An AI Spatiotemporal Digital Intelligence Platform for Pavement and Underground Diseases

Abstract

The present invention relates to the field of image processing technology, and discloses an AI spatio-temporal digital intelligence platform for pavement and underground diseases. The platform constructs a unified multi-scale wave field model by fusing micro-vibration sensor and underground radar data, then maps the two types of signals to the same space to generate a unified representation, then analyzes the mutual interference signals to form a conversion table, obtains an enhanced feature set across physical fields, generates an associated feature volume through three-dimensional registration, uses a super-resolution algorithm to obtain a super-resolution three-dimensional feature volume, and further generates a disease diagnosis report. The present invention solves the drawbacks of traditional technologies through a physically constrained depth fusion imaging system, realizes unified modeling and imaging of surface and underground diseases, improves the detection accuracy, and helps in the early and accurate diagnosis and treatment of diseases.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and more specifically, to an AI spatio-temporal digital intelligence platform for pavement and underground diseases. Background Art

[0002] In the field of road surface and underground structure disease detection and imaging, there are many key problems in traditional technologies: different technical means are used for surface and underground disease detection, resulting in data fragmentation and information silos, and it is impossible to present an integrated disease panorama; surface micro-vibration sensing and underground radar imaging technologies are independent of each other, and the mutual interference in signal processing is regarded as an interference factor; the heterogeneity of different physical field signals makes information fusion only stay at the post-processing level, and it is difficult to achieve deep fusion at the physical level; in harsh environments, the reliability of a single sensing modality drops significantly, seriously affecting the detection accuracy; existing technologies cannot effectively utilize the internal relationship between different physical fields, resulting in small diseases and early lesions being difficult to be reliably identified. To address these problems, the present invention proposes to construct a theory and application system for integrated surface and underground disease imaging with physical constraint deep fusion, realize unified physical modeling and collaborative imaging of surface and underground diseases, break through the limitations of existing technologies, and provide a new technical solution for the early accurate identification and treatment of road diseases. Summary of the Invention

[0003] The present invention provides an AI spatio-temporal digital intelligence platform for pavement and underground diseases, which solves the key problems in traditional road disease detection technologies in related technologies, such as data fragmentation, technology independence, limited information fusion, insufficient reliability, and failure to utilize internal relationships, and it is difficult to achieve integrated and accurate identification and collaborative treatment of surface and underground diseases.

[0004] The present invention provides an AI spatio-temporal digital intelligence method for pavement and underground diseases, including:

[0005] Receiving surface data collected by a micro-vibration sensor and imaging data obtained by an underground radar, mapping the micro-vibration waveform and the radar electromagnetic wave signal to the same physical space, and generating a unified multi-scale wave field model representing the common characteristics of the two physical fields;

[0006] Using an interference signal analysis method to extract valuable information in the interference signal, forming an interference information conversion table, and realizing cross-physical field information complementary enhancement based on the interference information conversion table to obtain a cross-physical field enhanced feature set;

[0007] Precisely registering the cross-physical field enhanced feature set in three-dimensional space to generate a three-dimensional surface and interior correlation feature body, applying a physically constrained super-resolution enhancement algorithm to the three-dimensional surface and interior correlation feature body to form a super-resolution three-dimensional feature body, analyzing the super-resolution three-dimensional feature body, and generating an integrated surface and underground disease diagnosis report.

[0008] Furthermore, the step of generating a unified multi-scale wave field model that characterizes the common characteristics of the two physical fields includes: extracting parameters and normalizing the representation of various physical field signals to obtain a unified multi-physical field parameter table; creating a set of nonlinear coupled wave field dynamics equations based on the unified multi-physical field parameter table; constructing a deep neural network architecture based on the cross-physical field coupling equation group, and obtaining the unified multi-scale wave field model by optimizing the objective function training.

[0009] Furthermore, the deep neural network architecture is trained with the following objective function:

[0010] ;

[0011] Information preservation loss function: ;

[0012] Physical constraint loss function: ;

[0013] in, and is a neural network model, is the physical model function, To balance the parameters, For input data, To output data, represents the L2 norm.

[0014] Furthermore, the step of extracting valuable information from the mutual interference signal includes: extracting the mutual interference component from the mixed signal using an adaptive iterative separation algorithm, generating a mutual interference signal characteristic spectrum, establishing a conversion model from the mutual interference signal to useful information, and forming a mutual interference information conversion table; the adaptive iterative separation algorithm is implemented by an iterative formula:

[0015] ;

[0016] ;

[0017] in, is the original mixed signal, For the The mutual interference component obtained by the iteration is For the The estimated iterative Pure signal, For the The estimated iterative Pure signal, is the signal extraction function, is the algorithm parameter set, where is the number of physical fields, is the number of iterations.

[0018] Furthermore, the conversion model from the mutual interference signal to useful information is implemented by the following mathematical transformation:

[0019] ;

[0020] in, is the mutual interference signal characteristic, is the intermediate information representation, and are the transformation matrix and bias vector, is the enhanced information after transformation, is a nonlinear activation function, is the feature transformation function, is element-wise multiplication.

[0021] Furthermore, the step of achieving complementary enhancement of cross-physical field information is implemented by the following enhancement function:

[0022] ;

[0023] in, For the Physics-like primitive features, To enhance the features, is the corresponding mutual interference information, For the Physics-like features, For the Physics to Physics-like characteristic transfer function: in, is a feature adaptive transformation network.

[0024] Furthermore, the step of accurately registering the cross-physical field enhanced feature set in three-dimensional space adopts the following spatial mapping equation:

[0025] ;

[0026] in, For spatial points The feature map value of For the The value of the characteristic field, is the position-related dynamic weight, which is calculated as follows:

[0027] ;

[0028] in, The characteristic field is at position The reliability measure of is the temperature parameter, is an exponential function.

[0029] Furthermore, the step of applying the physical constraint super-resolution enhancement algorithm to the three-dimensional surface-interior correlation feature body is achieved by solving the following optimization problem:

[0030] ;

[0031] in, is the reconstructed high-resolution feature volume, is the downsampling matrix, is the observation data, and is the regularization parameter, is the prediction function based on the physical model, is the total variation regularization term, is the physical constraint regularization term, represents the L2 norm.

[0032] Furthermore, the step of analyzing the super-resolution three-dimensional feature volume applies a multi-scale anomaly detection function:

[0033] ;

[0034] ;

[0035] in, For location The outliers, For super-resolution 3D feature bodies at position The eigenvalues of is the feature distance function, is a normal feature template, is the characteristic reliability function, is the Mahalanobis distance calculation, is the feature covariance matrix.

[0036] The present invention provides an AI spatiotemporal digital intelligence system for road surface and underground disease detection, comprising:

[0037] A data acquisition module, used to receive surface data collected by the micro-vibration sensor and imaging data acquired by the underground radar;

[0038] The wave field unified modeling module is used to build a multi-scale wave field unified model and map the micro-vibration waveform and radar electromagnetic wave signal into the same physical space;

[0039] The mutual interference signal analysis module is used to extract valuable information from the mutual interference signal and form a mutual interference information conversion table;

[0040] A feature enhancement module for achieving complementary enhancement of cross-physical-field information and obtaining a cross-physical-field enhanced feature set;

[0041] A spatial registration module for accurately registering the cross-physical-field enhanced feature set in a three-dimensional space to generate a three-dimensional table-related feature volume;

[0042] A super-resolution enhancement module for applying a super-resolution enhancement algorithm with physical constraints to the three-dimensional table-related feature volume to form a super-resolution three-dimensional feature volume;

[0043] A disease diagnosis module for analyzing the super-resolution three-dimensional feature volume and generating a table-and-interior integrated disease diagnosis report.

[0044] The beneficial effects of the present invention are as follows: By constructing a table-and-interior integrated disease imaging system with physically constrained deep fusion, the present invention solves problems in traditional technologies such as data fragmentation, improper handling of signal mutual interference, insufficient information fusion, low reliability in harsh environments, and difficulty in identifying minor diseases, realizes unified modeling and collaborative imaging of surface and underground diseases, improves detection accuracy and reliability, and provides an innovative solution for early and accurate identification and treatment of road diseases. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic flowchart of an AI spatio-temporal digital method for road surface and underground diseases of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] Now, the subject matter described herein will be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, changes can be made to the functions and arrangements of the elements discussed. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described in some examples can also be combined in other examples.

[0047] In at least one embodiment of the present invention, an AI spatio-temporal digital method for road surface and underground diseases is disclosed. As Figure 1 shown, it includes the following steps:

[0048] Step 100, constructing a multi-scale wave field unified model; using surface data collected by micro-vibration sensors and imaging data obtained by underground radar, a unified representation model characterizing the common features of the two physical fields is generated through a multi-scale wave field unified modeling method. This step maps the surface micro-vibration waveform and the underground radar electromagnetic wave signal to the same mathematical description framework to achieve unified expression of the two heterogeneous physical fields. Specifically, it includes the following steps:

[0049] Sub-step 101, parameterizing the physical field characteristics:

[0050] Input: surface micro-vibration raw signal data and ground radar echo raw data;

[0051] Execution process: Extract parameters and perform normalization representation on various physical field signals. Extract time-frequency domain features such as frequency, amplitude, and phase of the micro-vibration signal, obtain parameters such as the electromagnetic wave reflection characteristics, dielectric constant, and propagation speed of the radar signal, and apply a unified parameter mapping function, whose calculation formula is:

[0052]

[0053] Among them, is the parameterized mapping function of the th type of physical field, is the raw signal, are the th characteristic parameters extracted respectively;

[0054] Output: Obtain a unified multi-physical field parameter table, which contains the set of characteristic parameters and physical meaning explanations of each physical field signal, and is used for the subsequent construction of the wave field dynamics equation.

[0055] Sub-step 102, construction of the wave field dynamics equation:

[0056] Input: Unified multi-physical field parameter table;

[0057] Execution process: Create a set of non-linear coupled wave field dynamics equations according to the unified multi-physical field parameter table to establish the mutual influence relationship between different physical fields, and its calculation formula is:

[0058]

[0059] Among them, represents the second-order partial derivative of the th type of wave field (micro-vibration mechanical wave or radar electromagnetic wave) with respect to time , reflecting the change rate of the wave field with time; is the propagation speed corresponding to the th type of wave field; is the Laplace operator (second-order spatial derivative), which is used to describe the change of the wave field in space; represents the total influence of all other wave fields on the th type of wave field except the th type of wave field itself. Among them represents the gradient of the th type of wave field, is the coupling function between wave fields, representing the influence of the th type of wave field on the th type of wave field;

[0060] The specific form of the coupling function is:

[0061]

[0062] in, 、 and are linear coupling, gradient coupling and nonlinear coupling coefficients respectively;

[0063] Output: A set of coupled equations across physical fields is formed, which records the mathematical relationships and corresponding coefficients of the interactions between the physical fields, providing a theoretical basis for the co-representation space of physical constraints.

[0064] Sub-step 103, physical constraint co-representation space formation:

[0065] Input: Cross-physics coupled equations and raw signal data;

[0066] Execution process: A deep neural network architecture is constructed based on a set of coupled equations across physical fields, so that the neural network can simultaneously meet the dual goals of information retention and physical law constraints. The network is trained using the objective function, which is calculated as follows:

[0067]

[0068] in, is the parameter in the neural network model. By optimizing the parameter, the entire objective function reaches the minimum value;

[0069] It is the information retention loss, which is used to ensure that the feature representation retains the key information of the original signal. Its calculation formula is:

[0070]

[0071] in, For input data, is the expected output, is a neural network model, represents the L2 norm; : A trade-off parameter used to adjust the relative importance of data and physical constraints;

[0072] It is the physical constraint loss, which ensures that the feature representation conforms to the laws of physics. Its calculation formula is:

[0073]

[0074] in, is a neural network model, is the physical model function;

[0075] Output: Obtain the co-representation space of physical constraints, i.e., the unified multi-scale wave field model This model contains a set of neural network parameters for the unified representation of signals in different physical fields , physical constraint definitions and their weight configurations , which can be expressed as:

[0076]

[0077] where , respectively represent the th group of parameters of the neural network, , respectively represent the th physical constraint condition, , respectively represent the corresponding th constraint weight. This model can be used for subsequent analysis of cross-interference signals.

[0078] Step 200, construct a cross-interference signal analysis and feature enhancement mechanism; receive the unified multi-scale wave field model and the original mixed signal, and use the cross-interference signal analysis method to extract valuable information contained in the cross-interference signal. This step converts the cross-interference signal, which is usually regarded as interference, into an information source for enhancing the perception ability, and realizes mutual enhancement between signals. Specifically, it includes the following steps:

[0079] Sub-step 201, cross-interference signal feature separation:

[0080] Input: The mixed signal data of surface micro-vibration and ground penetrating radar, and the unified multi-scale wave field model;

[0081] Execution process: Use the adaptive iterative separation algorithm to extract the cross-interference components from the mixed signal, which is realized through the iterative formula. Its calculation formula is:

[0082]

[0083]

[0084] where is the original mixed signal, is the cross-interference component obtained in the th iteration, is the th estimated pure signal of the th class, is the signal extraction function, is the algorithm parameter set, is the number of physical fields, is the number of iterations;

[0085] Output: Generate a cross-interference signal feature map, which includes the feature representations of various cross-interference signals, time-frequency distribution maps, and the signal correlation strength matrix, for subsequent information enhancement.

[0086] Sub-step 202, cross-interference mode conversion:

[0087] Input: Cross-interference signal feature map and multi-scale wave field unified model;

[0088] Execution process: Establish a conversion model from cross-interference signals to useful information, implemented through mathematical transformation, and its calculation formula is:

[0089]

[0090]

[0091] where, is the cross-interference signal feature, is the intermediate information representation, and are the conversion matrix and bias vector, is the enhanced information after conversion, is the non-linear activation function, is the feature transformation function, is the element-wise multiplication;

[0092] Output: Form a cross-interference information conversion table, which includes the conversion rule set, information enhancement parameters, and feature mapping relationships, providing a basis for cross-physical field information enhancement.

[0093] Sub-step 203, cross-physical field complementary enhancement:

[0094] Input: Cross-interference information conversion table and original physical field features;

[0095] Execution process: Enhance the feature expression ability of a single physical field through the mutual complementation of information in each physical field, and its calculation formula is:

[0096]

[0097] where, is the original feature of the th type of physical field, is the enhanced feature, is the feature of the th type of physical field, is the corresponding cross-interference information; is from the th type of physical field to the th type of physical field feature conversion function, and its calculation formula is:

[0098]

[0099] Among them, is a feature adaptive transformation network that automatically adjusts the transformation strategy according to the specific physical field characteristics; <r

[0100] Output: Obtain a cross - physical - field enhanced feature set , which integrates the information advantages of each physical field and can be expressed as:

[0101]

[0102] Among them, , , are respectively the feature representations after enhancement of the th physical field, is the enhancement effect quantization index, is the feature quality evaluation result, is the total number of physical fields. This feature set provides a rich data basis for subsequent imaging.

[0103] Step 300, implement super - resolution integrated imaging of the surface and interior of diseases; based on receiving the cross - physical - field enhanced feature set, use the integrated surface - interior super - resolution imaging algorithm to achieve unified high - precision imaging of road surface and underground diseases. This step completes the transformation from independent detection to integrated imaging, generating a three - dimensional imaging result containing surface - interior correlation information. Specifically, it includes the following steps: [[ID=z

[0104] Sub - step 301, surface - interior correlation spatial registration:

[0105] Input: Cross - physical - field enhanced feature set and original road detection coordinate data - Execution process: Precisely align and map the surface and underground features in three - dimensional space, apply the spatial mapping equation, and its calculation formula is:

[0106]

[0107] Among them, is the feature mapping value of the spatial point , is the value of the th feature field, is the position - related dynamic weight, satisfying , is the number of feature fields.

[0108] The weight calculation uses the softmax function, and its calculation formula is:

[0109]

[0110] Among them, is the reliability measure of the feature field at the position , is the temperature parameter, is the exponential function;

[0111] Output: Generate a three-dimensional table-related feature body, which contains the spatial correspondence, correlation strength distribution, and integrated three-dimensional feature data of surface and underground features, providing a basis for super-resolution reconstruction.

[0112] Sub-step 302, physical constraint super-resolution enhancement:

[0113] Input: Three-dimensional table-related feature body and multi-scale wave field unified model;

[0114] Execution process: Use the regularization method with physical constraints to improve the spatial resolution of the feature body. This process is completed by the super-resolution enhancement algorithm, and its calculation formula is:

[0115]

[0116] Among them, is the reconstructed high-resolution feature body, is the downsampling matrix, is the observed data;

[0117] is the total variation regularization term:

[0118]

[0119] is the physical constraint regularization term:

[0120]

[0121] Among them, is the prediction function based on the physical model, and are regularization parameters, , , is the spatial coordinate of the feature body, where is the spatial gradient operator;

[0122] Output: Form a super-resolution three-dimensional feature body, which has a higher spatial resolution than the original data and contains refined road surface and underground structure features, resolution improvement quantification indicators, and quality assessment data.

[0123] Sub-step 303, integrated table and field disease location and analysis:

[0124] Input: Super-resolution three-dimensional feature body and pre-calibrated disease feature library;

[0125] Execution process: Disease detection, localization, and classification are performed on the super-resolution three-dimensional feature volume, and a multi-scale anomaly detection function is applied. Its calculation formula is:

[0126]

[0127] where, is the outlier value of the position ; is the feature distance function, is the normal feature template, is the feature reliability function.

[0128] The feature distance function adopts the Mahalanobis distance, and its calculation formula is:

[0129]

[0130] where, is the calculation of the Mahalanobis distance, is the feature covariance matrix;

[0131] Combining the disease type, location, and severity information obtained above, analyze the possible disease causes, and then match the predetermined disease repair plan library through the disease diagnosis information to give targeted repair suggestions;

[0132] Output: Integrate the above disease diagnosis information into an integrated diagnosis report, and output a table-integrated disease diagnosis report.

[0133] This report includes the disease type determination result, accurate three-dimensional position coordinates, disease severity level assessment, and repair plan suggestions, providing a scientific basis for road maintenance decision-making.

[0134] The physical constraint depth fusion table-integrated disease imaging system implemented through the above steps achieves the following technical effects:

[0135] Improved disease detection sensitivity: This system realizes the high-precision detection ability for fine road diseases, improves the surface micro-crack detection sensitivity from 0.5 mm of the traditional method to 0.1 mm, and extends the underground cavity identification depth from 2.5 meters of the traditional method to 4 meters, providing technical possibilities for the ultra-early detection of diseases;

[0136] Enhanced environmental adaptability: This system maintains a detection accuracy of more than 98% under harsh environmental conditions (such as heavy rain, high temperature, severe cold, vibration noise, and electromagnetic interference), which is significantly improved compared with the detection accuracy of only 75% of the traditional single-modal detection technology under the same conditions;

[0137] Improved accuracy of disease classification: This system has increased the accuracy of disease classification in the early stage from 70% of traditional methods to 95%, reducing false alarms and missed detections, and significantly lowering unnecessary maintenance costs caused by misjudgments.

[0138] Disease development prediction function: By obtaining integrated surface and underground disease information, this system can predict the development trend of diseases in the next 3 - 6 months with an accuracy rate of 85%, providing a scientific basis for road maintenance decision-making.

[0139] Optimized detection speed and energy consumption: This system has increased the road disease detection speed by 2.5 times while reducing energy consumption by 40%, breaking through the limitation of the trade-off between speed and accuracy in traditional detection technologies.

[0140] Surface - underground correlation analysis ability: This system has realized the correlation analysis between surface and underground diseases for the first time, being able to determine the influence mechanism of underground structure changes on the development of surface diseases and the reflection law of surface conditions on underground structures, forming a complete disease evolution chain and laying a foundation for the full - cycle road health management.

[0141] Most significantly, for the detection of composite diseases such as "surface micro - cracks + base moisture intrusion", the accuracy rate of this method reaches 93.8%, while that of traditional methods is only 59.3%, an increase of 34.5 percentage points.

[0142] In this application example, two key technical effects of the physical - constraint depth - integrated surface - underground disease imaging system are mainly verified: the ability to detect micro - diseases and the environmental adaptability.

[0143] To verify the system's ability to detect micro - diseases, 32 disease sample points of different types and severities were selected on the example road section, including surface micro - cracks (width 0.05 - 0.5mm), interlayer debonding, base segregation, etc. True labeled data was obtained through core - drilling sampling. The detection results of this method were compared with those of traditional methods, as shown in Table 1:

[0144] Table 1: Comparison of the ability to detect micro - diseases:

[0145]

[0146] Data shows that the average detection rate of this method reaches 93.6%, an increase of 32.5 percentage points compared with 61.1% of traditional methods. Especially in the detection of micro - diseases, the minimum surface crack width that can be identified reaches 0.08mm, which is 6.25 times higher than 0.5mm of traditional methods; the maximum underground disease depth that can be detected reaches 380cm, which is 58.3% higher than 240cm of traditional methods. This effect benefits from the innovative use of cross - interference signals as information carriers and the application of physical - constraint super - resolution enhancement algorithms.

[0147] To verify the robustness of the system under different environmental conditions, multiple tests were conducted at the same detection point on a sample road section under different weather conditions and road conditions to evaluate the system's detection accuracy. The results are shown in Table 2:

[0148] Table 2: Comparison of detection accuracy under different environmental conditions:

[0149]

[0150] Improvement in anti-interference capability = (error reduction rate of this method - error reduction rate of traditional method) / error reduction rate of traditional method × 100%

[0151] Data shows that this method maintains a high detection accuracy across a wide range of environmental conditions, averaging 96.5%, a 25.1 percentage point improvement over the 71.4% achieved by traditional methods. The advantages of this method are particularly pronounced under adverse environmental conditions (such as heavy rain, flooded roads, and vehicle vibration interference), with accuracy improving by over 30 percentage points and interference immunity increasing by over 80%. This is primarily due to the method's deep integration of information from different physical fields, enabling the system to maintain overall performance by leveraging complementary information from another modality when one sensing modality is disturbed.

[0152] In summary, this application example fully verifies the technical effectiveness of the physical constraint deep fusion integrated internal and external disease imaging system, especially its significant advantages in micro-disease detection and environmental adaptability, providing strong support for the early and accurate identification of road diseases and full life cycle management.

[0153] In one embodiment of the present invention, an example of the aforementioned AI spatiotemporal digital intelligence method for pavement and underground diseases is provided:

[0154] Application Scenario Description: This implementation method is used for pavement disease detection and diagnosis on the Beijing-Shanghai Expressway section from K1457+200 to K1457+400. This section is an asphalt concrete pavement built in 2008. After more than 10 years of use, it has developed various defects such as surface cracks, interlayer debonding, and moisture intrusion into the base layer. This section has the following characteristics:

[0155] The annual average daily traffic (AADT) of the road section is as high as 42,000 vehicles, with heavy traffic accounting for 28.6%;

[0156] The road section spans two geological conditions: K1457+200 to K1457+300 is a rock foundation, and K1457+300 to K1457+400 is a soft soil foundation;

[0157] The road section is significantly affected by seasonal rainfall, with an annual rainfall of 1,280 mm;

[0158] There is a large regional temperature difference, with the highest temperature reaching 40°C in summer and the lowest temperature reaching -10°C in winter;

[0159] The road section is located in the suburban area of the city, and there are various electromagnetic interference sources;

[0160] The historical maintenance records of the road section show that traditional single detection methods cannot effectively identify deep diseases, resulting in the recurrence of diseases in the short term after maintenance.

[0161] This application scenario covers complex pavement structures, diverse environmental conditions, and severe traffic loads, representing the typical difficulties in disease detection of high-grade highways. This implementation method aims to achieve integrated surface and subsurface disease imaging through the physical constraint depth fusion method, providing a scientific basis for accurate diagnosis and targeted repair.

[0162] On the section of the Beijing-Shanghai Expressway from K1457+200 to K1457+400, a detection vehicle equipped with the following equipment is used for data collection:

[0163] A set of laser scanning systems (scanning frequency: 2000Hz, scanning width: 3.6m);

[0164] A set of micro-vibration sensor arrays (frequency response: 0.5 - 5000Hz, sampling rate: 25kHz);

[0165] A set of ground penetrating radar systems (center frequency: 1.6GHz, scanning depth: 5m);

[0166] A set of infrared thermal imaging systems (temperature resolution: 0.05°C, spatial resolution: 3mm);

[0167] Data collection is carried out under normal traffic conditions. The detection vehicle travels at a constant speed (60km / h), and each sensor synchronously collects data. The collection process is shown in Table 3:

[0168] Table 3: Parameters of the data collection process:

[0169]

[0170] Multiple collections are carried out under different environmental conditions to verify the stability and adaptability of the system under various conditions.

[0171] Example of Step 100, constructing a multi-scale wave field unified model:

[0172] Apply the first-step processing to the collected raw data to extract the characteristic information of each physical field. The results are shown in Table 4:

[0173] Table 4: Results of cross-physical field feature extraction:

[0174]

[0175] Feature extraction uses the convolutional sparse transform algorithm, combines expert knowledge for feature selection, effectively removes redundant information, reduces the feature dimension, and at the same time retains the maximized effective information.

[0176] Step 200 example, construct an interference signal analysis and feature enhancement mechanism:

[0177] The adaptive iterative separation algorithm is used to process the interference components between each sensing signal. For the typical mixed signal at the example section K1457+250, the iterative separation results are shown in Table 5:

[0178] Table 5: Iterative separation results of interference signals (at section K1457+250):

[0179]

[0180] Special attention is paid to the analysis results of the interference signals at four typical disease locations in the section. The interference components are converted into valuable feature information through the interference information conversion model, as shown in Table 6:

[0181] Table 6: Information enhancement results of interference signals (at four typical disease locations):

[0182]

[0183] Analysis shows that the recovery rate of the interference signal of the settlement and water intrusion composite disease at section K1457+315 is the highest, reaching 90.4%. This is mainly because water intrusion has a significant impact on both the electromagnetic wave and the thermal field, and the interference signal contains rich water distribution information.

[0184] Step 300 example, achieve integrated imaging of super-resolution disease surface and interior:

[0185] The physical constraint super-resolution enhancement algorithm is applied to enhance the three-dimensional feature volume, focusing on the feature fusion situation at different depths, as shown in Table 7:

[0186] Table 7: Feature fusion parameters at different depths (section K1457+300 - K1457+320):

[0187]

[0188] The final generated integrated high-resolution three-dimensional disease image of the surface and interior is compared with the traditional method, as shown in Table 8:

[0189] Table 8: Comparison of integrated imaging results:

[0190]

[0191] At the typical disease area K1457+315, the integrated surface and inside imaging clearly shows the complete disease chain from surface settlement to moisture intrusion into the base layer, revealing the evolution mechanism of moisture intrusion leading to a decrease in roadbed strength and then inducing surface settlement, providing a scientific basis for precise maintenance.

[0192] In one embodiment of the present invention, for the dynamic evolution prediction and early warning of road surface and underground structure defects, an AI spatiotemporal digital intelligence method for road surface and underground defects further includes the following steps:

[0193] This implementation retains the capabilities of Implementation 1 and introduces a time dimension and multi-scale dynamic prediction model to accurately predict the future development of diseases. It proposes and implements the "Multi-level Prediction Theory of Spatiotemporal Evolution of Diseases" (MSEPT). Its main innovations include:

[0194] The multi-scale dynamic equations of disease evolution establish the relationship between internal and external disease evolution through the diffusion coefficient tensor, environmental response function and structural stress function;

[0195] Adaptive multi-timescale prediction mechanism, dynamically adjusts the prediction time window and model parameters; physical constraint-based prediction uncertainty quantification method realizes the reliability assessment of prediction results;

[0196] An algorithm for identifying key turning points in disease evolution can predict the acceleration and stabilization points of disease development;

[0197] The dynamic response model of environment-material-structure coupling can enhance the adaptability to changes in external conditions.

[0198] This implementation extends static disease detection to dynamic prediction, achieving a technological leap from "diagnosing what" to "predicting what will happen."

[0199] Steps 100 to 300 are identical to those in implementation mode 1. The unified physical field representation model, enhanced feature information, and high-precision disease diagnosis report obtained in these three steps will serve as input data for step 400 .

[0200] Step 400: Prediction and early warning of spatiotemporal disease evolution. This step is accomplished through multiple sub-steps, moving from static disease diagnosis to dynamic spatiotemporal disease evolution prediction and early warning.

[0201] Sub-step 401: Construction of disease spatiotemporal database:

[0202] Receive integrated disease diagnosis report data and historical monitoring records as input information;

[0203] Structurally process the received data according to spatio-temporal dimensions to obtain three types of time-series data:

[0204] Time-series data of disease status: including time-varying parameters of disease type, location, size, and severity;

[0205] Time-series data of environmental conditions: including environmental monitoring parameters of temperature, humidity, and precipitation;

[0206] Time-series data of load conditions: including load parameters of traffic flow, vehicle type, and axle load

[0207] The data is organized into a structured data set after spatio-temporal coordinate alignment, and its calculation formula is:

[0208]

[0209] Where, represents the state parameter vector of the th disease sample at time , is the corresponding environmental condition vector, is the load condition vector, is the total number of disease samples, is the th number of observation time points of the sample;

[0210] Generate a disease spatio-temporal data set, which contains complete records of disease status, environmental conditions, and load conditions with spatio-temporal alignment.

[0211] Sub-step 402, construction of a mathematical model for disease evolution:

[0212] Receive the disease spatio-temporal data set and the multi-scale wave field unified model as input information;

[0213] Based on the received data, construct a differential equation for disease state evolution, and its calculation formula is:

[0214]

[0215] Where, represents the disease state variable at spatial position at time , is the gradient operator, represents the disease diffusion coefficient tensor, represents the environmental response function, represents the structural stress function.

[0216] Establish parametric models for different types of diseases respectively:

[0217] Surface crack disease model:

[0218] An anisotropic diffusion coefficient is adopted, and this coefficient can be expressed by the formula as , where are the spatial position coordinates. This coefficient is mainly used to characterize the influence of temperature change and stress action on surface crack diseases. Assuming the temperature is , and the stress is , a relationship function can be established:

[0219]

[0220] The specific function form needs to be determined according to the actual physical laws and experimental data;

[0221] Interlayer debonding disease model:

[0222] Interface characteristic parameters are adopted. Let the interface characteristic parameters be , and these parameters mainly characterize the influence of interface stress and freeze-thaw action on interlayer debonding diseases. A relationship function can be established:

[0223]

[0224] where represents the freeze-thaw related parameters, and the specific form of the function needs to be determined according to the actual situation;

[0225] Underground disease model:

[0226] Material characteristic parameters are adopted. Let the material characteristic parameters be , and they mainly characterize the influence of moisture change and load transfer on underground diseases. A relationship function can be established:

[0227]

[0228] The specific form of the function needs to be determined according to the actual physical process and experimental data;

[0229] When determining the model parameters, it is achieved by minimizing the mean square error between the predicted value and the measured value. Let the predicted value be , the measured value be , and the mean square error is calculated by the formula:

[0230]

[0231] where, is the disease type, is the number of samples. By adjusting the model parameters, make Reach the minimum to determine each model parameter.

[0232] Finally, a mathematical model of disease evolution is generated. This model takes the current disease state and environmental conditions as inputs. Through the parameterized model established above and the determined parameters, it outputs the predicted values of the disease state at future time points .

[0233] Sub-step 403, time-scale adaptive prediction execution; receive the mathematical model of disease evolution and the current disease state data as input information;

[0234] Based on the received data, perform multi-time-scale prediction calculations:

[0235] Calculate the time interval sequence according to the disease type characteristics, and its calculation formula is:

[0236]

[0237] Among them, represents the th prediction time interval, represents the basic time interval parameter, represents the disease state change rate function, represents the disease state change acceleration function, and represent adjustment parameters;

[0238] Fuse the prediction results of multiple time-scale sub-models, and its calculation formula is:

[0239]

[0240] Among them, represents the fused prediction result, represents the th sub-model's prediction result, represents the time-related weight function, represents the total number of sub-models;

[0241] Apply physical constraint conditions to correct the prediction results to ensure compliance with the laws of material mechanics and structural deformation;

[0242] Generate disease state prediction data, which includes the predicted values of the disease state and their confidence intervals in the short term (within 3 months), medium term (6 months), and long term (1 year).

[0243] Sub-step 404, disease evolution inflection point identification; receive the disease state prediction data as input information. Based on the received data, perform key inflection point identification analysis: ​

[0244] Let the predicted trajectory be a function , whose time derivative is , by calculating Get the rate of change, and then Derivative to get acceleration ; Set the rate of change threshold and acceleration threshold ,when and The corresponding time point This is the time point when the rate of change and acceleration change significantly;

[0245] For the disease state vector , assuming its dimension is , mapping it to dimensional phase space , constructing the phase space trajectory By analyzing the geometric characteristics of the phase space trajectory, such as the curvature of the trajectory , the distance between tracks When these geometric features change significantly, it is identified as a structural change of the disease state vector in high-dimensional space;

[0246] Assume that the disease state variable is , environmental factors are , mutual information value The calculation formula is:

[0247]

[0248] in, and They are and The value set of yes and The joint probability of and They are and The marginal probability of Greater than a set threshold When the corresponding environmental factors These are key environmental triggers.

[0249] Assume that the change rate, acceleration, phase space trajectory geometric characteristics change and mutual information value obtained from the above steps are , assign weights to each factor , then the inflection point importance score The calculation formula is:

[0250]

[0251] Sort the inflection points according to the calculated values.

[0252] Generate the inflection point data of disease evolution, which contains the key turning time points within the predicted time range. Each inflection point includes the time position, the predicted value of the disease state, and the importance score.

[0253] Sub-step 405, disease risk level calculation; Receive the predicted disease state data and the inflection point data of disease evolution as input information;

[0254] Based on the received data, perform risk quantification calculation:

[0255] Apply the time decay function to calculate the uncertainty index at different predicted time points;

[0256] Calculate the risk value of the road section, and its calculation formula is:

[0257]

[0258] Among them, represents the risk value at position at time , represents the disease severity function, represents the prediction uncertainty, represents the infrastructure importance index;

[0259] Divide the risk into four levels according to the risk value threshold: normal, attention, warning, and emergency;

[0260] Generate a disease risk assessment report, which contains the spatio-temporal distribution map of disease risk, the division of grading warning thresholds, and the risk tips at key time points.

[0261] Sub-step 406, maintenance plan generation:

[0262] Data reception: Receive the disease risk assessment report and the maintenance technology parameter library as input information;

[0263] Maintenance plan calculation:

[0264] Calculation of the optimal intervention time window: Let the cost function be , and the risk function be , and calculate the optimal intervention time window through the formula , and its calculation formula is:

[0265]

[0266] Among them, is a functional relationship determined according to the actual situation, which comprehensively considers cost and risk factors to determine the most appropriate intervention time;

[0267] Selection of maintenance technology plan: Based on the disease type , development trend and road condition environment , select an appropriate maintenance technology plan from the maintenance technology parameter library , and its calculation formula is:

[0268]

[0269] where is a selection function that screens out the most suitable plan from numerous maintenance technology plans according to different input conditions;

[0270] Calculation of long-term effect and economic evaluation indicators: For different maintenance plans , calculate their long-term effect indicators and economic evaluation indicators . The long-term effect indicators can be obtained by comprehensively considering various factors such as road performance improvement and disease recurrence probability; the economic evaluation indicators can involve calculations of maintenance costs, expected revenues, etc.;

[0271] Report generation: Generate an auxiliary report for maintenance decision-making, which includes the recommended maintenance time window , suggestions for hierarchical intervention plans and prediction of intervention effects (including long-term effect indicators and economic evaluation indicators and other related contents).

[0272] The technical effects obtained through the above steps include:

[0273] Accuracy of multi-time scale prediction: Achieve accurate predictions for different time spans through disease state prediction data, with a short-term (within 3 months) prediction accuracy of 92%, a medium-term (6 months) prediction accuracy of 85%, and a long-term (1 year) prediction accuracy of 75%, providing reliable data for full life cycle management;

[0274] Early identification of disease development inflection points: Achieve early capture of key transition time points through disease evolution inflection point data, and on average, discover the transition of diseases from slow to rapid deterioration 15 - 30 days in advance;

[0275] Comprehensive consideration of environmental factors: Incorporate more than 15 environmental parameters into consideration through the environmental response function in the disease evolution mathematical model, significantly improving the adaptability of predictions in changing environments;

[0276] Quantification of the reliability of prediction results: The confidence interval estimate included in the disease risk assessment report provides a 95% confidence level, enabling maintenance decisions to be made while considering risks;

[0277] Optimization of maintenance costs and service life: By means of the optimal intervention time window and plan recommended in the maintenance decision-making assistance report, the road maintenance cost is reduced by 30%, and the service life is extended by 15% - 20%;

[0278] Precision of maintenance decision-making: The accuracy rate of the intervention plan recommended in the maintenance decision-making assistance report for different disease types reaches 87%, effectively supporting maintenance management decisions.

[0279] In summary, based on static disease detection, this embodiment realizes the technical leap from "disease identification" to "development prediction" by introducing the time dimension and constructing a disease spatio-temporal evolution prediction model, providing a complete technical solution for the whole-life cycle maintenance management of roads.

[0280] The above describes the embodiments of the present invention. However, these embodiments are not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative rather than restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make more equivalent embodiments in various forms, all of which fall within the protection scope of this embodiment.

Claims

1. An AI spatio-temporal digital intelligence method for pavement and underground diseases, characterized in that, Including the following steps: Receiving the surface data collected by the micro-vibration sensor and the imaging data obtained by the ground penetrating radar, mapping the micro-vibration waveform and the radar electromagnetic wave signal into the same physical space, and generating a unified multi-scale wave field model characterizing the common features of the two physical fields; Using the cross-interference signal analysis method to extract valuable information from the cross-interference signal, forming a cross-interference information conversion table, and realizing cross-physical field information complementary enhancement based on the cross-interference information conversion table to obtain a cross-physical field enhanced feature set; Precisely registering the cross-physical field enhanced feature set in three-dimensional space to generate a three-dimensional inside-outside associated feature volume, applying a physically constrained super-resolution enhancement algorithm to the three-dimensional inside-outside associated feature volume to form a super-resolution three-dimensional feature volume, analyzing the super-resolution three-dimensional feature volume, generating an inside-outside integrated disease diagnosis report, and using the following spatial mapping equation: ； Among them, is the feature mapping value of the spatial point , is the value of the th feature field is the position-related dynamic weight, and its calculation method is as follows: ； wherein, is the reliability measure of the characteristic field at the position , is the temperature parameter, is the exponential function.

2. The AI spatio-temporal digital method for pavement and underground diseases according to claim 1, wherein The step of generating the unified multi-scale wave field model characterizing the common features of the two physical fields includes: extracting and normalizing the parameters of various physical field signals to obtain a unified multi-physical field parameter table; creating a set of non-linear coupled wave field dynamics equations according to the unified multi-physical field parameter table; constructing a deep neural network architecture based on the cross-physical field coupled equations set, and training through optimizing the objective function to obtain the unified multi-scale wave field model.

3. The AI spatio-temporal digital method for pavement and underground diseases according to claim 2, characterized in that, The deep neural network architecture is trained through the following objective function: ； Information retention loss function: ; Physical constraint loss function: ; Among them, and are neural network models, is a physical model function, is a trade-off parameter, is the input data, is the output data, represents the L2 norm.

4. A pavement and underground disease AI spatio-temporal digital intelligence method according to claim 1, characterized in that, The step of extracting valuable information from the cross-interference signal includes: using the adaptive iterative separation algorithm to extract the cross-interference component from the mixed signal, generating a cross-interference signal feature map, establishing a conversion model from the cross-interference signal to useful information, and forming a cross-interference information conversion table; the adaptive iterative separation algorithm is realized through the iterative formula: ； ； Wherein, is the original mixed signal, is the interference component obtained in the th iteration, is the th iteration estimated th class pure signal, is the th iteration estimated th class pure signal, is the signal extraction function, is the algorithm parameter set, wherein is the number of physical fields, is the number of iterations.

5. The AI spatio-temporal digital method for pavement and underground diseases according to claim 4, characterized in that, The conversion model from the cross-interference signal to useful information is realized through the following mathematical transformation: ； Among them, is the mutual interference signal feature, is the intermediate information representation, and are the transformation matrix and the bias vector, is the transformed enhanced information, is the non-linear activation function, is the feature transformation function, is the element-wise multiplication.

6. The AI spatio-temporal digital method for pavement and underground diseases according to claim 1, characterized in that The step of realizing cross-physical field information complementary enhancement is realized through the following enhancement function: ； Among them, is the original feature of the nth type of physical field, is the enhanced feature, is the corresponding mutual interference information, is the feature of the nth type of physical field, is the feature transformation function from the nth type of physical field to the mth type of physical field: Among them, is the feature adaptive transformation network.

7. A pavement and underground disease AI spatio-temporal digital intelligence method according to claim 1, characterized in that, The step of applying the physically constrained super-resolution enhancement algorithm to the three-dimensional inside-outside associated feature volume is realized by solving the following optimization problem: ； Among them, is the reconstructed high-resolution feature volume, is the downsampling matrix, is the observed data, and are regularization parameters, is the prediction function based on the physical model, is the total variation regularization term, is the physical constraint regularization term, represents the L2 norm.

8. A pavement and underground disease AI spatio-temporal digital intelligence method according to claim 1, characterized in that The step of analyzing the super-resolution three-dimensional feature volume applies a multi-scale anomaly detection function: ； ； Among them, is an outlier at position , is the eigenvalue of the super-resolution three-dimensional feature volume at position , is the feature distance function, is the normal feature template, is the feature reliability function, is the Mahalanobis distance calculation, is the feature covariance matrix.

9. An AI spatio-temporal digital intelligence system for pavement and underground diseases, which is used to execute an AI spatio-temporal digital intelligence method for pavement and underground diseases as described in any one of claims 1-8, and is characterized in that, Including: A data acquisition module for receiving the surface data collected by the micro-vibration sensor and the imaging data obtained by the ground penetrating radar; A wave field unified modeling module for constructing a unified multi-scale wave field model and mapping the micro-vibration waveform and the radar electromagnetic wave signal into the same physical space; A cross-interference signal analysis module for extracting valuable information from the cross-interference signal and forming a cross-interference information conversion table; A feature enhancement module for realizing cross-physical field information complementary enhancement and obtaining a cross-physical field enhanced feature set; A spatial registration module for precisely registering the cross-physical field enhanced feature set in three-dimensional space to generate a three-dimensional inside-outside associated feature volume; A super-resolution enhancement module for applying a physically constrained super-resolution enhancement algorithm to the three-dimensional inside-outside associated feature volume to form a super-resolution three-dimensional feature volume; A disease diagnosis module for analyzing the super-resolution three-dimensional feature volume and generating an inside-outside integrated disease diagnosis report.

Images