Space-air-ground integrated high-speed rail line infrastructure state evaluation method

Abstract

The application discloses a kind of high-speed rail line infrastructure state evaluation methods based on space-air-ground integration, belong to high-speed rail line infrastructure state evaluation technical field, this method includes: obtaining multi-source data set;The data in multi-source data set is converted into multiple structured evaluation indexes, and evaluation index set is obtained;After the multiple structured evaluation indexes in evaluation index set are time series alignment, mile alignment, space alignment, multiple evaluation vectors are obtained;State evaluation model is used to process multiple evaluation vectors, and the prediction result of road bridge tunnel track is obtained, prediction result includes degradation state prediction result and residual life prediction result;Real-time degradation state and real-time residual life are determined based on real-time monitoring data;Based on the difference between real-time degradation state, real-time residual life and prediction result, the parameters of state evaluation model are updated.The method can accurately evaluate the state of road bridge tunnel track.

Description

Technical Field

[0001] This disclosure relates to the field of high-speed railway infrastructure condition assessment technology, and in particular to a high-speed railway infrastructure condition assessment method based on the integration of air, space, vehicle and ground systems. Background Technology

[0002] As my country's high-speed rail network continues to expand, the long-term performance evaluation of its infrastructure faces significant challenges.

[0003] Current high-speed rail lines include various structures such as tracks, bridges, and tunnels. Their condition is affected by multiple factors such as geological environment and load. Traditional manual inspection and single-point detection cannot comprehensively consider the impact of multiple factors on high-speed rail line infrastructure, resulting in inaccurate assessment results of high-speed rail line infrastructure condition. Summary of the Invention

[0004] This disclosure provides a method for assessing the condition of high-speed railway infrastructure based on an integrated space-air-ground system, enabling accurate condition assessment of roads, bridges, tunnels, and tracks. The technical solution includes at least the following components: Firstly, a method for assessing the condition of high-speed railway infrastructure based on an integrated air-space-vehicle-ground system is provided, comprising: acquiring a multi-source data set, the multi-source data set including data collected by multi-source sensors for monitoring roads, bridges, tunnels, and tracks; converting the data in the multi-source data set into multiple structured evaluation indicators to obtain an evaluation indicator set, the structured evaluation indicators being used to assess the condition of roads, bridges, tunnels, and tracks; aligning the multiple structured evaluation indicators in the evaluation indicator set according to time sequence, mileage, and space to obtain multiple evaluation vectors; processing the multiple evaluation vectors using a condition assessment model to obtain prediction results for roads, bridges, tunnels, and tracks, the prediction results including degradation state prediction results and remaining life prediction results; determining the real-time degradation state and real-time remaining life based on real-time monitoring data; and updating the parameters of the condition assessment model based on the differences between the real-time degradation state, the real-time remaining life, and the prediction results.

[0005] Optionally, the state assessment model includes a convolutional neural network (CNN), a long short-term memory network (LSTM), a cross-modal feature interaction layer, and a multi-task output layer connected in sequence. The CNN is used to extract spatial features, the LSTM is used to extract temporal features, the cross-modal feature interaction layer is used to fuse the spatial features and the temporal features, and the multi-task output layer is used to output the prediction results for roads, bridges, tunnels, and tracks. The parameters of the state assessment model are updated based on the difference between the real-time degradation state, the real-time remaining lifetime, and the prediction results, including the difference between the real-time degradation state of the roadbed, the real-time remaining lifetime, and the prediction results. If the roadbed difference threshold is met, update the forget gate weight matrix parameters of the LSTM; if the difference between the real-time degradation state of the bridge, the real-time remaining lifetime, and the prediction result meets the bridge difference threshold, update the convolution kernel parameters of the CNN; if the difference between the real-time degradation state of the tunnel, the real-time remaining lifetime, and the prediction result meets the tunnel difference threshold, update the multi-head attention weights of the cross-modal feature interaction layer; if the difference between the real-time degradation state of the track, the real-time remaining lifetime, and the prediction result meets the track difference threshold, update the fully connected layer weights of the multi-task output layer.

[0006] Optionally, updating the forget gate weight matrix parameters of the LSTM includes: reweighting the forget gate weight matrix parameters of the LSTM based on a temporal attention mechanism; updating the convolution kernel parameters of the CNN includes: updating the convolution kernel parameters of the CNN based on gradient pruning; updating the multi-head attention weights of the cross-modal feature interaction layer includes: updating the multi-head attention weights of the cross-modal feature interaction layer using a federated learning collaborative fine-tuning algorithm; updating the fully connected layer weights of the multi-task output layer includes: updating the fully connected layer weights of the multi-task output layer using an adaptive momentum optimization algorithm.

[0007] Optionally, the step of aligning multiple structured evaluation indicators in the evaluation indicator set in a temporal sequence includes: using the Dynamic Time Warping (DTW) algorithm to align multiple structured evaluation indicators in the evaluation indicator set in a temporal sequence, obtaining an alignment path set, the alignment path set including multiple time alignment paths; for any index pair in the first time alignment path, linear interpolation is used to achieve data synchronization, the first time alignment path being any time alignment path in the alignment path set; the step of aligning multiple structured evaluation indicators in the evaluation indicator set in mileage includes: setting multiple mileage points on the line, and using a piecewise cubic spline function to align adjacent mileage points; the step of aligning multiple structured evaluation indicators in spatially includes: using a seven-parameter Helmholtz transform to transform the coordinates of the multiple structured evaluation indicators to the same spatial coordinate system.

[0008] Optionally, the first Time of the first The evaluation vector corresponding to the mileage point is represented in the following way:

[0009] in, For the first Time of the first The evaluation vector corresponding to the mileage point For spatial alignment, the first Time of the first The coordinates of the mileage point For the first Time of the first Mileage point One evaluation indicator, , , All are positive integers. The value range is 1 to , The total number of mileage points. The value range is 1 to , The length of the time series. The value range is 1 to , This represents the total number of evaluation indicators.

[0010] Optionally, the CNN is a three-dimensional CNN, and the CNN extracts the spatial features of the evaluation vector in the following manner:

[0011] in, for Spatial characteristics, For activation function, , This is the time neighborhood offset. This is the spatial neighborhood offset. Indicates the time neighborhood offset as Spatial neighborhood offset is In this case, The Convolutional kernel weights in each dimension, Here is the convolutional kernel weight matrix of the CNN. , For element-wise multiplication, This is the bias of the CNN.

[0012] Secondly, a high-speed rail infrastructure condition assessment device based on integrated space-air-ground system is also provided, comprising: an acquisition module for acquiring a multi-source data set, the multi-source data set including data collected by multi-source sensors for monitoring roads, bridges, tunnels, and tracks; a conversion module for converting the data in the multi-source data set into multiple structured assessment indicators to obtain an assessment indicator set, the structured assessment indicators being used for condition assessment of roads, bridges, tunnels, and tracks; an alignment module for performing time-series alignment, mileage alignment, and spatial alignment on the multiple structured assessment indicators in the assessment indicator set to obtain multiple assessment vectors; a prediction module for processing the multiple assessment vectors using a condition assessment model to obtain prediction results for roads, bridges, tunnels, and tracks, the prediction results including degradation state prediction results and remaining life prediction results; a real-time monitoring module for determining real-time degradation state and real-time remaining life based on real-time monitoring data; and an update module for updating the parameters of the condition assessment model based on the differences between the real-time degradation state, the real-time remaining life, and the prediction results.

[0013] Optionally, the state evaluation model includes a convolutional neural network (CNN), a long short-term memory network (LSTM), a cross-modal feature interaction layer, and a multi-task output layer connected in sequence. The CNN is used to extract spatial features, the LSTM is used to extract temporal features, the cross-modal feature interaction layer is used to fuse the spatial features and the temporal features, and the multi-task output layer is used to output the prediction results for roads, bridges, tunnels, and tracks. The update module is further used to update the forget gate weight matrix parameters of the LSTM when the difference between the real-time degradation state of the roadbed, the real-time remaining lifetime, and the prediction results meets a roadbed difference threshold; update the convolutional kernel parameters of the CNN when the difference between the real-time degradation state of the bridge, the real-time remaining lifetime, and the prediction results meets a bridge difference threshold; update the multi-head attention weights of the cross-modal feature interaction layer when the difference between the real-time degradation state of the tunnel, the real-time remaining lifetime, and the prediction results meets a tunnel difference threshold; and update the fully connected layer weights of the multi-task output layer when the difference between the real-time degradation state of the track, the real-time remaining lifetime, and the prediction results meets a track difference threshold.

[0014] Optionally, the updating module is further configured to: reweight the forget gate weight matrix parameters of the LSTM based on a temporal attention mechanism; update the convolution kernel parameters of the CNN based on gradient pruning; update the multi-head attention weights of the cross-modal feature interaction layer using a federated learning collaborative fine-tuning algorithm; and update the fully connected layer weights of the multi-task output layer using an adaptive momentum optimization algorithm.

[0015] Optionally, the alignment module is further configured to use the Dynamic Time Warping (DTW) algorithm to perform time-series alignment of multiple structured evaluation indicators in the evaluation indicator set to obtain an alignment path set, the alignment path set including multiple time alignment paths; for any index pair in the first time alignment path, linear interpolation is used to achieve data synchronization, the first time alignment path being any time alignment path in the alignment path set; multiple mileage points are set on the line, and piecewise cubic spline functions are used to align adjacent mileage points; a seven-parameter Helmholtz transform is used to transform the coordinates of the multiple structured evaluation indicators to the same spatial coordinate system.

[0016] Thirdly, a computer device is also provided, comprising: a memory and a processor, wherein the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to perform the high-speed rail line infrastructure status assessment method based on the integrated air-space-vehicle-ground system described in the above embodiments.

[0017] Fourthly, a computer-readable storage medium is also provided, wherein at least one computer program is stored in the computer-readable storage medium, and the at least one computer program is loaded and executed by a processor to perform the high-speed rail line infrastructure status assessment method based on the integrated air-space-vehicle-ground system described in the above embodiments.

[0018] Fifthly, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement the method described in the first aspect.

[0019] The beneficial effects of the technical solutions provided in this disclosure include at least the following: In this embodiment, a multi-source data set is acquired; the data in the multi-source data set is converted into multiple structured evaluation indicators to obtain an evaluation indicator set; the multiple structured evaluation indicators in the evaluation indicator set are aligned temporally, mileage-wise, and spatially to obtain multiple evaluation vectors; a state evaluation model is used to process the multiple evaluation vectors to obtain prediction results for roads, bridges, tunnels, and railways, including degradation state prediction results and remaining life prediction results; real-time degradation state and real-time remaining life are determined based on real-time monitoring data; based on the differences between the real-time degradation state, the real-time remaining life, and the prediction results, the parameters of the state evaluation model are updated, thus realizing a dynamic state evaluation and model closed-loop optimization mechanism for high-speed railway infrastructure. Specifically, the state evaluation model enables feature extraction and fusion analysis of multi-source data, establishing an end-to-end mapping relationship from raw data to degradation state and remaining life. The parameters of the condition assessment model are updated based on the differences between the real-time degradation state, the real-time remaining lifetime, and the prediction results. This is equivalent to using the differences between the real-time degradation state, the real-time remaining lifetime, and the prediction results to provide closed-loop feedback to the parameters of the condition assessment model, thereby continuously optimizing the model parameters based on real-time monitoring data and effectively improving the accuracy of the prediction results output by the condition assessment model. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart is shown as an exemplary embodiment of the present disclosure of a method for assessing the condition of high-speed rail line infrastructure based on the integration of air, space, and vehicle-ground systems. Figure 2A flowchart is shown for another exemplary embodiment of the present disclosure of a method for assessing the condition of high-speed rail line infrastructure based on the integration of air, space, and vehicle-ground systems. Figure 3 A schematic diagram of the structure of a high-speed rail line infrastructure condition assessment device based on the integration of air, space, and vehicle-ground systems provided in an exemplary embodiment of this disclosure is shown. Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this disclosure. Detailed Implementation

[0022] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this disclosure pertains. The terms “first,” “second,” “third,” and similar terms used in this patent application specification and claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an” or “a” and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms “comprising” or “including” and similar terms mean that the elements or objects preceding “comprising” or “including” encompass the elements or objects listed following “comprising” or “including” and their equivalents, and do not exclude other elements or objects. The terms “connected” or “linked” and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0023] To make the objectives, technical solutions, and advantages of this disclosure clearer, the embodiments of this disclosure will be described in further detail below with reference to the accompanying drawings.

[0024] Figure 1 A flowchart illustrating an exemplary embodiment of this disclosure provides a method for assessing the condition of high-speed rail line infrastructure based on an integrated space-air-ground system, which can be executed by a computer device. See also... Figure 1 The method includes: In step 101, a multi-source data set is obtained.

[0025] The multi-source dataset includes data collected by multiple sensors used to monitor roads, bridges, tunnels, and tracks.

[0026] Among them, roads, bridges, tunnels and tracks include: high-speed rail tracks, bridges, tunnels and roadbeds.

[0027] The roadbed is the foundation structure of a high-speed railway line, including geotechnical structures such as embankments, cuttings, and slopes. Bridges are elevated structures that span obstacles, including piers, beams, and supports. Tunnels are underground or mountain-crossing structures, including linings, inverts, and drainage systems. Tracks are the load-bearing system for train operation, composed of rails, fasteners, and ballast. All the road, bridge, tunnel, and track components mentioned here are related to high-speed rail.

[0028] Rails are a subset of tracks and are the core load-bearing components of tracks; bridge piers are a subset of bridges and are the vertical support structures of bridges. Track monitoring includes but is not limited to rails (it also includes fasteners, track bed, etc.), while bridge monitoring includes but is not limited to bridge piers (it also includes beams, supports, etc.).

[0029] According to the classification of roads, bridges, tunnels, and tracks, the multi-source data set can include four types of data: roadbed inspection data, track inspection data, bridge inspection data, and tunnel inspection data.

[0030] Roadbed inspection data can be acquired using spaceborne multispectral imaging, distributed optical fiber, and other equipment. For multispectral images acquired by spaceborne multispectral imaging, the slope moisture content of the roadbed can be retrieved using NDVI (Normalized Difference Vegetation Index) (spatial resolution 2m). For distributed optical fiber, the roadbed strain distribution can be monitored based on OTDR (Optical Time-Domain Reflectometer) (spatial positioning accuracy 0.5m). Preprocessing of the roadbed inspection data includes atmospheric correction of multispectral images and fiber optic strain moving average.

[0031] Bridge inspection data can be acquired using equipment such as spaceborne InSAR (Interferometric Synthetic Aperture Radar), airborne LiDAR (Light Detection and Ranging), and GNSS (Global Navigation Satellite System) monitoring stations. For point cloud data acquired by spaceborne InSAR, differential interferometry can be used to extract pier settlement sequence data (accuracy ±2mm). For point cloud data acquired by airborne LiDAR, the Time-of-Flight (ToF) method can be used to construct a 3D point cloud of the bridge surface, automatically marking the location and width of cracks on the bridge surface. For satellite signals acquired by GNSS monitoring stations, carrier phase calculation can be used to achieve dynamic displacement monitoring of the bridge (horizontal accuracy 1mm, vertical accuracy 2mm). Preprocessing of bridge inspection data includes InSAR atmospheric delay correction and LiDAR point cloud radius filtering.

[0032] Tunnel inspection data can be collected using equipment such as UAV infrared thermal imagers and vehicle-mounted laser scanners. For infrared image data collected by UAV infrared thermal imagers, the lining voids in the tunnel can be identified based on the principle of radiation thermometry (temperature difference threshold 2℃). For point cloud data collected by vehicle-mounted laser scanners, tunnel clearance deformation can be detected using laser triangulation (axial resolution 5mm). Preprocessing of tunnel inspection data includes infrared thermal image non-uniformity correction and lidar intensity normalization.

[0033] Track inspection data can be acquired using equipment such as vehicle-mounted ultrasonic rail flaw detectors, track geometry inspection systems, and trackside fiber optic sensors. For ultrasonic data acquired by vehicle-mounted ultrasonic rail flaw detectors, internal defects in the rail head / web can be detected using the pulse-echo method, generating damage levels (0-4) and location coordinates. For data acquired by the track geometry inspection system, deformation parameters such as gauge, level, and elevation can be output (with millimeter-level accuracy) based on inertial reference methods and photoelectric ranging. For sensor data acquired by trackside fiber optic sensors, the longitudinal stress variation curve of the rail can be analyzed using the Brillouin scattering principle (sampling frequency 1kHz). Preprocessing of track inspection data includes wavelet packet denoising of the ultrasonic rail data and Kalman filtering of the track geometry data.

[0034] After preprocessing the above four types of data, the signal-to-noise ratio and spatial resolution (such as signal-to-noise ratio) of the above four types of data can be unified. With a spatial resolution of 0.05m, it lays the data foundation for subsequent fusion analysis. The collection of the above four types of data involves multiple sensors including satellites, drones, trains, and ground-based sensors, enabling the integrated collection of multi-source data from high-speed rail infrastructure, thereby obtaining a multi-source data set.

[0035] The multi-source data set in this embodiment is historical multi-source data of roads, bridges, tunnels and tracks, which can be used to train subsequent state assessment models.

[0036] In step 102, the data in the multi-source dataset is converted into multiple structured evaluation metrics to obtain an evaluation metric set.

[0037] Structured evaluation indicators are used to assess the condition of roads, bridges, tunnels, and railways.

[0038] In this embodiment of the disclosure, the structured evaluation indicators are classified according to road, bridge, tunnel, and track, resulting in four categories: structured evaluation indicators for roadbeds, bridges, tunnels, and tracks. Examples of these four categories of structured evaluation indicators are given below.

[0039] (1) The structural evaluation indicators corresponding to the roadbed include: settlement acceleration and slope stability coefficient.

[0040] (2) The structural evaluation indicators for bridges include: support displacement warning value and crack propagation rate.

[0041] (3) The structural evaluation indicators corresponding to the tunnel include: lining defect density and convergence deformation.

[0042] (4) The structured evaluation indicators corresponding to the track include: rail damage index and track quality index (TQI).

[0043] After obtaining the multi-source data set, any method from relevant technologies can be used to transform the multi-source data set into the aforementioned structured evaluation indicators.

[0044] In step 103, multiple structured evaluation indicators in the evaluation indicator set are aligned in terms of time sequence, mileage, and space to obtain multiple evaluation vectors.

[0045] Optionally, aligning multiple structured evaluation indicators in the evaluation indicator set in a time sequence involves the following two steps.

[0046] The first step is to use a dynamic time warping algorithm to perform time-series alignment of multiple structured evaluation indicators in the evaluation indicator set, thereby obtaining an alignment path set.

[0047] The alignment path set includes multiple time alignment paths.

[0048] Dynamic Time Warping (DTW) algorithms align time series by calculating the optimal matching path between two time series, resulting in an aligned time path. Data from different sensors or monitoring devices belong to different time series in the time dimension. After transforming multi-source data into evaluation metrics, data at each time point can also be transformed into evaluation metrics. This means that evaluation metrics derived from multi-source data from different sensors also belong to different time series in the time dimension. Therefore, the DTW algorithm can be used here to perform time-series alignment of multiple structured evaluation metrics from a multi-source evaluation metric set.

[0049] After time series alignment based on DTW, multiple alignment paths are obtained. Each alignment path includes multiple index pairs. An index pair includes data from two different time series. An index pair is actually a time point obtained by aligning data from two different time series.

[0050] The second step is to use linear interpolation to synchronize data for any index pair in the first-time alignment path.

[0051] The first time alignment path is any time alignment path in the set of alignment paths.

[0052] For example, linear interpolation is achieved using formula (1).

[0053] (1) In formula (1), This is an index pair in the first-time alignment path. , These are data from different time series. , These are the interpolation weighting coefficients. , These are interpolated index pairs, which are the time-aligned data.

[0054] The time path obtained by the DTW algorithm is actually composed of multiple time points with certain intervals between them, so the resulting time curve is not continuous. By performing linear interpolation on the time points in the time path, the time path is transformed into a continuous and smooth time curve, thereby achieving time sequence alignment.

[0055] Optionally, multiple structured evaluation indicators in the evaluation indicator set can be aligned by mileage, including setting multiple mileage points on the line and using a piecewise cubic spline function to align adjacent mileage points.

[0056] For example, a mileage interval value can be preset, and multiple mileage points can be set on the line based on this mileage interval value. Here, since this embodiment of the disclosure is for condition assessment of high-speed rail line infrastructure, the line here is the high-speed rail line to be studied. For example, the mileage interval value can be 500 meters, for example, mileage points can be set every 500 meters along the line. In addition, it is necessary to record the GNSS coordinates of each mileage point, the mileage value of each mileage point on the line (e.g., the mileage distance from the starting point of the line to the mileage point), and the multiple evaluation indicators included in the mileage point.

[0057] In implementation, a Linear Referencing System (LRS) can be used to implement mileage-based data encoding and control. Each mileage can be encoded as: Line Number + Mileage Sequence Number + Structure Type Code.

[0058] The structure type code indicates which structure (road, bridge, tunnel, or track) the mileage point belongs to. For example, the structure type codes are: 01 (road), 02 (bridge), 03 (tunnel), 04 (track).

[0059] For example, the bridge at the 500th mileage point of the K123 line of the Beijing-Shanghai High-Speed ​​Railway can be coded as JHG123-500-02, where JHG123 represents the K123 line of the Beijing-Shanghai High-Speed ​​Railway, 500 represents the 500th mileage point, and 02 represents the bridge.

[0060] A mileage point can be abstractly understood as a spatial location that may include any one of the following structures: road, bridge, tunnel, or track, or multiple structures (e.g., establishing a mileage point that includes a tunnel and the track within the tunnel). In this case, when encoding the mileage point, a numerical value for the structure type can be added. For example, if the 500th mileage point on the Beijing-Shanghai High-Speed ​​Railway K123 includes both a tunnel and track, then this spatial location can be encoded as JHG123-500-03-04.

[0061] After setting multiple mileage points, the data collected by each sensor or monitoring device can actually be inserted into the mileage point. The data collected by the sensor or monitoring device at the same mileage point is equivalent to achieving mileage alignment. Furthermore, the evaluation indicators calculated from these mileage-aligned data can also achieve mileage alignment.

[0062] However, simply aligning mileage points is insufficient, as this type of alignment is discontinuous and interrupted. Therefore, interpolation is required between adjacent mileage points, inserting multiple interpolated mileage points to achieve continuous mileage alignment. In this embodiment, a piecewise cubic spline function is used to insert multiple interpolated mileage points between adjacent mileage points to achieve continuous mileage alignment. This process can be represented by formula (2).

[0063] (2) In formula (2), For use at two adjacent mileage points Piecewise cubic spline function for interpolation within an interval These are the spline coefficients, which can be solved using boundary conditions.

[0064] For example, the boundary conditions are expressed by formula (3).

[0065] (3) In formula (3), Mileage point The corresponding multiple evaluation indicators, Mileage point Corresponding evaluation indicators For use at two adjacent mileage points Piecewise cubic spline function for interpolation within an interval This demonstrates that the first derivative of the piecewise cubic spline function is continuous. This demonstrates the continuity of the second derivative of the piecewise cubic spline function. The meanings of the other parameters in formula (3) are the same as those in formula (2), and are omitted here for further explanation.

[0066] Optionally, multiple structured evaluation indicators in the evaluation indicator set can be spatially aligned, including: using a seven-parameter Helmert transform to convert the coordinates of multiple structured evaluation indicators to the same spatial coordinate system, such as the Earth-Centered Earth-Fixed (ECEF) coordinate system.

[0067] The coordinates of the evaluation index are the coordinates of the sensor or monitoring device that calculates the evaluation index. Different sensors or monitoring devices may output coordinate formats that differ. Through the seven-parameter Helmholtz transformation, the coordinates of multiple structured evaluation indices can be transformed to the same spatial coordinate system.

[0068] There are many implementation methods for the seven-parameter Helmert transform, which will not be detailed here.

[0069] After time-series alignment, mileage alignment, and spatial alignment, evaluation metrics for the same mileage point at the same time belong to the same evaluation vector. In this way, multiple evaluation vectors can be obtained.

[0070] For example, the first Time of the first The evaluation vector corresponding to the mileage point can be represented by formula (4).

[0071] (4) in, For the first Time of the first The evaluation vector corresponding to the mileage point For spatial alignment, the first Time of the first The coordinates of the mileage point For the first Time of the first Mileage point One evaluation indicator, , , All are positive integers. The value range is 1 to , The total number of mileage points. The value range is 1 to , The length of the time series. The value range is 1 to , This represents the total number of evaluation indicators.

[0072] All evaluation vectors can form the input tensor of the state evaluation model, which is represented by formula (5).

[0073] (5) In formula (5), For the input tensor of the state evaluation model, the meanings of the other parameters in formula (5) are the same as those in formula (4), and are omitted here.

[0074] In this embodiment, an integrated air-ground-vehicle monitoring system is constructed. Through the collaborative configuration of multi-platform, multi-scale sensors, the system achieves full-element, all-weather status awareness of high-speed rail infrastructure. By establishing a unified data preprocessing workflow, the system achieves standardized conversion and quality improvement of multi-source heterogeneous data, providing a reliable data foundation for subsequent analysis.

[0075] In step 104, a state assessment model is used to process multiple assessment vectors to obtain the prediction results for roads, bridges, tunnels, and tracks.

[0076] The prediction results include degradation status prediction results and remaining lifetime prediction results.

[0077] Optionally, the state evaluation model includes a convolutional neural network (CNN), a long short-term memory network (LSTM), a cross-modal feature interaction layer, and a multi-task output layer connected in sequence. The CNN is used to extract spatial features, the LSTM is used to extract temporal features, the cross-modal feature interaction layer is used to fuse spatial and temporal features, and the multi-task output layer is used to output the prediction results of roads, bridges, tunnels, and railways.

[0078] Optionally, the CNN is a three-dimensional CNN. In this case, the CNN uses the following formula (6) to extract the spatial features of the evaluation vector.

[0079] (6) In formula (6), for Spatial characteristics, For activation function, , This is the time neighborhood offset. This is the spatial neighborhood offset. Indicates the time neighborhood offset as Spatial neighborhood offset is In this case, The Convolutional kernel weights in each dimension, This is the convolutional kernel weight matrix of the CNN. , For element-wise multiplication, This is the bias of the CNN.

[0080] Based on LSTM, temporal feature extraction and degradation process modeling can be achieved.

[0081] In LSTM, the forget gate controls the memory strength of historical states, thereby preserving long-term degradation trends (such as rail fatigue accumulation); the input gate adjusts the weights of current features, effectively capturing short-term mutations (such as sudden track geometric deformation); and the output gate determines the output ratio of the hidden state, achieving dynamic feature focusing.

[0082] LSTM's memory cells fuse historical and current information through a gating mechanism to form a temporal degradation trajectory of infrastructure, enabling modeling of the degradation process. Examples include the gradual change in bridge support displacement and the propagation rate of tunnel lining cracks.

[0083] The cross-modal feature interaction layer achieves the fusion of spatial and temporal features through a multi-head attention mechanism to obtain fused features. These fused features are input into the multi-task output layer to obtain the prediction results for roads, bridges, tunnels, and railways. The prediction results include the degradation state prediction results and the remaining life prediction results.

[0084] For example, the degradation state includes 5 levels (more levels can be set as needed). The multi-task output layer can output the degradation state prediction result by outputting the probability distribution of the road, bridge, tunnel and rail at each degradation state. In addition, the degradation state prediction result includes the degradation state of the road, bridge, tunnel and rail at each time point in the future.

[0085] Similarly, the remaining life prediction results include the remaining life of roads, bridges, tunnels, and rails at every moment over a future period.

[0086] In the prediction results output by the state prediction model, each mileage point will output a quaternary prediction result for roads, bridges, tunnels, and tracks. For structures included at the mileage point, the corresponding part of the quaternary prediction result will be output as the corresponding prediction result; for structures not included at the mileage point, the corresponding part of the quaternary prediction result will be set to 0. For example, if a mileage point includes tunnels and tracks but does not include roadbeds and bridges, then the roadbed and bridge parts of the quaternary prediction result for that mileage point will be set to 0.

[0087] For different degradation states and degradation lives of roads, bridges, tunnels, and railways, some maintenance suggestions can be preset, and these suggestions can be output simultaneously with the prediction results. For example, when the state prediction model predicts that the degradation state of the roadbed is at a certain level, it can output the corresponding maintenance suggestions for that level.

[0088] In step 105, the real-time degradation status and real-time remaining lifetime are determined based on real-time monitoring data.

[0089] Here, the acquisition methods for real-time monitoring data and multi-source datasets are exactly the same. The difference is that the data in the multi-source dataset is historical data, while the real-time monitoring data is data collected in real time. Similarly, the real-time monitoring data also needs to be processed in steps 101 to 103 before it can be used to calculate the real-time degradation state and real-time remaining lifetime.

[0090] After obtaining multiple evaluation vectors based on real-time monitoring data, the real-time degradation status and remaining lifespan of roads, bridges, tunnels, and rails can be determined by setting preset thresholds. However, this method can only determine the degradation status and remaining lifespan corresponding to current or historical data, and cannot predict the degradation status and remaining lifespan of roads, bridges, tunnels, and rails, such as predicting the degradation status and remaining lifespan of roads, bridges, tunnels, and rails in the future. Therefore, in this embodiment of the present disclosure, steps 101 and 104 are used to predict the degradation status and remaining lifespan of roads, bridges, tunnels, and rails in the future.

[0091] In step 106, the parameters of the state assessment model are updated based on the differences between the real-time degradation state, the real-time remaining lifetime, and the prediction results.

[0092] In this embodiment, steps 101 to 106 establish a dynamic state assessment and closed-loop optimization mechanism for high-speed rail infrastructure. Specifically, the state assessment model enables feature extraction and fusion analysis of multi-source data, establishing an end-to-end mapping from raw data to degradation state and remaining lifetime. Based on the differences between real-time degradation state, real-time remaining lifetime, and prediction results, the parameters of the state assessment model are updated. This is equivalent to using the differences between real-time degradation state, real-time remaining lifetime, and prediction results to provide closed-loop feedback to the parameters of the state assessment model, achieving continuous optimization of model parameters based on real-time monitoring data and effectively improving the accuracy of the prediction results output by the state assessment model.

[0093] Figure 2 A flowchart illustrating a high-speed rail line infrastructure condition assessment method based on an integrated space-air-ground system provided in another exemplary embodiment of this disclosure is shown. This method can be executed by a computer device. See also Figure 1 The method includes: In step 201, a multi-source data set is obtained.

[0094] The multi-source dataset includes data collected by multiple sensors for monitoring roads, bridges, tunnels, and tracks; In step 202, the data in the multi-source dataset is converted into multiple structured evaluation indicators to obtain an evaluation indicator set.

[0095] Structured evaluation indicators are used to assess the condition of roads, bridges, tunnels, and railways.

[0096] In step 203, multiple structured evaluation indicators in the evaluation indicator set are aligned in terms of time sequence, mileage, and space to obtain multiple evaluation vectors.

[0097] In step 204, a state assessment model is used to process multiple assessment vectors to obtain the prediction results for roads, bridges, tunnels, and tracks.

[0098] The prediction results include degradation status prediction results and remaining lifetime prediction results.

[0099] In step 205, the real-time degradation status and real-time remaining lifetime are determined based on real-time monitoring data.

[0100] The relevant content of steps 201 to 205 is the same as that of steps 101 to 105 mentioned above, and will not be described in detail here.

[0101] In step 206, the parameters of the state assessment model are updated based on the differences between the real-time degradation state, the real-time remaining lifetime, and the prediction results.

[0102] Optionally, step 206 includes the following four cases.

[0103] (1) When the difference between the real-time degradation state of the roadbed, the real-time remaining life and the prediction result meets the roadbed difference threshold, update the forget gate weight matrix parameters of LSTM.

[0104] In implementation, the forget gate weight matrix parameters of the LSTM can be reweighted based on the temporal attention mechanism to update the forget gate weight matrix parameters of the LSTM.

[0105] Subgrade degradation and lifespan are significantly affected by long-term deformation. If the difference between the real-time degradation status and real-time remaining life of the subgrade and the predicted results meets the subgrade difference threshold, it indicates that the model parameters related to the long-term degradation trend in the condition assessment model need to be adjusted, specifically the parameters of the LSTM forget gate weight matrix. Updating the LSTM forget gate weight matrix parameters can enhance the sensitivity to slow deterioration trends such as settlement and slope slippage.

[0106] (2) Update the CNN convolution kernel parameters when the difference between the bridge’s real-time degradation state, real-time remaining life and prediction results meets the bridge difference threshold.

[0107] In implementation, the convolution kernel parameters of the CNN can be updated based on gradient clipping.

[0108] Bridge degradation and lifespan are significantly affected by localized damage, which is reflected in spatial features. If the difference between the bridge's real-time degradation status, real-time remaining lifespan, and predicted results meets the bridge difference threshold, it indicates that the relevant parameters of the spatial features in the condition assessment model need adjustment. Updating the convolutional kernel parameters of the CNN can adjust the condition assessment model's ability to extract spatial features, improving the accuracy of identifying localized damage such as cracks and support displacements.

[0109] (3) When the difference between the real-time degradation state of the tunnel, the real-time remaining lifetime and the prediction result meets the tunnel difference threshold, update the multi-head attention weight of the cross-modal feature interaction layer.

[0110] In implementation, federated learning collaborative fine-tuning algorithms (such as the FedProx algorithm) can be used to update the multi-head attention weights of the cross-modal feature interaction layer.

[0111] Tunnel degradation and lifespan require consideration of multiple parameters, such as seepage and deformation data; that is, tunnel degradation and lifespan are directly related to multi-parameter coupling. If the difference between the tunnel's real-time degradation state, real-time remaining lifespan, and predicted results meets a tunnel difference threshold, it indicates that the model parameters related to multi-parameter coupling in the state assessment model need adjustment, specifically the multi-head attention weights in the cross-modal feature interaction layer. Updating the multi-head attention weights in the cross-modal feature interaction layer can optimize the correlation analysis capability between seepage, deformation, and convergence.

[0112] (4) When the difference between the real-time degradation state of the orbit, the real-time remaining lifetime and the prediction result meets the orbit difference threshold, update the weights of the fully connected layer of the multi-task output layer.

[0113] In implementation, an adaptive momentum optimization algorithm (such as the AdamW algorithm) can be used to update the weights of the fully connected layers in the multi-task output layer.

[0114] Tracks involve multiple components (such as rails, ballast, fasteners, etc.). Typically, different weights are assigned to these components in the fully connected layer of the multi-task output layer to balance their impact on the track. If the difference between the track's real-time degradation state, real-time remaining life, and the predicted results meets the track difference threshold, it indicates that the weight settings for the multiple track components in the fully connected layer of the multi-task output layer are unreasonable. The weights of the fully connected layer in the multi-task output layer need to be updated to calibrate the overall weights of the rails, ballast, fasteners, and other components in the track.

[0115] Depending on whether the output of the state evaluation model satisfies the above four conditions, an appropriate update strategy can be selected. For example, if the output of the state evaluation model satisfies conditions (1) and (2), the relevant update strategies in conditions (1) and (2) (updating the forget gate weight matrix parameters of the LSTM and updating the convolution kernel parameters of the CNN) are used to update the parameters of the state evaluation model.

[0116] In this embodiment, steps 201 to 206 establish a dynamic state assessment and closed-loop optimization mechanism for high-speed rail infrastructure. The state assessment model enables feature extraction and fusion analysis of multi-source data, establishing an end-to-end mapping relationship from raw data to degradation state and remaining lifetime. Specifically, the parameters of the state assessment model are updated based on the differences between the real-time degradation state, real-time remaining lifetime, and prediction results. This is equivalent to a closed-loop feedback mechanism between the differences in real-time degradation state, real-time remaining lifetime, and prediction results, continuously optimizing the model parameters based on real-time monitoring data and maintaining the accuracy of the assessment results.

[0117] The following are device embodiments of this application. For details not described in detail in the device embodiments, please refer to the above method embodiments.

[0118] Figure 3 A schematic diagram of the structure of a high-speed rail line infrastructure condition assessment device based on an exemplary embodiment of this disclosure is shown. See also: Figure 3 The high-speed rail line infrastructure status assessment device 300 based on the integration of air, space, vehicle and ground includes: an acquisition module 301, a conversion module 302, an alignment module 303, a prediction module 304, a real-time monitoring module 305 and an update module 306.

[0119] The acquisition module 301 is used to acquire a multi-source data set, which includes data collected by multi-source sensors for monitoring roads, bridges, tunnels and tracks.

[0120] The conversion module 302 is used to convert the data in the multi-source data set into multiple structured evaluation indicators to obtain an evaluation indicator set, which is used to evaluate the condition of roads, bridges, tunnels and railways.

[0121] The alignment module 303 is used to perform time-series alignment, mileage alignment, and spatial alignment on multiple structured evaluation indicators in the evaluation indicator set to obtain multiple evaluation vectors.

[0122] The prediction module 304 is used to process the multiple evaluation vectors using a state assessment model to obtain prediction results for roads, bridges, tunnels and railways. The prediction results include degradation state prediction results and remaining life prediction results.

[0123] The real-time monitoring module 305 is used to determine the real-time degradation status and real-time remaining life based on real-time monitoring data.

[0124] The update module 306 is used to update the parameters of the state assessment model based on the difference between the real-time degradation state, the real-time remaining lifetime, and the prediction result.

[0125] Optionally, the state evaluation model includes a convolutional neural network (CNN), a long short-term memory network (LSTM), a cross-modal feature interaction layer, and a multi-task output layer connected in sequence. The CNN is used to extract spatial features, the LSTM is used to extract temporal features, the cross-modal feature interaction layer is used to fuse the spatial features and the temporal features, and the multi-task output layer is used to output the prediction results for roads, bridges, tunnels, and tracks. The update module 306 is further used to update the forget gate weight matrix parameters of the LSTM when the difference between the real-time degradation state of the roadbed, the real-time remaining lifetime, and the prediction results meets the roadbed difference threshold; update the convolutional kernel parameters of the CNN when the difference between the real-time degradation state of the bridge, the real-time remaining lifetime, and the prediction results meets the bridge difference threshold; update the multi-head attention weights of the cross-modal feature interaction layer when the difference between the real-time degradation state of the tunnel, the real-time remaining lifetime, and the prediction results meets the tunnel difference threshold; and update the fully connected layer weights of the multi-task output layer when the difference between the real-time degradation state of the track, the real-time remaining lifetime, and the prediction results meets the track difference threshold.

[0126] Optionally, the updating module 306 is further configured to: reweight the forget gate weight matrix parameters of the LSTM based on a temporal attention mechanism; update the convolution kernel parameters of the CNN based on gradient pruning; update the multi-head attention weights of the cross-modal feature interaction layer using a federated learning collaborative fine-tuning algorithm; and update the fully connected layer weights of the multi-task output layer using an adaptive momentum optimization algorithm.

[0127] Optionally, the alignment module 303 is further configured to use the Dynamic Time Warping (DTW) algorithm to perform time-series alignment of multiple structured evaluation indicators in the evaluation indicator set to obtain an alignment path set, the alignment path set including multiple time alignment paths; to use linear interpolation to achieve data synchronization for any index pair in the first time alignment path, the first time alignment path being any time alignment path in the alignment path set; to set multiple mileage points on the line, and to use a piecewise cubic spline function to achieve mileage alignment for adjacent mileage points; and to use a seven-parameter Helmholtz transform to transform the coordinates of the multiple structured evaluation indicators to the same spatial coordinate system.

[0128] It should be noted that the above-described embodiments of the high-speed rail infrastructure status assessment device based on the integrated space-air-vehicle-ground system are only illustrative examples of the functional module divisions. In practical applications, the functions can be assigned to different functional modules as needed, i.e., the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the high-speed rail infrastructure status assessment device based on the integrated space-air-vehicle-ground system and the high-speed rail infrastructure status assessment method based on the integrated space-air-vehicle-ground system belong to the same concept, and their specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0129] The module division in this embodiment is illustrative and represents only one logical functional division. In actual implementation, other division methods are possible. Furthermore, the functional modules in the various embodiments of this disclosure can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0130] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a terminal device (which may be a personal computer, mobile phone, or communication device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0131] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this disclosure. For example... Figure 4 As shown, the computer device 400 includes a processor 401 and a memory 402.

[0132] Processor 401 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 401 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 401 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 401 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 401 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0133] The memory 402 may include one or more computer-readable storage media, which may be non-transitory. The memory 402 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 402 is used to store at least one instruction, which is executed by the processor 401 to implement the high-speed rail line infrastructure condition assessment method based on integrated air-space-vehicle-ground systems provided in this disclosure embodiment.

[0134] Those skilled in the art will understand that Figure 4 The structure shown does not constitute a limitation on computer device 400, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0135] This disclosure also provides a non-transitory computer-readable storage medium, which, when the instructions in the storage medium are executed by the processor of a computer device, enables the computer device to execute the high-speed rail line infrastructure status assessment method based on the integrated air-space-vehicle-ground system provided in this disclosure.

[0136] This disclosure also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the high-speed rail line infrastructure status assessment method based on the integration of air, space, vehicle, and ground provided in this disclosure.

[0137] The above description is merely an optional embodiment of this disclosure and is not intended to limit this disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.

Claims

1. A method for assessing the condition of high-speed railway infrastructure based on integrated air-space-vehicle-ground systems, characterized in that, The method includes: Acquire a multi-source data set, which includes data collected by multi-source sensors for monitoring roads, bridges, tunnels and tracks, including high-speed railway tracks, bridges, tunnels and roadbeds; The data in the multi-source data set is converted into multiple structured evaluation indicators to obtain an evaluation indicator set, which is used to evaluate the condition of roads, bridges, tunnels and railways. After performing time-series alignment, mileage alignment, and spatial alignment on multiple structured evaluation indicators in the set of evaluation indicators, multiple evaluation vectors are obtained. The multiple evaluation vectors are processed using a state assessment model to obtain prediction results for roads, bridges, tunnels, and railways. The prediction results include degradation state prediction results and remaining life prediction results. Real-time degradation status and real-time remaining lifetime are determined based on real-time monitoring data; The parameters of the state assessment model are updated based on the difference between the real-time degradation state, the real-time remaining lifetime, and the prediction result. The state evaluation model comprises a convolutional neural network (CNN), a long short-term memory network (LSTM), a cross-modal feature interaction layer, and a multi-task output layer, connected sequentially. The CNN is used to extract spatial features, the LSTM is used to extract temporal features, the cross-modal feature interaction layer is used to fuse the spatial and temporal features, and the multi-task output layer is used to output the prediction results for roads, bridges, tunnels, and railways. The step of updating the parameters of the state assessment model based on the difference between the real-time degradation state, the real-time remaining lifetime, and the prediction result includes: If the difference between the real-time degradation state of the roadbed, the real-time remaining lifetime, and the prediction result meets the roadbed difference threshold, the forget gate weight matrix parameters of the LSTM are updated. If the difference between the bridge's real-time degradation state, the real-time remaining lifetime, and the prediction result meets the bridge difference threshold, the convolution kernel parameters of the CNN are updated. If the difference between the real-time degradation state of the tunnel, the real-time remaining lifetime, and the prediction result meets the tunnel difference threshold, update the multi-head attention weights of the cross-modal feature interaction layer. If the difference between the real-time degradation state of the orbit, the real-time remaining lifetime, and the prediction result meets the orbit difference threshold, the weights of the fully connected layer of the multi-task output layer are updated.

2. The method according to claim 1, characterized in that, The step of updating the forget gate weight matrix parameters of the LSTM includes: reweighting the forget gate weight matrix parameters of the LSTM based on a temporal attention mechanism; The updating of the convolution kernel parameters of the CNN includes: updating the convolution kernel parameters of the CNN based on gradient clipping. The step of updating the multi-head attention weights of the cross-modal feature interaction layer includes: updating the multi-head attention weights of the cross-modal feature interaction layer using a federated learning collaborative fine-tuning algorithm; The step of updating the fully connected layer weights of the multi-task output layer includes: updating the fully connected layer weights of the multi-task output layer using an adaptive momentum optimization algorithm.

3. The method according to claim 2, characterized in that, The step of aligning multiple structured evaluation indicators in the evaluation indicator set according to time sequence includes: The Dynamic Time Warping (DTW) algorithm is used to perform time-series alignment of multiple structured evaluation indicators in the evaluation indicator set to obtain an alignment path set, which includes multiple time alignment paths. For any index pair in the first time alignment path, linear interpolation is used to achieve data synchronization, where the first time alignment path is any time alignment path in the set of alignment paths. The step of mileage alignment of multiple structured evaluation metrics in the evaluation metric set includes: Multiple mileage points are set on the line, and mileage alignment is achieved by using a piecewise cubic spline function for adjacent mileage points; The step of spatially aligning multiple structured evaluation indicators in the set of evaluation indicators includes: The coordinates of the multiple structured evaluation indicators are transformed to the same spatial coordinate system using a seven-parameter Helmert transform.

4. The method according to claim 3, characterized in that, No. Time of the first The evaluation vector corresponding to the mileage point is represented in the following way: in, For the first Time of the first The evaluation vector corresponding to the mileage point For the space alignment of the first Time of the first The coordinates of the mileage point For the first Time of the first Mileage point One evaluation indicator, , , All are positive integers. The value range is 1 to , The total number of mileage points. The value range is 1 to , The length of the time series. The value range is 1 to , This represents the total number of evaluation indicators.

5. The method according to claim 4, wherein the CNN is a three-dimensional CNN, and the CNN extracts spatial features of the evaluation vector in the following manner: in, for Spatial characteristics, For activation function, , This is the time neighborhood offset. This is the spatial neighborhood offset. Indicates the time neighborhood offset as Spatial neighborhood offset is In this case, The Convolutional kernel weights in each dimension, Here is the convolutional kernel weight matrix of the CNN. , For element-wise multiplication, This is the bias of the CNN.

6. A high-speed railway line infrastructure condition assessment device based on integrated space-air-ground system, characterized in that, The device includes: The acquisition module is used to acquire a multi-source data set, which includes data collected by multi-source sensors for monitoring roads, bridges, tunnels and tracks, including high-speed railway tracks, bridges, tunnels and roadbeds; The conversion module is used to convert the data in the multi-source data set into multiple structured evaluation indicators to obtain an evaluation indicator set, which is used to evaluate the status of roads, bridges, tunnels and railways. The alignment module is used to perform time-series alignment, mileage alignment, and spatial alignment on multiple structured evaluation indicators in the evaluation indicator set to obtain multiple evaluation vectors. The prediction module is used to process the multiple evaluation vectors using a state assessment model to obtain prediction results for roads, bridges, tunnels and railways. The prediction results include degradation state prediction results and remaining life prediction results. The real-time monitoring module is used to determine the real-time degradation status and real-time remaining lifetime based on real-time monitoring data; An update module is used to update the parameters of the state assessment model based on the difference between the real-time degradation state, the real-time remaining lifetime, and the prediction result; The state evaluation model comprises a convolutional neural network (CNN), a long short-term memory network (LSTM), a cross-modal feature interaction layer, and a multi-task output layer, connected sequentially. The CNN is used to extract spatial features, the LSTM is used to extract temporal features, the cross-modal feature interaction layer is used to fuse the spatial and temporal features, and the multi-task output layer is used to output the prediction results for roads, bridges, tunnels, and railways. The update module is also used for: If the difference between the real-time degradation state of the roadbed, the real-time remaining lifetime, and the prediction result meets the roadbed difference threshold, the forget gate weight matrix parameters of the LSTM are updated. If the difference between the bridge's real-time degradation state, the real-time remaining lifetime, and the prediction result meets the bridge difference threshold, the convolution kernel parameters of the CNN are updated. If the difference between the real-time degradation state of the tunnel, the real-time remaining lifetime, and the prediction result meets the tunnel difference threshold, update the multi-head attention weights of the cross-modal feature interaction layer. If the difference between the real-time degradation state of the orbit, the real-time remaining lifetime, and the prediction result meets the orbit difference threshold, the weights of the fully connected layer of the multi-task output layer are updated.

7. A computer device, characterized in that, The computer device includes a memory and a processor, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to implement the method of any one of claims 1 to 5.

9. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the method described in any one of claims 1 to 5.

Images