynamic load simulation is performed on each candidate reinforcement material using tools such as finite element analysis software. During the simulation process, the key performance parameters are used as input parameters, and dynamic loads similar to the actual situation in the high-risk monitoring area are applied to solve the mechanical equilibrium equation of the model to obtain simulated performance data such as deformation and stress distribution of each material in the high-risk monitoring area. Finally, the simulated performance data are sorted and analyzed to provide a basis for subsequent material screening.

[0126] Step S54B4: Based on the coupling correlation between the material compressive attenuation rate and the environmental adaptability coefficient in the simulated performance data, a subset of candidate materials that meet the preset durability conditions is screened.

[0127] The material compressive decay rate is the rate at which the material's compressive strength decays under long-term stress, reflecting the material's durability. The environmental adaptability coefficient is the coefficient of variation in a material's performance under different environmental conditions (such as temperature, humidity, and pH), reflecting the material's adaptability to the environment. The coupling correlation is the relationship between the material compressive decay rate and the environmental adaptability coefficient. Preset durability conditions are pre-set conditions for material durability based on the actual conditions and project requirements of the high-risk monitoring area. The candidate material subset is a collection of material types selected from the set of candidate reinforcement material types that meet the preset durability conditions.

[0128] The process of screening a subset of candidate materials that meet the preset durability conditions based on the coupling correlation between the material compressive decay rate and the environmental adaptability coefficient in the simulated performance data is as follows. First, the compressive decay rate and environmental adaptability coefficient of each candidate reinforcement material are extracted from the simulated performance data. Then, the coupling correlation between the material compressive decay rate and the environmental adaptability coefficient is analyzed. This correlation can be analyzed by establishing a mathematical model, drawing a scatter plot, etc. For example, by establishing a regression model, the linear relationship between the compressive decay rate and the environmental adaptability coefficient is analyzed. Then, according to the preset durability conditions, the screening criteria are determined. The preset durability conditions may include the upper limit of the compressive decay rate, the lower limit of the environmental adaptability coefficient, etc. Finally, according to the screening criteria, a subset of candidate materials that meet the preset durability conditions are screened from the set of candidate reinforcement material types. Materials that do not meet the conditions are excluded.

[0129] Step S54B5: Perform long-term stability prediction processing on each material in the candidate material subset, combine the soil stratification data and moisture content distribution data of the high-risk monitoring area, and calculate the matching degree between the material penetration resistance and the interlayer shear strength.

[0130] Long-term stability prediction involves predicting the performance changes of each material in a subset of candidate materials over the long term. Material penetration resistance is the material's ability to resist the penetration of liquids or gases. Interlaminar shear strength is the shear strength between adjacent soil layers. The degree of fit between material penetration resistance and interlaminar shear strength reflects the material's suitability for use in soil layers within high-risk monitoring areas.

[0131] The process for predicting the long-term stability of each material in the candidate material subset and calculating the matching degree between the material's penetration resistance and interlaminar shear strength is as follows. First, the long-term stability of each material in the candidate material subset is predicted using methods such as accelerated aging tests and numerical simulations. Accelerated aging tests simulate the environmental conditions experienced by a material during long-term use, such as high temperature, high humidity, and UV exposure, accelerating the material's aging process and thereby predicting long-term performance changes. Numerical simulations use computer software to build a material aging model and, by inputting relevant parameters, predict performance changes over long-term use. Next, the relationship between the material's penetration resistance and interlaminar shear strength is analyzed by combining soil stratification data and moisture content distribution data from high-risk monitoring areas. For example, soil moisture content can affect both the material's penetration resistance and interlaminar shear strength; higher moisture content may reduce both. Next, the matching degree between the material's penetration resistance and interlaminar shear strength is calculated. Methods such as normalization and weighted averaging can be used to calculate the matching degree. For example, the material penetration resistance and interlayer shear strength are normalized separately, and then different weights are assigned according to their influence on the performance of the material in the soil layer, and the weighted average is calculated as the matching degree.

[0132] Step S54B6: Based on the constraints of matching degree and resource allocation ratio, generate reinforcement material configuration parameters including material type optimization sequence, unit area usage gradient and inter-layer layout rules.

[0133] The matching degree reflects the suitability of a material for the soil layers in the high-risk monitoring area, and the resource allocation ratio specifies the amount of resources that can be used for each material. The preferred material type sequence is the sequence obtained by sorting the materials in the candidate material subset according to the matching degree, with the materials with the highest ranking being given priority. The unit area usage gradient shows the variation in the amount of reinforcement material per unit area at different locations in the high-risk monitoring area. The interlayer layout rule describes the method and order in which the reinforcement material is laid across different soil layers.

[0134] The process for generating reinforcement material configuration parameters based on the constraints of matching degree and resource allocation ratio is as follows. First, the materials in the candidate material subset are ranked according to matching degree to form a preferred material type sequence. Materials with higher matching degrees are ranked higher in the sequence. Then, the unit area usage gradient is determined based on the resource allocation ratio and the actual conditions of the high-risk monitoring area. For materials with higher matching degrees, the unit area usage can be appropriately increased, if resources permit; for materials with lower matching degrees, the usage can be reduced. Furthermore, considering the degree of deformation and stress distribution at different locations in the high-risk monitoring area, the unit area usage of reinforcement material is increased in areas with larger deformation and concentrated stress, while the usage is reduced in areas with smaller deformation and relatively uniform stress. Next, interlayer layout rules are developed based on the material performance characteristics and soil stratification data for the high-risk monitoring area. For example, geogrids with greater flexibility can be laid in the upper soil layer to increase its tensile strength; while cement-mixed piles with higher rigidity can be placed deeper into the soil layer to improve its bearing capacity. Finally, the preferred sequence of material types, unit area usage gradient and interlayer layout rules are sorted out to form the reinforcement material configuration parameters.

[0135] As an embodiment, the above-mentioned process of generating the reinforcement material configuration parameters includes the following steps S54C1 to S54C5:

[0136] Step S54C1: Determine a set of candidate reinforcement material types based on the material type requirements of the optimal reinforcement solution.

[0137] This step is similar in principle and implementation to step S54B1. Based on the specific requirements of the optimal reinforcement solution for the type of reinforcement material, reinforcement materials that meet the requirements are collected through reference, expert consultation, and market research to form a set of candidate reinforcement material types.

[0138] Step S54C2: Perform historical performance parameter analysis on each material in the candidate reinforcement material type set, and extract a set of key performance parameters associated with the stress distribution uniformity index of the high-risk monitoring area.

[0139] This step follows the same principles and implementation as step S54B2. The historical performance data for each material in the candidate reinforcement material set is collected, organized, and analyzed to establish a performance database. Then, based on the factors influencing the stress distribution uniformity index in the high-risk monitoring area, key performance parameters associated with that index are extracted to form a key performance parameter set.

[0140] Step S54C3: Based on the mapping relationship between the key performance parameter set and the predicted deformation variable of the deformation response model, dynamic load simulation processing is performed on each candidate reinforcement material to generate simulated performance data of different materials in the high-risk monitoring area.

[0141] This step follows the same principles and implementation as step S54B3. A mapping relationship is established between the set of key performance parameters and the predicted deformation variables of the deformation response model. Finite element analysis software and other tools are used to perform dynamic load simulations on each candidate reinforcement material, generating simulated performance data for different materials in high-risk monitoring areas, such as deformation and stress distribution.

[0142] Step S54C4: Based on the coupling correlation between the material compressive attenuation rate and the environmental adaptability coefficient in the simulated performance data, a subset of candidate materials that meet the preset durability conditions is screened.

[0143] This step follows the same principles and implementation as step S54B4. The compressive decay rate and environmental adaptability coefficient for each candidate reinforcement material are extracted from the simulated performance data. The coupling correlation between these factors is analyzed. Based on the preset durability conditions, screening criteria are determined, and a subset of candidate materials that meet the criteria are selected from the set of candidate reinforcement material types.

[0144] Step S54C5: Perform long-term stability prediction processing on each material in the candidate material subset, combine the soil stratification data and moisture content distribution data of the high-risk monitoring area, and calculate the matching degree between the material penetration resistance and the interlayer shear strength.

[0145] This step follows the same principles and implementation as step S54B5. Accelerated aging tests and numerical simulations are used to predict the long-term stability of each material in the candidate material subset. Combined with soil stratification data and moisture content distribution data from high-risk monitoring areas, the relationship between material penetration resistance and interlaminar shear strength is analyzed, and the degree of compatibility between them is calculated.

[0146] Step S54C6: Based on the constraints of the matching degree and the resource allocation ratio, generate reinforcement material configuration parameters including the material type optimization sequence, unit area usage gradient, and interlayer layout rules.

[0147] This step shares the same principles and implementation as step S54B6. The materials in the candidate material subset are sorted based on their matching degree to form a preferred material type sequence. The unit area usage gradient is determined based on the resource allocation ratio and the actual conditions of the high-risk monitoring area. Interlayer layout rules are developed based on material performance characteristics and soil stratification data, ultimately generating reinforcement material configuration parameters.

[0148] As an implementation manner, the method provided in the embodiment of the present invention may further include the following steps S600 to S1000:

[0149] Step S600: periodically acquiring an updated distributed optical fiber sensing data set of the target soil roadbed.

[0150] Periodic acquisition involves collecting distributed fiber optic sensing data of the target soil roadbed at intervals. The updated distributed fiber optic sensing data set is the distributed fiber optic sensing data set of the target soil roadbed collected at the current time point, including the latest fiber optic strain data sequence and temperature compensation data sequence.

[0151] The process of periodically acquiring updated distributed fiber optic sensing data sets for the target soil roadbed can be achieved in the following manner. First, the data collection cycle is determined based on the actual conditions of the target soil roadbed and monitoring requirements. For example, for soil roadbeds with rapidly changing deformation, a shorter collection cycle, such as daily, can be set; for soil roadbeds with relatively stable deformation, a longer collection cycle, such as weekly, can be set. Then, using the distributed fiber optic sensors and fiber optic sensor demodulators pre-installed in the target soil roadbed, data is collected according to the set collection cycle. During the collection process, the accuracy and integrity of the data are ensured. For example, the collected data is verified in real time to promptly identify and correct errors during the data collection process. Finally, the collected, updated distributed fiber optic sensing data sets are stored and managed for subsequent analysis and processing.

[0152] Step S700: regenerate a multi-dimensional deformation feature set of the monitoring area according to the updated distributed optical fiber sensing data set.

[0153] The principle and implementation process of this step are similar to those of step S300. According to the optical fiber strain data sequence and temperature compensation data sequence in the updated distributed optical fiber sensing data set, a coupling analysis process is performed on them to generate a multi-dimensional deformation feature set of the monitoring area. The specific process includes performing noise filtering and baseline calibration on the optical fiber strain data sequence to obtain a denoised standardized strain data sequence; constructing a temperature strain compensation coefficient matrix according to the temperature change trend of the temperature compensation data sequence, performing temperature drift correction on the standardized strain data sequence, and generating a corrected target strain data sequence; performing a time-frequency domain joint analysis on the target strain data sequence to extract the time domain deformation accumulation and frequency domain disturbance response intensity of the monitoring area; calculating the stress accumulation correlation characteristic value of the monitoring area based on the nonlinear correlation between the time domain deformation accumulation and the frequency domain disturbance response intensity; fusing the time domain deformation accumulation, the frequency domain disturbance response intensity and the stress accumulation correlation characteristic value to generate a multi-dimensional deformation feature set of the monitoring area.

[0154] Step S800: comparing the difference between the newly generated multi-dimensional deformation feature set and the historical multi-dimensional deformation feature set, and extracting the spatiotemporal propagation path features of the difference.

[0155] The historical multidimensional deformation feature set is the multidimensional deformation feature set of the monitoring area generated during previous monitoring processes. The difference is the difference between the newly generated multidimensional deformation feature set and the historical multidimensional deformation feature set, reflecting the changes in deformation in the monitoring area at different time points. The spatiotemporal propagation path characteristics are the propagation patterns and characteristics of the difference across time and space.

[0156] The process for comparing the differences between the regenerated multidimensional deformation feature set and the historical multidimensional deformation feature set and extracting the spatiotemporal propagation path characteristics of the differences is as follows. First, the regenerated multidimensional deformation feature set is subtracted element-by-element from the historical multidimensional deformation feature set to obtain the differences. For example, for the deformation rate characteristics, stress accumulation correlation characteristics, and environmental perturbation response characteristics in the multidimensional deformation feature set, the differences between the regenerated set and the historical set are calculated. Then, spatial analysis techniques and time series analysis methods are used to analyze the differences and extract their spatiotemporal propagation path characteristics. For example, geographic information system (GIS) technology is used to visualize the differences spatially and observe their propagation across different monitoring areas. Time series cluster analysis methods are used to analyze the temporal variation of the differences and determine characteristics such as the starting time and propagation speed of the differences.

[0157] Step S900: Determine the deformation source region and the affected region chain according to the spatiotemporal propagation path characteristics, and update the dynamic matching rule of the preset deformation threshold condition based on the coupling attenuation coefficient of the deformation rate characteristics of the deformation source region and the stress accumulation correlation characteristics of the affected region chain.

[0158] The deformation source region is the area within the soil roadbed where deformation begins, representing the starting point of soil roadbed deformation. The affected region chain is a collection of monitoring areas affected by the propagation of deformation from the deformation source region, arranged in the order of deformation propagation. The coupling attenuation coefficient is the attenuation relationship coefficient between the deformation rate characteristics of the deformation source region and the stress accumulation correlation characteristics of the affected region chain, reflecting the degree of deformation attenuation during propagation. The dynamic matching rule for the preset deformation threshold condition is used to determine whether the deformation of the monitoring area exceeds the preset threshold. This rule is dynamically updated based on the actual deformation of the soil roadbed.

[0159] The process for determining the deformation source region and affected region chain based on spatiotemporal propagation path characteristics and updating the dynamic matching rules for the preset deformation threshold conditions is as follows. First, based on the spatiotemporal propagation path characteristics, the deformation source region is determined through retrospective analysis. For example, the monitoring region that first shows a significant difference along the difference propagation path is identified as the deformation source region. Next, the affected region chain is determined along the difference propagation path. Affected monitoring regions can be sequentially determined based on the spatial adjacency between monitoring regions and the propagation direction of the difference, and these regions are arranged in order of propagation to form an affected region chain. Next, the relationship between the deformation rate characteristics of the deformation source region and the stress accumulation correlation characteristics of the affected region chain is analyzed to calculate the coupling attenuation coefficient. Methods such as regression analysis can be used to establish a regression model between the deformation rate characteristics and the stress accumulation correlation characteristics, and the coupling attenuation coefficient is determined based on the model parameters. Finally, the dynamic matching rules for the preset deformation threshold conditions are updated based on the coupling attenuation coefficient and the actual conditions of the deformation source region and the affected region chain. For example, for monitoring areas in the affected area chain that are closer to the deformation source area, the preset deformation threshold conditions are appropriately lowered; for monitoring areas that are farther away, the preset deformation threshold conditions are appropriately increased according to the coupling attenuation coefficient.

[0160] As an embodiment, step S900, determining the deformation source region and the affected region chain according to the spatiotemporal propagation path characteristics, may specifically include the following steps S910 to S960:

[0161] Step S910: performing time series cluster analysis on the difference values ​​to identify a set of monitoring areas having continuous deformation conduction characteristics.

[0162] Time series cluster analysis groups variances according to time series, ensuring that variances within the same group have similar trends and characteristics. Continuous deformation conduction refers to the continuous propagation and influence of deformations between monitoring areas. A monitoring area set is a collection of monitoring areas that exhibit continuous deformation conduction.

[0163] The process for performing time series cluster analysis on the difference values ​​to identify sets of monitoring areas with continuous deformation conduction characteristics is as follows. First, the difference values ​​are arranged by monitoring area and time, forming multiple time series data. Each time series data set represents the temporal evolution of the difference values ​​for a monitoring area. Next, an appropriate clustering algorithm, such as K-means clustering or hierarchical clustering, is selected to perform cluster analysis on these time series data. Taking the K-means clustering algorithm as an example, the number of clusters K is first determined. K initial cluster centers are then randomly selected. Each time series data point is assigned to the cluster with the closest cluster center. The cluster centers are then updated, and the assignment and updating process is repeated until the cluster centers no longer change. Finally, based on the clustering results, sets of monitoring areas with continuous deformation conduction characteristics are identified. Spatially adjacent monitoring areas belonging to the same cluster are considered to have continuous deformation conduction characteristics and are grouped as a single set of monitoring areas.

[0164] Step S920: constructing a deformation conduction topology network according to the deformation rate change direction of the monitoring area set and the transfer polarity of the stress accumulation correlation characteristics.

[0165] The direction of deformation rate change is the temporal trend of the deformation rate in the monitored area, for example, whether the deformation rate is increasing or decreasing. The transfer polarity of the stress accumulation correlation feature is the direction and positive / negative nature of the stress accumulation correlation feature's transfer between monitored areas. The deformation conduction topology network is used to describe the deformation conduction relationship between monitored areas. Nodes in the network represent monitored areas, and edges represent the deformation conduction relationship between monitored areas.

[0166] The process of constructing a deformation conduction topological network based on the direction of deformation rate change and the transfer polarity of stress accumulation correlation characteristics of a set of monitoring areas is as follows. First, the direction of deformation rate change and the transfer polarity of stress accumulation correlation characteristics of the monitoring areas in each monitoring area set are analyzed. The direction of deformation rate change can be determined by calculating the first-order derivative of the deformation rate, and the transfer polarity is determined by analyzing the difference in stress accumulation correlation characteristics between adjacent monitoring areas. Then, based on the direction of deformation rate change and transfer polarity, the deformation conduction relationship between monitoring areas is determined. If the deformation rate of a monitoring area increases and its stress accumulation correlation characteristics are transferred to adjacent monitoring areas, it can be considered that a deformation conduction relationship exists from this monitoring area to the adjacent monitoring area. Finally, the deformation conduction topological network is constructed, using the monitoring areas as nodes and the deformation conduction relationships as edges. In the network, the direction of deformation conduction can be represented by directed edges, and the strength of the conduction can be represented by edge weights.

[0167] Step S930: Backtrack and determine the initial conduction node in the deformation source area through the node connection strength and directional weight in the deformation conduction topology network.

[0168] Node connection strength is the tightness of connections between nodes in the deformation conduction topology network, reflecting the strength of deformation conduction between monitoring areas. Directional weight is the weight of directed edges in the deformation conduction topology network, indicating the direction and strength of deformation conduction. The initial conduction node is the starting node of the deformation source region in the deformation conduction topology network, where deformation propagation begins.

[0169] The process of backtracking to determine the initial transmission node in the deformation source region using the node connection strength and directional weight in the deformation transmission topology network is as follows. First, one or more nodes with large in-degree (i.e., a large number of edges pointing to them) are selected from the deformation transmission topology network as candidate nodes. Nodes with large in-degrees are likely to be the source nodes of the deformation. Then, based on the node connection strength and directional weight, backtracking analysis is performed starting from the candidate nodes. The connection relationships between nodes are traced back step by step along the reverse direction of the directed edges, and a backtracking score is calculated for each node. The backtracking score can be calculated based on the node connection strength and directional weight. For example, the product of the node connection strength and directional weight is used as the backtracking score. Finally, the node with the highest backtracking score is selected as the initial transmission node in the deformation source region. This node is the node from which the deformation propagation begins.

[0170] Step S940: generating a conduction path map including an affected area chain based on the spatial distribution relationship between the initial conduction node and the set of monitoring areas.

[0171] The spatial distribution relationship represents the spatial relationship between the initial conduction node and the other monitoring areas in the monitoring area set. The conduction path map is a diagram that shows the path of deformation propagation from the initial conduction node to the chain of affected areas. It can intuitively display the direction of deformation propagation and the affected areas.

[0172] The process for generating a transmission path map containing a chain of affected regions based on the spatial distribution relationship between the initial transmission node and the set of monitoring areas is as follows. First, the corresponding monitoring area is determined spatially based on the initial transmission node's location in the deformation transmission topology network. Then, starting from this monitoring area, the affected monitoring areas are sequentially identified along the directed edges in the deformation transmission topology network, forming a chain of affected regions. Next, using tools such as a geographic information system (GIS), the initial transmission node and the chain of affected regions are marked on a map. These monitoring areas are then connected by lines based on the direction of deformation transmission to generate a transmission path map. In the map, different colors or line styles can be used to represent different intensities of deformation transmission, providing a more intuitive representation of the propagation of deformation.

[0173] Step S950: Divide the deformation source region and affected region chains of different levels according to the node attenuation gradient and the inter-region coupling strength in the conduction path map.

[0174] The node attenuation gradient is the degree to which node-related characteristics (such as deformation rate and stress accumulation correlation characteristics) decay as deformation propagates from the initial conduction node to the affected region chain in the conduction path map. The inter-region coupling strength is the degree of mutual influence between monitored regions, reflecting the ease with which deformation propagates between regions.

[0175] The process for demarcating deformation source regions and affected region chains of varying levels based on the node attenuation gradients and inter-regional coupling strengths in the conduction path map is as follows. First, the changes in node-related characteristics in the conduction path map with propagation distance are analyzed to calculate the node attenuation gradients. For example, the attenuation ratio of deformation rate and stress accumulation correlation characteristics between adjacent nodes is calculated. Then, based on the inter-regional coupling strength, the difficulty of deformation propagation between different monitoring areas is determined. The inter-regional coupling strength can be determined by analyzing factors such as the geological conditions and topography between the monitoring areas. Next, a demarcation criterion is established based on the node attenuation gradients and inter-regional coupling strengths. For example, when the node attenuation gradient reaches a set threshold or the inter-regional coupling strength is weak, the region containing that node is demarcated into affected region chains of varying levels. Finally, based on the demarcation criterion, the deformation source regions and affected region chains of varying levels are labeled and demarcated in the conduction path map.

[0176] Step S960: The division result is used to configure the blocking reinforcement sequence for the conduction path and the cross-regional resource scheduling path in the construction timing planning.

[0177] The delineation results clearly define deformation source regions and chains of affected regions at different levels. This provides important guidance for determining the blocking reinforcement sequence and cross-regional resource scheduling during construction timeline planning. The blocking reinforcement sequence involves reinforcing the monitored areas along the transmission path in order to block the propagation of deformation during soil roadbed reinforcement construction. The cross-regional resource scheduling path involves allocating human, material, and financial resources between different monitoring areas.

[0178] The process for configuring the blocking reinforcement sequence and cross-regional resource scheduling for transmission paths in construction sequence planning is as follows. First, based on the demarcation results, the reinforcement priorities for the deformation source area and the chains of affected areas at different levels are determined. The deformation source area is the starting point of deformation and requires priority reinforcement to prevent further deformation propagation. For the chains of affected areas at different levels, the reinforcement sequence is determined based on their distance from the deformation source area and the degree of impact. Areas closer to the deformation source area and with greater impact receive priority reinforcement. Then, based on the reinforcement sequence, a blocking reinforcement plan for the transmission path is developed. For example, reinforcement measures such as wall reinforcement and support reinforcement can be implemented at key nodes along the transmission path to block the propagation of deformation. Next, cross-regional resource scheduling is considered. Based on the resource needs and distribution of different monitoring areas, human, material, and financial resources are rationally allocated. For example, for a chain of affected areas with limited resources, resources can be allocated from other areas with abundant resources. When allocating resources, factors such as transportation costs and time should be considered to select the optimal cross-regional resource scheduling path. Finally, the blocking reinforcement sequence and cross-regional resource scheduling path are incorporated into the construction timing planning to ensure that the construction process can effectively block the propagation of deformation and improve the stability of the earth roadbed.

[0179] Step S1000: regenerate deformation warning information of the target soil roadbed according to the updated dynamic matching rules, and adjust the resource allocation ratio and construction timing planning in the soil roadbed deformation optimization strategy based on the priority of the deformation source area.

[0180] The updated dynamic matching rules are used to determine whether the deformation in a monitored area exceeds a preset threshold, based on factors such as the coupling attenuation coefficient between the deformation rate characteristics of the deformation source area and the stress accumulation correlation characteristics of the affected area chain. Deformation warning information is provided regarding the deformation of the target soil roadbed, including warning trigger conditions, risk area location identification, and recommended response timelines. The resource allocation ratio is the distribution ratio of human, material, and financial resources across different monitoring areas within the soil roadbed deformation optimization strategy. Construction sequence planning is the timing and sequence planning for the reinforcement construction process.

[0181] The process for regenerating deformation warning information for the target soil roadbed based on the updated dynamic matching rules and adjusting the resource allocation ratio and construction schedule in the soil roadbed deformation optimization strategy is as follows. First, the regenerated multi-dimensional deformation feature set of the monitoring area is spatiotemporally matched with the updated dynamic matching rules. Based on the matching results, the warning status of each monitoring area is redefined, including whether an alert is triggered and the alert level. Then, based on the redefined alert status, deformation warning information for the target soil roadbed is generated. The alert triggering conditions are determined based on the updated dynamic matching rules. The risk area location can be marked on a map using tools such as a geographic information system (GIS). The recommended response time is determined based on the alert level and the actual situation in the monitoring area. Next, the resource allocation ratio in the soil roadbed deformation optimization strategy is adjusted based on the priority of the deformation source area. The deformation source area is a key area for soil roadbed deformation and requires priority in resource supply. Therefore, the resource allocation ratio to the deformation source area is increased, such as by increasing the number of construction personnel and deploying more reinforcement materials and equipment. For the affected area chain, resource allocation ratios are rationally adjusted based on the degree of impact and warning level. Finally, construction schedules are adjusted based on the adjusted resource allocation ratios and deformation warning information. Reinforcement work in the deformation source area is prioritized to ensure timely control of deformation progression. For the affected area chain, construction time and sequence are rationally arranged based on reinforcement priorities and resource availability. At the same time, cross-regional resource scheduling paths are considered to ensure smooth construction. Through these steps, dynamic monitoring and optimization of the target soil roadbed are achieved, improving its stability and safety.

[0182] It is understandable that the various algorithms involved in the above-mentioned introductions of the embodiments of the present invention, such as filtering algorithms, differential methods, etc., can be learned from the relevant content in the prior art. In order to save space, they will not be expanded too much in the embodiments of this application. In addition, when implementing the scheme of this application, those skilled in the art can supplement the details according to the common knowledge in this field. For example, according to the common knowledge in this field, normalization can be used to eliminate dimensional conflicts before feature fusion, interpolation can be used to eliminate dimensional differences, and historical data, experience or business scenario requirements can be combined to reasonably set thresholds, and models can be trained based on general model training methods, etc. This application will no longer provide redundant introductions to the overly detailed implementation process.

[0183] See also Figure 2 , Figure 2This is a schematic diagram of the structure of a monitoring system provided in an embodiment of the present invention. The monitoring system can be a backend computer system, such as a server, and includes at least a processor 101, a communication interface 102, and a memory 103. The processor 101, communication interface 102, and memory 103 can be connected via a bus or other means. The processor 101 (also known as the Central Processing Unit (CPU)) is the computing and control core of the monitoring system, capable of parsing various instructions within the monitoring system and processing various types of data within the monitoring system. The communication interface 102 can optionally include a standard wired interface or a wireless interface (such as Wi-Fi or a mobile communication interface), and can be used to send and receive data under the control of the processor 101. The communication interface 102 can also be used for data transmission and interaction within the monitoring system. The memory 103 is a storage device within the monitoring system, used to store programs and data. It is understood that the memory 103 herein can include both the monitoring system's built-in memory and, of course, the extended memory supported by the monitoring system. The memory 103 provides storage space, which stores the monitoring system's operating system, but this is not limited to this disclosure. In one embodiment, the processor 101 executes the soil roadbed deformation monitoring method based on distributed optical fiber provided in the above embodiment of the present invention by running the computer program in the memory 103 .

Claims

1. A soil roadbed deformation monitoring method based on distributed optical fiber, characterized in that: include: Acquire a distributed optical fiber sensing data set of a target soil roadbed, wherein the distributed optical fiber sensing data set includes a plurality of temporally and spatially continuous optical fiber strain data sequences and temperature compensation data sequences; Dividing the target soil roadbed into a plurality of monitoring areas with different deformation characteristics according to the deformation intensity distribution of the optical fiber strain data sequence; performing coupled analysis processing on the optical fiber strain data sequence and the temperature compensation data sequence of each monitoring area to generate a multi-dimensional deformation feature set of the monitoring area; The multi-dimensional deformation feature set includes deformation rate features, stress accumulation correlation features and environmental disturbance response features; Generating deformation warning information of the target soil roadbed according to the spatiotemporal matching result of the multi-dimensional deformation feature set and the preset deformation threshold condition; A soil roadbed deformation optimization strategy is generated based on the deformation warning information, and the soil roadbed deformation optimization strategy is fed back to the roadbed maintenance system to trigger a reinforcement response operation.

2. The method according to claim 1, wherein The target soil roadbed is divided into a plurality of monitoring areas with different deformation characteristics according to the deformation intensity distribution of the optical fiber strain data sequence, including: Performing segmented gradient calculation on the optical fiber strain data sequence to extract the maximum deformation gradient value and average deformation fluctuation amplitude of each optical fiber segment; assigning a deformation intensity level to each optical fiber segment based on a ratio of the maximum deformation gradient value to the average deformation fluctuation amplitude; Clustering continuously distributed optical fiber segments with the same or adjacent deformation intensity levels into initial candidate monitoring areas, and adjusting the area boundaries of the initial candidate monitoring areas according to the deformation intensity level fluctuation range of the initial candidate monitoring areas; Merging or segmenting the adjusted initial candidate monitoring areas according to a preset area division density requirement to generate a final area division result including the multiple monitoring areas; The differential deformation characteristics are characterized by the median and standard deviation of the deformation intensity level of each monitoring area.

3. The method according to claim 2, wherein The coupling analysis processing of the optical fiber strain data sequence and the temperature compensation data sequence of each monitoring area to generate a multi-dimensional deformation feature set of the monitoring area includes: performing noise filtering and baseline calibration on the optical fiber strain data sequence to obtain a denoised standardized strain data sequence; constructing a temperature strain compensation coefficient matrix according to the temperature variation trend of the temperature compensation data sequence, and performing temperature drift correction on the standardized strain data sequence using the temperature strain compensation coefficient matrix to generate a corrected target strain data sequence; Performing a joint time-domain and frequency-domain analysis on the target strain data sequence to extract the time-domain deformation accumulation and frequency-domain disturbance response intensity of the monitoring area; Calculating a stress accumulation correlation characteristic value of the monitoring area according to a nonlinear correlation relationship between the time domain deformation accumulation amount and the frequency domain disturbance response intensity; The time domain deformation accumulation, frequency domain disturbance response intensity and stress accumulation correlation characteristic value are integrated to generate the multi-dimensional deformation feature set of the monitoring area.

4. The method according to claim 3, wherein The generating of deformation warning information of the target soil roadbed according to the spatiotemporal matching result of the multi-dimensional deformation feature set and the preset deformation threshold condition includes: Acquire a historical deformation feature library of the monitoring area, wherein the historical deformation feature library contains multi-dimensional deformation feature data of multiple historical deformation cases and corresponding deformation evolution results; Performing similarity matching between the multi-dimensional deformation feature set and feature data in the historical deformation feature library to determine a target matching case between the current deformation pattern of the monitoring area and a historical deformation case; Predicting the deformation development trend of the monitoring area based on the deformation evolution results of the target matching case, and generating a preliminary warning level based on the degree of deviation between the deformation development trend and the preset deformation threshold condition; Combined with the spatial distribution correlation of the preliminary warning levels of adjacent monitoring areas, the preliminary warning levels are subjected to spatiotemporal correction processing to generate the final warning level of the monitoring area; Based on the spatial superposition results of the final warning levels of all monitoring areas, the overall deformation warning information of the target soil roadbed is generated; the deformation warning information includes warning trigger conditions, risk area location identification and recommended response time.

5. The method according to claim 4, wherein The generating of the soil roadbed deformation optimization strategy based on the deformation warning information includes: Determining at least one high-risk monitoring area in the target soil roadbed that needs to be treated first according to the risk area positioning identifier; Acquire geological structure data and historical maintenance records of the high-risk monitoring area, and construct a deformation response model of the high-risk monitoring area; Based on the deformation response model, simulate the deformation suppression effect of different reinforcement schemes on the high-risk monitoring area, and select the optimal reinforcement scheme according to the simulation results; Generating the soil roadbed deformation optimization strategy including reinforcement material configuration parameters, construction timing planning, and resource allocation ratio according to the recommended response time and the optimal reinforcement plan; The resource allocation ratio is adjusted according to the deformation warning level of the high-risk monitoring area and the resource requirements of the remaining monitoring areas.

6. The method according to claim 5, wherein The construction process of the deformation response model includes: Collect soil stratification data, moisture content distribution data and load history data in the high-risk monitoring area; constructing a layered geomechanical model based on the soil layer stratification data, and correcting the interlayer shear strength parameters of the layered geomechanical model using the moisture content distribution data; Inputting the load history data into the revised layered geomechanical model to calculate the theoretical deformation response curve of the high-risk monitoring area under different load scenarios; By comparing the theoretical deformation response curve with the actual deformation monitoring data of the high-risk monitoring area, the parameter error of the layered geomechanical model is calibrated to obtain the calibrated deformation response model; The parameter error calibration process adopts an iterative feedback mechanism until the matching degree between the theoretical deformation response curve and the actual deformation monitoring data reaches a preset threshold.

7. The method according to claim 6, wherein The simulation of the deformation suppression effect of different reinforcement schemes on the high-risk monitoring area includes: Obtaining a set of candidate reinforcement solutions, the set of candidate reinforcement solutions including a plurality of combinations of reinforcement material types, layout densities, and structural forms; Inputting the parameters of each candidate reinforcement scheme into the deformation response model, and calculating the predicted deformation and stress distribution uniformity index of the high-risk monitoring area after reinforcement; Assigning a comprehensive suppression score to each candidate reinforcement scheme based on the difference between the predicted deformation and a preset safe deformation threshold, and the deviation between the stress distribution uniformity index and the ideal distribution; sorting all candidate reinforcement schemes based on the comprehensive inhibition score, and selecting the candidate reinforcement scheme with the highest score as the optimal reinforcement scheme; The comprehensive suppression score is obtained by weighted fusion of the difference and the normalized value of the deviation, and the weight coefficient is configured according to the warning level of the high-risk monitoring area.

8. The method according to claim 7, wherein The generation process of the construction timing plan includes: Determine the supply cycle and transportation route of required materials and equipment according to the reinforcement material configuration parameters; Based on the distribution density and traffic accessibility of the high-risk monitoring area, divide it into multiple construction sub-areas and assign construction priorities; Based on the resource allocation ratio and supply cycle, a construction time window and human resource scheduling plan are configured for each construction sub-area; Optimizing the parallel execution strategy of the human resource scheduling plan based on the overlap of the construction time windows and the equipment reuse requirements; The construction timing plan including the construction sub-areas, construction time windows, parallel execution strategies and emergency adjustment mechanisms is generated.

9. The method according to claim 1, wherein The method further comprises: Periodically acquiring an updated distributed optical fiber sensing data set of the target soil roadbed; regenerating a multi-dimensional deformation feature set of the monitoring area according to the updated distributed optical fiber sensing data set; Comparing the difference between the regenerated multi-dimensional deformation feature set and the historical multi-dimensional deformation feature set, and extracting the spatiotemporal propagation path characteristics of the difference; Determine the deformation source region and the affected region chain according to the spatiotemporal propagation path characteristics, and update the dynamic matching rule of the preset deformation threshold condition based on the coupling attenuation coefficient between the deformation rate characteristics of the deformation source region and the stress accumulation correlation characteristics of the affected region chain; Regenerate deformation warning information of the target soil roadbed according to the updated dynamic matching rules, and adjust the resource allocation ratio and construction timing planning in the soil roadbed deformation optimization strategy based on the priority of the deformation source area; The step of determining the deformation source region and the affected region chain according to the spatiotemporal propagation path characteristics includes: Performing a time series cluster analysis on the difference quantities to identify a set of monitoring areas having continuous deformation conduction characteristics; Constructing a deformation conduction topological network according to the deformation rate change direction and the transfer polarity of the stress accumulation correlation characteristics of the monitoring area set; Backtracking to determine the initial conduction node in the deformation source area through the node connection strength and directional weight in the deformation conduction topological network; generating a conduction path map including the affected area chain based on a spatial distribution relationship between the initial conduction node and the set of monitoring areas; Dividing the deformation source region and affected region chains of different levels according to the node attenuation gradient and the inter-region coupling strength in the conduction path map; The division result is used to configure the blocking reinforcement sequence for the conduction path and the cross-regional resource scheduling path in the construction timing planning.

10. A monitoring system, characterized in that: include: a memory, wherein the computer program is stored in the memory; A processor is used to load the computer program to implement the soil roadbed deformation monitoring method based on distributed optical fiber as described in any one of claims 1 to 9.

Images