Method, system, equipment and method for intelligently evaluating airport runway based on multi-source data

Abstract

The invention relates to an airport runway intelligent evaluation method, system, device and method based on multi-source data, and the method comprises the steps: scanning the runway surface of an airport runway through a laser radar, resolving and generating a runway surface three-dimensional point cloud model, and carrying out the point cloud analysis; the method comprises the following steps of: scanning a road surface by using camera equipment to capture road surface image information of an airport road, intelligently identifying the road surface image information, marking a damage position on the road, and further calculating a road condition index PCI (Peripheral Component Interconnect); constructing a multi-source data fusion field expert evaluation model, importing a point cloud analysis result, a field condition index PCI and field basic data into the model, and outputting a field comprehensive evaluation index; and introducing a spatio-temporal interpolation algorithm and gradient analysis, and analyzing the comprehensive evaluation index of the field channel to obtain a spatio-temporal change prediction result of the condition of the field channel. According to the method, various advanced technologies and algorithms are comprehensively applied, and the problems of singleness, static performance and low efficiency of a traditional airport track evaluation method are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of airport runway safety assessment, and in particular to an airport runway intelligent assessment method, system, equipment and method based on multi-source data. Background Art

[0002] In the field of civil aviation, with the rapid development of the aviation industry and the increasing requirements for flight safety, the safety assessment of airport runways has become particularly critical. As the factors affecting runway safety gradually increase, the requirements for runway safety also increase. Traditional runway assessment methods are usually based on a single evaluation indicator. However, this assessment method cannot fully consider the multiple characteristics of runways due to its singleness. Therefore, this traditional assessment is difficult to provide an objective view of the actual situation of airport runways.

[0003] More importantly, the condition of airport runways changes dynamically in real time. This dynamism requires that the evaluation method can capture and reflect any changes in runway conditions in a timely manner. However, traditional evaluation methods have significant deficiencies in this regard, and their evaluation results are often static and cannot effectively respond to real-time changes in runway conditions. This leads to inefficiency in the evaluation and cannot meet the current needs for real-time monitoring and intelligent analysis of airport runways. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides an airport runway intelligent assessment method, system, equipment and method based on multi-source data, which solves the technical problems that traditional airport runway assessment methods are single, static and inefficient, and cannot meet the needs of modern airports for real-time monitoring and intelligent analysis.

[0006] (II) Technical solution

[0007] In order to achieve the above object, the main technical solutions adopted by the present invention include:

[0008] In a first aspect, an embodiment of the present invention provides an airport runway intelligent assessment method based on multi-source data, comprising:

[0009] Scanning the pavement of the airport runway by laser radar, solving and generating a three-dimensional point cloud model of the pavement, and performing at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result;

[0010] Use camera equipment to scan the pavement and capture the pavement image information of the airport runway, intelligently identify the microscopic road defects on the pavement image information, mark the damaged locations on the runway, and then calculate the runway condition index PCI;

[0011] Based on the AHP and TOPSIS algorithms, a multi-source data fusion field and track expert evaluation model is constructed. The point cloud analysis results, field and track condition index PCI and acquired field and track basic data are imported into the multi-source data field and track expert evaluation model. The expert knowledge graph library obtained through graph neural network training is weighted averaged to output the field and track comprehensive evaluation index.

[0012] The spatiotemporal interpolation algorithm and gradient analysis are introduced to analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and the spatiotemporal change prediction results of the field and track conditions are obtained.

[0013] Optionally, the pavement of the airport runway is scanned by a laser radar, a pavement 3D point cloud model is solved and generated, and at least one point cloud analysis is performed based on the pavement 3D point cloud model, and the point cloud analysis results obtained include:

[0014] Control the precision laser radar to scan the pavement of the airport runway, and receive and store the scan files wirelessly transmitted by the laser radar;

[0015] Utilize the pre-deployed solution function module and solution software to solve the received scan files at different time nodes and generate the 3D point cloud model of the corresponding time node;

[0016] Based on the generated 3D point cloud model, the pavement geometry information including the shape, elevation, curvature and cross-section information of the road surface and the microscopic surface structure information is extracted;

[0017] Use the point cloud M3C2 algorithm to perform deformation analysis on the three-dimensional point cloud model at two or more time nodes, and output the deformation of the field track in all directions by calculating the position change of each point in the point cloud;

[0018] According to the preset neighborhood range or the number of points in the neighborhood, the adjacent point set of each point in the point cloud model is determined, the fitting plane of the adjacent point set is calculated by the least squares method, and the normal vector of the fitting plane is obtained as the normal vector of each point;

[0019] According to the difference between the normal vectors of adjacent points, the curvature of each point is solved, and the flatness information of the track is output based on the obtained curvature data and one or more pre-set flatness thresholds;

[0020] The point cloud analysis results are output based on the road surface geometry information, the deformation of the field road in various directions and the field road flatness information.

[0021] Optionally, the pavement is scanned by a camera to capture the pavement image information of the airport runway, the pavement image information is intelligently identified for microscopic road defects, and the damaged locations on the runway are marked, and then the runway condition index PCI is calculated, including:

[0022] Control high-speed camera equipment to scan the pavement of the airport runway to obtain pavement image information;

[0023] Each image in the road surface image is analyzed through the deep learning network U-net to extract the road surface crack feature image;

[0024] According to the number, density and presentation form of cracks in the pavement crack feature image, the microscopic damage on the road is classified, and after the damage location is identified, it is automatically marked on the pavement image;

[0025] Divide each picture in the obtained road surface image into a plurality of virtual units;

[0026] The damage density D of various diseases in each virtual unit is determined based on the total unit area and the equivalent total damage area of ​​various types of damage. ij ;

[0027] According to the damage density and the damage-reduction curve obtained by fitting the historical data of damage density and reduction value, the reduction value DV of each damage type in each virtual unit is found. i ;

[0028] Sort all the reduction values ​​to form an ordered reduction value array;

[0029] When the reduction value DV in the reduction value array i When the index i=1 and the reduction value is greater than 5, the reduction value DV is taken i As the modified reduction value MaxCDV;

[0030] When the reduction value DV in the reduction value array i When index i>1, the reduction value DV is determined based on the reduction value array i The maximum reduction value DV in imax , find the maximum reduction value DV imax The number of damage types included is used to form a new array to calculate the new total reduction value;

[0031] According to the new total number of reduction values, the number of damage types and the total number of reduction values-corrected value curve obtained by fitting the historical data of the total number of reduction values ​​and the corrected reduction values, the corrected reduction value MaxCDV is found;

[0032] The value of the track condition index PCI is obtained by subtracting the obtained corrected reduction value from the set upper limit value.

[0033] Optionally, the damage density is:

[0034] D ij =A ij / A×100;

[0035] Where A is the total area of ​​the virtual unit, A ij is the equivalent total damaged area of ​​the jth damage level of the i-th damage type;

[0036] The reduction value array is:

[0037] {DV i (i=1-n)};

[0038] The number of damage types are:

[0039] m=1+[0.095(100-DV i max )];

[0040] The new array is:

[0041] {DV i (i=1~m)};

[0042] The new reduction in value is:

[0043]

[0044] The value of the track condition index PCI is:

[0045] PCI=100-MaxCDV.

[0046] Optionally, a multi-source data fusion field expert evaluation model is constructed based on the AHP and TOPSIS algorithms, and the point cloud analysis results, field condition index PCI and acquired field basic data are imported into the multi-source data field expert evaluation model. The expert knowledge graph library obtained through graph neural network training is weighted averaged, and the output field comprehensive evaluation index includes:

[0047] A multi-level structure for the field evaluation system is constructed using AHP, where the first level of the multi-level structure contains various evaluation dimensions, and the second level of the multi-level structure contains the evaluation indicators under each evaluation dimension;

[0048] The TOPSIS algorithm is introduced to configure corresponding evaluation criteria for each evaluation index, and the multi-source data field expert evaluation model is formed by integrating evaluation dimensions, evaluation indicators and evaluation indicators in a multi-level structure;

[0049] Collect and analyze the results of expert evaluations of existing projects and the associated data of safety levels, and use graph neural network technology to build and train an expert knowledge graph library;

[0050] Obtaining basic track data including track location coordinates, track plate type and structure, and service life through at least one pre-deployed monitoring device;

[0051] The point cloud analysis results, the field condition index PCI and the field basic data are imported into the multi-source data field expert evaluation model, and the expert knowledge graph library trained by the graph neural network is used for weighted average processing to output the field comprehensive evaluation index and field classification standard;

[0052] Among them, the multi-source data expert evaluation model includes:

[0053] Z + =(max{z 11 ,z 21 ,…,z n1},max{z 12 ,z 22 ,…,z n2},…,max{z 1m ,z 2m ,…,z nm})

[0054] Z - =(min{z 11 ,z 21 ,…,z n1},min{z 12 ,z 22 ,…,z n2},…,min{z 1m ,z 2m ,…,z nm})

[0055]

[0056]

[0057]

[0058]

[0059] In the formula, m is the dimension of the included evaluation indicators, n is the evaluation data of each evaluation indicator, and Z + With Z - They are the standardized matrices Z ij The positive and negative ideal solutions of , the standardized matrix represents the jth term of the positive and negative ideal solutions, w i is the weight coefficient, which is the weight of each factor in the system obtained by comprehensive analysis of the AHP method, D + With D - are the i-th evaluation object and the positive ideal solution Z + and negative ideal solution Z - The degree of closeness, S iis the rating value, and ζ is the evaluation weight.

[0060] Optionally, the space-time interpolation algorithm and gradient analysis are introduced to analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and the prediction results of the space-time change of the field and track conditions are obtained, including:

[0061] Collect and organize comprehensive evaluation index data of the field in multiple time periods;

[0062] The Kriging spatial interpolation algorithm is used to analyze the spatial distribution of the collected field comprehensive evaluation index data, and the field comprehensive evaluation index of each point on the field is calculated;

[0063] By performing interpolation calculation on the comprehensive evaluation index data of the field and track at the same position in different time periods, a spatiotemporal interpolation result containing multiple groups of interpolation data is obtained. The multiple groups of interpolation data reflect the changes in the field and track conditions in the time dimension.

[0064] Based on multiple sets of interpolation data, the change gradient of the field and track comprehensive evaluation index is calculated. The change gradient reflects the change rate and direction of the field and track conditions in the spatial dimension.

[0065] By analyzing the obtained change gradient, the gradient analysis result including the problem area to be concerned is output;

[0066] Based on the spatiotemporal interpolation results and gradient analysis results, the spatiotemporal change prediction results of the field conditions are output.

[0067] Optionally, the collected field track comprehensive evaluation index data is analyzed for spatial distribution by using the Kriging spatial interpolation algorithm, and the field track comprehensive evaluation index of each point on the field track is calculated, including:

[0068] Selecting estimation points in the field space;

[0069] According to the distance between each virtual unit and the estimated point and the variability of the field comprehensive evaluation index, the weight coefficient of each virtual unit relative to the estimated point is determined;

[0070] Traverse all virtual units, assign corresponding weight coefficients to each virtual unit, and then calculate the comprehensive evaluation index value of the field track at each point;

[0071] Among them, the Kriging interpolation calculation method is as follows:

[0072]

[0073] In the formula, Z(s i ) is the comprehensive evaluation index of the field of the ith virtual unit, λ i Based on the estimated point S 0 The weight coefficient determined by the distance and the variability of the comprehensive evaluation index of the field, S 0is the estimated point, and N is the total number of virtual units involved in the calculation.

[0074] In a second aspect, an embodiment of the present invention provides an airport runway intelligent assessment system based on multi-source data, comprising:

[0075] The point cloud analysis module is configured to scan the pavement of the airport runway by using a laser radar, solve and generate a three-dimensional point cloud model of the pavement, and perform at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result;

[0076] The runway condition index output module is configured to capture the pavement image information of the airport runway by scanning the pavement with a camera, intelligently identify the microscopic road defects on the pavement image information, and mark the damaged locations on the runway, thereby calculating the runway condition index PCI;

[0077] The comprehensive evaluation module is configured to build a multi-source data fusion field expert evaluation model based on the AHP and TOPSIS algorithms, and import the point cloud analysis results, field condition index PCI and acquired field basic data into the multi-source data field expert evaluation model, perform weighted average processing on the expert knowledge graph library obtained through graph neural network training, and output the field comprehensive evaluation index;

[0078] The spatiotemporal change prediction module is configured to introduce spatiotemporal interpolation algorithms and gradient analysis, analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and obtain the spatiotemporal change prediction results of the field and track conditions.

[0079] In a third aspect, an embodiment of the present invention provides an airport runway intelligent assessment device based on multi-source data, comprising: at least one database; and a memory communicatively connected to the at least one database; wherein the memory stores instructions executable by the at least one database, and the instructions are executed by the at least one database so that the at least one database can execute the airport runway intelligent assessment method based on multi-source data as described above.

[0080] In a fourth aspect, an embodiment of the present invention provides a computer-readable medium having computer-executable instructions stored thereon, which, when executed by a processor, implement the airport runway intelligent assessment method based on multi-source data as described above.

[0081] (III) Beneficial effects

[0082] The beneficial effects of the present invention are:

[0083] First, through the laser radar scanning technology, the three-dimensional point cloud model of the road surface can be accurately calculated and generated. This not only provides a highly accurate data basis for field and road evaluation, but also further reveals the detailed features of the road surface through point cloud analysis.

[0084] Secondly, the road surface images are captured by camera equipment, and the microscopic defects and damage locations on the road are marked through intelligent recognition technology, and then the road condition index PCI is accurately calculated. This method not only improves the efficiency and accuracy of road defect identification, but also provides a clear target location for subsequent maintenance and repair work.

[0085] Furthermore, the multi-source data fusion runway expert evaluation model built based on the AHP and TOPSIS algorithms integrates information from multiple data sources, including point cloud analysis results, runway condition index PCI, and runway basic data. The expert knowledge graph library obtained through graph neural network training is weighted averaged to output a comprehensive and objective runway comprehensive evaluation index. This not only solves the problems of the singleness and staticness of traditional evaluation methods, but also better meets the needs of modern airports for real-time monitoring and intelligent analysis.

[0086] Finally, by introducing the spatiotemporal interpolation algorithm and gradient analysis, the present invention can make in-depth spatiotemporal change predictions for the comprehensive evaluation index of the runway, which can not only reveal the distribution pattern and change law of the runway conditions, but also provide powerful decision-making support for airport managers, enabling them to carry out targeted optimization and maintenance work based on the prediction results.

[0087] Therefore, the present invention comprehensively uses a variety of advanced technologies and algorithms to achieve a comprehensive, dynamic and efficient evaluation of airport runways, effectively solving the problems of singleness, staticness and inefficiency of traditional airport runway evaluation methods, and providing strong technical support for real-time monitoring and intelligent analysis of modern airports. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 A schematic diagram of a flow chart of a method provided by an embodiment of the present invention;

[0089] Figure 2 A schematic diagram of a specific flow chart of step S1 of the method provided in an embodiment of the present invention;

[0090] Figure 3 A schematic diagram of a specific flow chart of step S2 of the method provided in an embodiment of the present invention;

[0091] Figure 4 A schematic diagram of a damage-reduction curve of the method provided in an embodiment of the present invention;

[0092] Figure 5 A schematic diagram of a curve of the total reduction value and the corrected value of the method provided in an embodiment of the present invention;

[0093] Figure 6 A specific schematic diagram of step S3 of the method provided in an embodiment of the present invention;

[0094] Figure 7 A schematic diagram of the architecture of an expert evaluation system for the method provided by an embodiment of the present invention;

[0095] Figure 8 A specific schematic diagram of step S4 of the method provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0096] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0097] like Figure 1 As shown, an intelligent evaluation method for airport runways based on multi-source data proposed in an embodiment of the present invention includes: scanning the pavement of an airport runway by a laser radar, solving and generating a three-dimensional point cloud model of the pavement, and performing at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result; scanning the pavement by a camera to capture the pavement image information of the airport runway, intelligently identifying microscopic road defects on the pavement image information, and marking the damaged locations on the runway, thereby solving the runway condition index PCI; constructing a runway expert evaluation model based on multi-source data fusion based on the AHP and TOPSIS algorithms, and importing the point cloud analysis results, the runway condition index PCI and the acquired runway basic data into the multi-source data runway expert evaluation model, performing weighted average processing on the expert knowledge graph library obtained through graph neural network training, and outputting a runway comprehensive evaluation index; introducing a spatiotemporal interpolation algorithm and gradient analysis to analyze the distribution pattern and change law of the runway comprehensive evaluation index to obtain a spatiotemporal change prediction result of the runway condition.

[0098] First, through the laser radar scanning technology, the three-dimensional point cloud model of the road surface can be accurately calculated and generated. This not only provides a highly accurate data basis for field and road evaluation, but also further reveals the detailed features of the road surface through point cloud analysis.

[0099] Secondly, the road surface images are captured by camera equipment, and the microscopic defects and damage locations on the road are marked through intelligent recognition technology, and then the road condition index PCI is accurately calculated. This method not only improves the efficiency and accuracy of road defect identification, but also provides a clear target location for subsequent maintenance and repair work.

[0100] Furthermore, the multi-source data fusion runway expert evaluation model built based on the AHP and TOPSIS algorithms integrates information from multiple data sources, including point cloud analysis results, runway condition index PCI, and runway basic data. The expert knowledge graph library obtained through graph neural network training is processed by weighted average to output a comprehensive and objective runway comprehensive evaluation index. This not only solves the problems of the singleness and staticness of traditional evaluation methods, but also can better meet the needs of modern airports for real-time monitoring and intelligent analysis.

[0101] Finally, by introducing the spatiotemporal interpolation algorithm and gradient analysis, the present invention can make in-depth spatiotemporal change predictions for the comprehensive evaluation index of the runway, which can not only reveal the distribution pattern and change law of the runway conditions, but also provide powerful decision-making support for airport managers, enabling them to carry out targeted optimization and maintenance work based on the prediction results.

[0102] Therefore, the present invention comprehensively uses a variety of advanced technologies and algorithms to achieve a comprehensive, dynamic and efficient evaluation of airport runways, effectively solving the problems of singleness, staticness and inefficiency of traditional airport runway evaluation methods, and providing strong technical support for real-time monitoring and intelligent analysis of modern airports.

[0103] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0104] Specifically, an embodiment of the present invention provides an airport runway intelligent evaluation method based on multi-source data, including:

[0105] S1. Scan the pavement of the airport runway by using a laser radar, solve and generate a three-dimensional point cloud model of the pavement, and perform at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result.

[0106] Furthermore, if Figure 2 As shown, step S1 includes:

[0107] S11. Control the precision laser radar to scan the pavement of the airport runway, and receive and store the scanned files wirelessly transmitted by the laser radar. The laser radar used is a high-precision laser radar with sub-millimeter accuracy and a wireless data transmission function, which can transmit the scanned files to the data transmission module. The data transmission module is connected to the central server and can transmit data back in real time without interruption.

[0108] S12. Utilize the pre-deployed solution function module and solution software to solve the received scan files at different time nodes, and generate a three-dimensional point cloud model of the corresponding time node.

[0109] S13. Based on the generated 3D point cloud model, extract the pavement geometry information including the shape, elevation, curvature, cross-section information and microscopic surface structure information of the road surface. The central server should have a fast solution function, and can quickly generate a 3D high-precision point cloud model of the time node through mainstream solution software. The pavement geometry information covers the characteristics of the road surface shape, elevation, curvature, cross-section information and microscopic surface structure information.

[0110] S14. Use the point cloud M3C2 algorithm to perform deformation analysis on the three-dimensional point cloud model of two or more time nodes, and output the deformation of the field track in each direction by calculating the position change of each point in the point cloud.

[0111] In a specific embodiment, it is first necessary to obtain a three-dimensional point cloud model of the airport runway at two or more different time nodes. Secondly, the point cloud M3C2 algorithm is used for deformation analysis. The algorithm is designed to detect complex changes directly on the point cloud data without meshing, thereby simplifying the processing flow. When the M3C2 algorithm performs change calculations, it is less affected by spatial point density, surface roughness, and different sampling positions, which ensures the accuracy and reliability of deformation analysis.

[0112] Next, the M3C2 algorithm is used to perform comparative calculations on the selected two or more 3D point cloud models. This process involves accurate matching and comparison of the positions of the points in each point cloud. Through the algorithm calculation, the position changes of each point in the point cloud can be identified, and these changes reflect the actual deformation of the track in different time periods. Furthermore, based on the calculation results of the M3C2 algorithm, the deformation of the track in various directions is output.

[0113] S15. Determine the adjacent point set of each point in the point cloud model according to the preset neighborhood range or the number of points in the neighborhood, calculate the fitting plane of the adjacent point set by the least squares method, and obtain the normal vector of the fitting plane as the normal vector of each point.

[0114] S16. According to the difference between the normal vectors of adjacent points, the curvature of each point is solved, and based on the solved curvature data and one or more pre-set smoothing thresholds, the smoothing information of the track is output.

[0115] S17. Output the point cloud analysis result according to the road surface geometry information, the deformation of the field road in various directions and the flatness information of the field road.

[0116] In another specific embodiment, first, a preset neighborhood range or the number of points in the neighborhood is set for each point in the point cloud model. This helps to determine which points are adjacent to each other. Alternatively, based on the preset neighborhood conditions, the neighboring points of each point are automatically found and determined by the algorithm. Secondly, for each point and its adjacent points, a plane is fitted using the least squares method. This fitting plane represents the local geometric features of the point and its neighborhood. A normal vector is calculated from the fitting plane, and this normal vector will be regarded as the normal vector of the point.

[0117] Next, compare the normal vectors of adjacent points and calculate the difference between them. This difference reflects the curvature of the local surface. Based on the difference in normal vectors, the curvature of each point is solved by an algorithm. The larger the curvature, the higher the curvature of the local surface at that point. Compare the solved curvature data with one or more pre-set flatness thresholds. If the curvature of a point exceeds the set threshold, the flatness of that location is judged to be poor.

[0118] Finally, according to the curvature data and the flatness judgment results, the flatness information of the track is output, which can include the location of points with poor flatness, curvature values, etc.

[0119] S2. Use camera equipment to scan the pavement and capture the pavement image information of the airport runway, intelligently identify the microscopic road defects on the pavement image information, mark the damaged locations on the runway, and then calculate the runway condition index PCI.

[0120] Furthermore, if Figure 3 As shown, step S2 includes:

[0121] S21, control the high-speed camera equipment to scan the pavement of the airport runway to obtain pavement image information. The camera used is a high-resolution high-speed camera equipment, which can quickly save the image data during the movement process, keep the size of the saved pictures consistent with the parameters such as the camera focal length, and mark the coordinates of each photo position.

[0122] S22. Each image in the road surface image is analyzed through the deep learning network U-net to extract the road surface crack feature image.

[0123] S23. According to the number, density and presentation form of cracks in the pavement crack feature image, the microscopic damage on the road is classified, and after the damage location is identified, it is automatically marked on the pavement image. Concrete pavements can be divided into: single cracks, corner fractures, cross cracks, settlement, expansion cracks, filler / joint damage, potholes, plate corner peeling, patches, etc.; asphalt pavements can be divided into cracks, cracks, slippage, aging, bulges, collapses, pushes, patches, etc.

[0124] S24, dividing each image in the obtained road surface image into a plurality of virtual units, and determining the damage density of various diseases in each virtual unit based on the total area of ​​the unit and the equivalent total damage area of ​​various types of damage.

[0125] S25. Find the reduction value DV of each damage type in each virtual unit based on the damage density and the damage-reduction curve obtained by fitting the historical data of damage density and reduction value. i .

[0126] S26. Sort all the reduction values ​​to form an ordered reduction value array.

[0127] S27, when the reduction value DV in the reduction value array i When the index i=1 and the reduction value is greater than 5, the reduction value DV is taken i As the modified reduction value MaxCDV. When the reduction value DV in the reduction value array i When index i>1, the reduction value DV is determined based on the reduction value array i The maximum reduction value DV in imax , find the maximum reduction value DV i max The number of damage types included is used to form a new array to calculate the new total reduction value.

[0128] S28. Find and obtain the corrected reduction value MaxCDV based on the new total reduction value, the number of damage types, and the total reduction value-corrected value curve obtained by fitting the historical data of the total reduction value and the corrected reduction value.

[0129] S29. Obtain the value of the field track condition index PCI by subtracting the obtained corrected reduction value from the set upper limit value.

[0130] In a specific embodiment, the damage density of the pavement damage reduction value is first calculated as:

[0131] D ij =A ij / A×100;

[0132] Where A is the total area of ​​the virtual unit, A ij is the equivalent total damaged area of ​​the jth damage degree of the i-th damage type.

[0133] Secondly, the damage density is determined Figure 4 The damage-reduction curve shown in the figure determines the reduction value DV for each unit under each damage type. i .like Figure 4 As shown in the figure, according to the degree of damage, there are three situations: high (H), medium (M), and low (L).

[0134] After that, after calculating the deduction value of each item, arrange the deduction values ​​from large to small to form a reduction value array:

[0135] {DV i (i=1-n)};

[0136] Next, we will deal with two situations separately:

[0137] (1) When i = 1DV i >5 when MaxCDV=DV i

[0138] (2) When i>1, the calculation of MaxCDV includes the number of damage types m.

[0139] m=1+[0.095(100-DV i max )];

[0140] Among them, DV imax For DV i The maximum value in , and form a new array as:

[0141] {DV i (i=1~m)};

[0142] And calculate the new total reduction value as:

[0143]

[0144] according to Figure 5 The total reduction value-corrected value curve shown is corrected according to different numbers of m to obtain the corrected reduction value MaxCDV.

[0145] Finally, the value of the track condition index PCI is calculated according to the following formula:

[0146] PCI=100-MaxCDV.

[0147] S3. Based on the AHP and TOPSIS algorithms, a field and track expert evaluation model with multi-source data fusion is constructed, and the point cloud analysis results, field and track condition index PCI and the acquired field and track basic data are imported into the multi-source data field and track expert evaluation model. The expert knowledge graph library obtained through graph neural network training is weighted averaged to output the field and track comprehensive evaluation index.

[0148] Furthermore, if Figure 6 As shown, step S3 includes:

[0149] S31. Use AHP to construct a multi-level structure for the field evaluation system, wherein the first level of the multi-level structure contains various evaluation dimensions, and the second level of the multi-level structure contains evaluation indicators under each evaluation dimension.

[0150] A multi-level structure was constructed using the analytic hierarchy process (AHP), which was specifically used for the field evaluation system. In this multi-level structure, two core levels were established to comprehensively reflect the status of the field.

[0151] Level 1: Evaluation dimensions, including: geometric information, covering the basic geometric characteristics of the field, damage degree, reflecting the damage to the field, and basic status, describing the overall condition of the field. These evaluation dimensions constitute the basic framework for evaluating the condition of the field.

[0152] The second level: evaluation indicators. Under each evaluation dimension, specific evaluation indicators are further refined to ensure the comprehensiveness and accuracy of the evaluation. Such as the cross-sectional shape of the track, track settlement and deformation, track flatness, track damage density, track PCI index, track materials, track maintenance, track drainage, and track operation. Specifically, the cross-sectional shape of the track, track settlement and deformation, and track flatness belong to the geometric information dimension, which describes the physical form of the track in detail. The track damage density and track PCI index are: Under the dimension of degree of damage, these indicators quantify the damage to the track. And, track materials, track maintenance, track drainage, and track operation: These indicators are classified into the basic status dimension, providing detailed information on track operation and maintenance.

[0153] S32. Introduce the TOPSIS algorithm to configure corresponding evaluation standards for each evaluation indicator, and form a multi-source data field expert evaluation model by integrating evaluation dimensions, evaluation indicators and evaluation indicators in a multi-level structure.

[0154] In order to improve the evaluation system, the TOPSIS (Topic Selection-based Multi-attribute Decision Making) algorithm was introduced. This algorithm configures corresponding evaluation criteria for each evaluation indicator, and can integrate evaluation dimensions, evaluation indicators and corresponding evaluation criteria in a multi-level structure, thus forming a comprehensive and comprehensive multi-source data field expert evaluation model.

[0155] S33. Collect and analyze the results of expert evaluations of existing projects and the associated data on safety levels, and use graph neural network technology to build and train an expert knowledge graph library.

[0156] By collecting and analyzing the correlation data between the expert evaluation results and safety levels in past projects, an expert knowledge graph library was constructed using graph neural network (GNN) technology. This graph library not only maps the deep connection between expert evaluation and safety level, but also improves its accuracy and reliability in field evaluation through continuous training and optimization.

[0157] S34. Obtaining basic track data including track location coordinates, track plate type and structure, and service life through at least one pre-deployed monitoring device.

[0158] S35. Import the point cloud analysis results, the field and track condition index PCI and the field and track basic data into the multi-source data field and track expert evaluation model, and use the expert knowledge graph library trained by the graph neural network to perform weighted average processing to output the field and track comprehensive evaluation index and field and track grading standards.

[0159] Specifically, we first establish a multi-source evaluation index Xnm, where m is the included evaluation index dimension and n is the evaluation data of each evaluation index. Then, we standardize the multi-source data matrix and establish a standardized matrix Z. To calculate the positive ideal solution and negative ideal solution Z + With Z - , define the closeness D between the i-th evaluation object and the positive ideal solution and the negative ideal solution + With D - Finally, the score value S is obtained. After S is normalized, the evaluation weight ζ is obtained as the fusion coefficient, and then the multi-source data expert evaluation model is obtained:

[0160] Z + =(max{z 11 ,z 21 ,…,z n1},max{z 12 ,z 22 ,…,z n2},…,max{z 1m ,z 2m ,…,z nm})

[0161] Z - =(min{z 11 ,z 21 ,…,z n1},min{z 12 ,z 22 ,…,z n2},…,min{z 1m ,z 2m ,…,z nm})

[0162]

[0163]

[0164]

[0165]

[0166] During the evaluation process, the basic data of the track is first obtained through pre-deployed monitoring equipment, including key information such as location coordinates, type and structure of track slabs, and service life. Then, the point cloud analysis results, track condition index (PCI) and these basic data are imported into the multi-source data track expert evaluation model built previously.

[0167] Using the expert knowledge graph library trained by the graph neural network, the correlation weights between the various indicators in the expert evaluation system and the degree of influence between the indicators and the grade division are determined. Then, the weighted average method is used to assign weights to data from different sources according to the importance of each indicator, and then the weighted average is performed, and finally a comprehensive evaluation index of the field and the corresponding field classification standard are output. This comprehensive evaluation index is a quantitative representation of the overall condition of the field and can intuitively reflect the safety level and performance of the field. Reference Figure 7 The grading standard divides the tracks into four levels: very safe, basically safe, general, and high-risk according to the comprehensive evaluation index, providing clear guidance for subsequent maintenance and management.

[0168] S4. Introduce spatiotemporal interpolation algorithm and gradient analysis to analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and obtain the spatiotemporal change prediction results of the field and track conditions.

[0169] Furthermore, if Figure 8 As shown, step S4 includes:

[0170] S41. Collect and organize the comprehensive evaluation index data of the field and track for multiple time periods. In this step, it is necessary to collect the comprehensive evaluation index data of the field and track from different time nodes. These data should cover multiple time periods to ensure that the subsequent analysis can capture the changing trend of the field and track conditions over time.

[0171] S42. Perform spatial distribution analysis on the collected field and track comprehensive evaluation index data through Kriging spatial interpolation algorithm, and calculate the field and track comprehensive evaluation index of each point on the field and track.

[0172] Furthermore, S42 includes: selecting a number of estimated points in the field space as target positions for analysis. Then, according to the distance between each virtual unit and the estimated point and the variability of the field comprehensive evaluation index, the weight coefficient of each virtual unit relative to the estimated point is determined. Furthermore, all virtual units are traversed, and a corresponding weight coefficient is assigned to each virtual unit, thereby calculating the field comprehensive evaluation index value of each point.

[0173] In a specific embodiment, first, the distance between each virtual unit (i.e., known observation point) and the estimated point needs to be calculated. Then, based on the variability of the field comprehensive evaluation index (usually described by the semivariance function), the influence of each virtual unit on the estimated point, i.e., the weight coefficient, is determined.

[0174] The estimated value of the field comprehensive evaluation index of the estimated point is obtained by weighted summation of the weight coefficients of all virtual units. This process needs to be repeated for all estimated points in the field space to calculate the comprehensive evaluation index of each point on the field.

[0175] Among them, the Kriging interpolation calculation method is as follows:

[0176]

[0177] In the formula, Z(s i ) is the comprehensive evaluation index of the field of the ith virtual unit, λ i Based on the estimated point S 0 The weight coefficient determined by the distance and the variability of the comprehensive evaluation index of the field, S 0 is the estimated point, and N is the total number of virtual units involved in the calculation.

[0178] S43. By performing change interpolation calculation on the field and track comprehensive evaluation index data at the same position in different time periods, a spatiotemporal interpolation result including multiple groups of interpolation data is obtained. The multiple groups of interpolation data reflect the change of the field and track conditions in the time dimension.

[0179] S44. Based on multiple sets of interpolation data, the change gradient of the field and track comprehensive evaluation index is calculated. The change gradient reflects the change rate and direction of the field and track conditions in the spatial dimension.

[0180] S45, by analyzing the obtained change gradient, output the gradient analysis result including the problem area to be concerned. This analysis result can help the management personnel quickly locate the area in the field where there may be problems or areas that need to be paid attention to, such as determining the location with a larger gradient.

[0181] S46. Based on the spatiotemporal interpolation results and gradient analysis results, the spatiotemporal change prediction results of the field and track conditions are output. This prediction result not only takes into account the field and track conditions at the current time point, but also predicts its possible change trend in the future, thereby providing powerful decision-making support for the management and maintenance of the field and track.

[0182] In addition, an embodiment of the present invention provides an airport runway intelligent assessment system based on multi-source data, comprising:

[0183] The point cloud analysis module is configured to scan the pavement of the airport runway through a laser radar, solve and generate a three-dimensional point cloud model of the pavement, and perform at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result.

[0184] The runway condition index output module is configured to use a camera to scan the runway to capture the pavement image information of the airport runway, intelligently identify microscopic road defects in the pavement image information, and mark the damaged locations on the runway, thereby calculating the runway condition index PCI.

[0185] The comprehensive evaluation module is configured to build a field and track expert evaluation model based on multi-source data fusion based on the AHP and TOPSIS algorithms, and import the point cloud analysis results, field and track condition index PCI and the acquired field and track basic data into the multi-source data field and track expert evaluation model, and perform weighted averaging processing on the expert knowledge graph library obtained through graph neural network training to output the field and track comprehensive evaluation index.

[0186] The spatiotemporal change prediction module is configured to introduce spatiotemporal interpolation algorithms and gradient analysis, analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and obtain the spatiotemporal change prediction results of the field and track conditions.

[0187] Furthermore, an embodiment of the present invention provides an airport runway intelligent assessment device based on multi-source data, comprising: at least one database; and a memory communicatively connected to the at least one database; wherein the memory stores instructions executable by the at least one database, and the instructions are executed by the at least one database so that the at least one database can execute the airport runway intelligent assessment method based on multi-source data as described above.

[0188] Specifically, this device contains at least one database, which is well designed and can efficiently store and manage massive amounts of multi-source data, including but not limited to various monitoring data of airport runways, historical maintenance records, and real-time operation information.

[0189] In addition, the device is also equipped with a memory that is connected to the at least one database. The role of this memory is very critical. It not only has a large capacity and can store a large amount of data and instructions, but more importantly, it stores instructions that can be executed by the at least one database. These instructions are carefully designed and optimized, and they can be executed by the database, thereby driving the entire device to complete complex data processing and analysis tasks.

[0190] It is worth mentioning that when these instructions are executed by the database, they enable the database to execute the above-mentioned intelligent airport runway assessment method based on multi-source data. This means that through this device, users can easily achieve a comprehensive, accurate, and intelligent assessment of airport runways, timely discover potential safety hazards, and provide strong guarantees for the safe operation of the airport. In general, the device provided by the embodiment of the present invention is not only technologically advanced, but also highly practical and has a high market promotion value.

[0191] Furthermore, an embodiment of the present invention provides a computer-readable medium having computer-executable instructions stored thereon, and when the executable instructions are executed by a processor, the airport runway intelligent assessment method based on multi-source data as described above is implemented.

[0192] The embodiment of the present invention also provides a special computer-readable medium. This medium stores carefully designed computer-executable instructions. When these instructions are executed by the processor, they will accurately implement the airport runway intelligent evaluation method based on multi-source data described above. This means that through this computer-readable medium, users can easily obtain and implement this advanced evaluation method, thereby improving the intelligent level of airport runway management and ensuring the safety and efficiency of the runway. This medium is not only easy to carry and transmit, but also ensures the consistency and accuracy of the evaluation method, providing strong technical support for airport operations.

[0193] In summary, the embodiments of the present invention provide an intelligent evaluation method, system, device and method for airport runways based on multi-source data. The present invention uses a laser radar to scan the runway surface and an intelligent visual image recognition algorithm to deeply evaluate the real-time status of the airport runway from multiple dimensions. The key data such as the geometric information, settlement deformation, and degree of damage of the runway surface are calculated, and a comprehensive runway expert evaluation system with multi-source data is constructed by combining the AHP and TOPSIS methods. In addition, the Kriging interpolation algorithm can be used to predict the changing trend of the runway comprehensive evaluation index and accurately locate key positions.

[0194] The core advantage of this invention is that it can use intelligent algorithms to conduct real-time and comprehensive evaluation of airport runways, ensuring the high accuracy and practical value of the evaluation results. By integrating multi-source data, including information of various types and sources, a more comprehensive and global runway evaluation strategy is provided. This strategy not only helps to comprehensively consider various influencing factors, thereby gaining a deeper understanding of the runway status, but also can make up for the limitations of single data source evaluation and more accurately reflect the true status of the runway.

[0195] In addition, the expert system intelligent analysis method adopted by the present invention has strong adaptability and robustness, enabling it to maintain excellent performance in the complex and changing airport environment. Multi-source data monitoring technology provides a broader perspective for evaluation and richer information support for decision-making by introducing data from different spatial and temporal dimensions.

[0196] Ultimately, these innovative technologies work together to enable airport managers and decision makers to gain a deeper understanding of runway performance, obtain more comprehensive runway status reports, and make more targeted optimization decisions based on them. At the same time, the automated and intelligent evaluation process significantly improves work efficiency and reduces labor costs, thereby greatly improving overall economic benefits.

[0197] Since the system / device described in the above embodiments of the present invention is a system / device used to implement the method of the above embodiments of the present invention, a person skilled in the art can understand the specific structure and deformation of the system / device based on the method described in the above embodiments of the present invention, and thus will not be described in detail here. All systems / devices used in the method of the above embodiments of the present invention belong to the scope of protection of the present invention.

[0198] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0199] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0200] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0201] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments after knowing the basic creative concept. Therefore, the claims should be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0202] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include these modifications and variations.

Claims

1. An airport runway intelligent assessment method based on multi-source data, characterized in that: include: Scanning the pavement of the airport runway by laser radar, solving and generating a three-dimensional point cloud model of the pavement, and performing at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result; Use camera equipment to scan the pavement and capture the pavement image information of the airport runway, intelligently identify the microscopic road defects on the pavement image information, mark the damaged locations on the runway, and then calculate the runway condition index PCI; Based on the AHP and TOPSIS algorithms, a multi-source data fusion field and track expert evaluation model is constructed. The point cloud analysis results, field and track condition index PCI and acquired field and track basic data are imported into the multi-source data field and track expert evaluation model. The expert knowledge graph library obtained through graph neural network training is weighted averaged to output the field and track comprehensive evaluation index. The spatiotemporal interpolation algorithm and gradient analysis are introduced to analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and the spatiotemporal change prediction results of the field and track conditions are obtained.

2. The airport runway intelligent assessment method based on multi-source data according to claim 1, characterized in that: Scan the pavement of the airport runway by laser radar, solve and generate a 3D point cloud model of the pavement, and perform at least one point cloud analysis based on the 3D point cloud model of the pavement. The point cloud analysis results include: Control the precision laser radar to scan the pavement of the airport runway, and receive and store the scan files wirelessly transmitted by the laser radar; Utilize the pre-deployed solution function module and solution software to solve the received scan files at different time nodes and generate the 3D point cloud model of the corresponding time node; Based on the generated 3D point cloud model, the pavement geometry information including the shape, elevation, curvature and cross-section information of the road surface and the microscopic surface structure information is extracted; Use the point cloud M3C2 algorithm to perform deformation analysis on the three-dimensional point cloud model at two or more time nodes, and output the deformation of the field track in all directions by calculating the position change of each point in the point cloud; According to the preset neighborhood range or the number of points in the neighborhood, the adjacent point set of each point in the point cloud model is determined, the fitting plane of the adjacent point set is calculated by the least squares method, and the normal vector of the fitting plane is obtained as the normal vector of each point; According to the difference between the normal vectors of adjacent points, the curvature of each point is solved, and the flatness information of the track is output based on the obtained curvature data and one or more pre-set flatness thresholds; The point cloud analysis results are output based on the road surface geometry information, the deformation of the field road in various directions and the field road flatness information.

3. The airport runway intelligent assessment method based on multi-source data according to claim 1, characterized in that: The pavement image information of the airport runway is captured by scanning the pavement with camera equipment, and the microscopic road defects are intelligently identified on the pavement image information, and the damaged locations on the runway are marked, and then the runway condition index PCI is calculated, including: Control high-speed camera equipment to scan the pavement of the airport runway to obtain pavement image information; Each image in the road surface image is analyzed through the deep learning network U-net to extract the road surface crack feature image; According to the number, density and presentation form of cracks in the pavement crack feature image, the microscopic damage on the road is classified, and after the damage location is identified, it is automatically marked on the pavement image; Divide each picture in the obtained road surface image into a plurality of virtual units; The damage density D of various diseases in each virtual unit is determined based on the total unit area and the equivalent total damage area of ​​various types of damage. ij ; According to the damage density and the damage-reduction curve obtained by fitting the historical data of damage density and reduction value, the reduction value DV of each damage type in each virtual unit is found. i ; Sort all the reduction values ​​to form an ordered reduction value array; When the reduction value DV in the reduction value array i When the index i=1 and the reduction value is greater than 5, the reduction value DV is taken i As the modified reduction value MaxCDV; When the reduction value DV in the reduction value array i When index i>1, the reduction value DV is determined based on the reduction value array i The maximum reduction value DV in imax , find the maximum reduction value DV imax The number of damage types included is used to form a new array to calculate the new total reduction value; According to the new total number of reduction values, the number of damage types and the total number of reduction values-corrected value curve obtained by fitting the historical data of the total number of reduction values ​​and the corrected reduction values, the corrected reduction value MaxCDV is found; The value of the track condition index PCI is obtained by subtracting the obtained corrected reduction value from the set upper limit value.

4. The airport runway intelligent assessment method based on multi-source data as claimed in claim 3 is characterized in that: The damage density is: D ij =A ij / A×100; Where A is the total area of ​​the virtual unit, A ij is the equivalent total damaged area of ​​the jth damage level of the i-th damage type; The reduction value array is: {DV i (i=1~n)}; The number of damage types are: m=1+[0.095(100-DV imax )]; The new array is: {DV i (i=1~m)}; The new reduction in value is: The value of the track condition index PCI is: PCI=100-MaxCDV.

5. The airport runway intelligent assessment method based on multi-source data according to claim 1, characterized in that: Based on the AHP and TOPSIS algorithms, a multi-source data fusion field expert evaluation model is constructed. The point cloud analysis results, field condition index PCI and acquired field basic data are imported into the multi-source data field expert evaluation model. The expert knowledge graph library obtained through graph neural network training is weighted averaged, and the output field comprehensive evaluation index includes: A multi-level structure for the field evaluation system is constructed using AHP, where the first level of the multi-level structure contains various evaluation dimensions, and the second level of the multi-level structure contains the evaluation indicators under each evaluation dimension; The TOPSIS algorithm is introduced to configure corresponding evaluation criteria for each evaluation index, and the multi-source data field expert evaluation model is formed by integrating evaluation dimensions, evaluation indicators and evaluation indicators in a multi-level structure; Collect and analyze the results of expert evaluations of existing projects and the associated data of safety levels, and use graph neural network technology to build and train an expert knowledge graph library; Obtaining basic track data including track location coordinates, track plate type and structure, and service life through at least one pre-deployed monitoring device; The point cloud analysis results, the field condition index PCI and the field basic data are imported into the multi-source data field expert evaluation model, and the expert knowledge graph library trained by the graph neural network is used for weighted average processing to output the field comprehensive evaluation index and field classification standard; Among them, the multi-source data expert evaluation model includes: WITH + =(max{z 11 ,With 21 ,…,With n1 },max{from 12 ,With 22 ,…,With n2 },…,max{from 1m ,With 2m ,…,With nm }) WITH - =(min{z 11 ,With 21 ,…,With n1 },min{z 12 ,With 22 ,…,With n2 },…,min{z 1m ,With 2m ,…,With nm }) In the formula, m is the dimension of the included evaluation indicators, n is the evaluation data of each evaluation indicator, and Z + With Z - They are the standardized matrices Z ij The positive and negative ideal solutions of , the standardized matrix represents the jth term of the positive and negative ideal solutions, w i is the weight coefficient, D + With D - are the i-th evaluation object and the positive ideal solution Z + and negative ideal solution Z - The degree of closeness, S i is the rating value, and ζ is the evaluation weight.

6. The airport runway intelligent assessment method based on multi-source data according to claim 1, characterized in that: By introducing the spatiotemporal interpolation algorithm and gradient analysis, the distribution pattern and change law of the comprehensive evaluation index of the field and track are analyzed, and the spatiotemporal change prediction results of the field and track conditions are obtained, including: Collect and organize comprehensive evaluation index data of the field in multiple time periods; Through the Kriging spatial interpolation algorithm, the spatial distribution of the collected field comprehensive evaluation index data is analyzed to calculate the field comprehensive evaluation index of each point on the field. By performing interpolation calculation on the comprehensive evaluation index data of the field and track at the same position in different time periods, a spatiotemporal interpolation result containing multiple groups of interpolation data is obtained. The multiple groups of interpolation data reflect the changes in the field and track conditions in the time dimension. Based on multiple sets of interpolation data, the change gradient of the field and track comprehensive evaluation index is calculated. The change gradient reflects the change rate and direction of the field and track conditions in the spatial dimension. By analyzing the obtained change gradient, the gradient analysis result including the problem area to be concerned is output; Based on the spatiotemporal interpolation results and gradient analysis results, the spatiotemporal change prediction results of the field conditions are output.

7. The airport runway intelligent assessment method based on multi-source data according to claim 6 is characterized in that: Through the Kriging spatial interpolation algorithm, the spatial distribution analysis of the collected field comprehensive evaluation index data is carried out, and the field comprehensive evaluation index of each point on the field is calculated, including: Selecting estimation points in the field space; According to the distance between each virtual unit and the estimated point and the variability of the field comprehensive evaluation index, the weight coefficient of each virtual unit relative to the estimated point is determined; Traverse all virtual units, assign corresponding weight coefficients to each virtual unit, and then calculate the comprehensive evaluation index value of the field track at each point; Among them, the Kriging interpolation calculation method is as follows: In the formula, Z(s i ) is the comprehensive evaluation index of the field of the ith virtual unit, λ i is the weight coefficient determined according to the distance of the estimation point S0 and the variability of the comprehensive evaluation index of the field, S0 is the estimation point, and N is the total number of virtual units involved in the calculation.

8. An airport runway intelligent assessment system based on multi-source data, characterized in that: include: The point cloud analysis module is configured to scan the pavement of the airport runway by using a laser radar, solve and generate a three-dimensional point cloud model of the pavement, and perform at least one point cloud analysis based on the three-dimensional point cloud model of the pavement to obtain a point cloud analysis result; The runway condition index output module is configured to capture the pavement image information of the airport runway by scanning the pavement with a camera, intelligently identify the microscopic road defects on the pavement image information, and mark the damaged locations on the runway, thereby calculating the runway condition index PCI; The comprehensive evaluation module is configured to build a multi-source data fusion field expert evaluation model based on the AHP and TOPSIS algorithms, and import the point cloud analysis results, field condition index PCI and acquired field basic data into the multi-source data field expert evaluation model, perform weighted average processing on the expert knowledge graph library obtained through graph neural network training, and output the field comprehensive evaluation index; The spatiotemporal change prediction module is configured to introduce spatiotemporal interpolation algorithms and gradient analysis, analyze the distribution pattern and change law of the field and track comprehensive evaluation index, and obtain the spatiotemporal change prediction results of the field and track conditions.

9. An intelligent airport runway assessment device based on multi-source data, characterized in that: include: at least one database; And a memory communicatively connected to the at least one database; wherein the memory stores instructions executable by the at least one database, and the instructions are executed by the at least one database so that the at least one database can execute the airport runway intelligent assessment method based on multi-source data as described in any one of claims 1-7.

10. A computer-readable medium having computer-executable instructions stored thereon, characterized in that: When the executable instructions are executed by the processor, the airport runway intelligent assessment method based on multi-source data as described in any one of claims 1 to 7 is implemented.

Images