Engineering project data management system and method based on data analysis

Abstract

The invention discloses an engineering project data management system and method based on data analysis, and belongs to the technical field of data analysis. The method comprises the following steps: in a highway engineering project, obtaining engineering project data, and carrying out data preprocessing; extracting route design data and topographic and geomorphic data of the expressway, and performing spatial matching on the route design data and the topographic and geomorphic data; carrying out adaptability analysis on terrain types and line trends, marking abnormal road sections, and forming an abnormal road section data set; extracting abnormal road section data features from the abnormal road section data set, and obtaining an association mode of the abnormal road section data features; clustering the association modes into different categories according to similar features to form an abnormal road section problem combination; fusing the sensor data and the abnormal road section problem combination, and outputting a plurality of candidate route design adjustment schemes through an artificial intelligence algorithm; different candidate route design adjustment schemes are simulated through a geographic information system, and the optimal route design adjustment scheme is selected.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an engineering project data management system and method based on data analysis. Background Art

[0002] As the scale of highway construction continues to expand and the complexity of projects continues to rise, massive amounts of multi-dimensional data are pouring in at all stages of the project. Design drawings have evolved from traditional two-dimensional CAD to BIM three-dimensional models, containing rich information such as geometry, materials, and progress; construction logs record daily on-site work details, problems, and solutions; sensors are spread across the construction site to provide real-time feedback on working conditions; construction material procurement lists control resource allocation and cost accounting; geographic information system data depicts the terrain, land use, and surrounding environment. However, in the past, there were many "islands" of data, and design, construction, and operation and maintenance data belonged to different departments and software platforms, with different formats and standards, which hindered collaborative management and in-depth analysis and mining, and made it difficult to efficiently support the refined management and control demands of modern highways.

[0003] Existing technologies mostly rely on manual statistics and simple chart viewing, and lack the ability to mine the associations of multi-source data. They do not dig deep into the logic of design and terrain adaptation and the root causes of construction problems. For example, it is difficult to see the intrinsic connection between frequent construction delays on a certain road section and the terrain undulations and material supply, resulting in delayed identification of abnormal sections and only treating the symptoms rather than the root causes. The project repeatedly falls into the same predicament, with rising costs and frequent delays. Route design adjustments are often based on subjective experience and local surveys, and data is not fully utilized to comprehensively evaluate the advantages and disadvantages of solutions based on terrain, construction, and cost. Summary of the invention

[0004] The purpose of the present invention is to provide an engineering project data management system and method based on data analysis to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for engineering project data management based on data analysis, the method comprising the following steps:

[0007] S100. In highway engineering projects, obtain engineering project data, including design drawings, construction logs, sensor data, construction material purchase lists and geographic information system data, and summarize them through interface programs; screen engineering project data for duplicate, erroneous or missing data, standardize and convert data in different formats, and unify the timestamp format;

[0008] S200. Based on the project data, the route design data of the expressway is extracted from the design drawings through AutoCAD Civil 3D, the terrain data is extracted based on the geographic information system data, and the route design data and the terrain data are spatially matched by using the coordinate conversion method; the elevation and slope of each point of the route design are compared with the terrain and landform at the corresponding position point by point, the compatibility analysis of the terrain type and the route direction is performed, and the abnormal sections are marked to form an abnormal section data set;

[0009] S300, extracting abnormal road section data features from the abnormal road section data set, including abnormal severity, road section length and terrain complexity coefficient, and using the Apriori algorithm to obtain the association patterns of the abnormal road section data features for the construction log and the construction material purchase list; using the K-Means clustering algorithm to cluster the association patterns into different categories according to similar features to form an abnormal road section problem combination;

[0010] S400, the sensor data is combined with the abnormal road section problem to obtain the fused data of the abnormal road section to be adjusted; the generative adversarial network algorithm architecture is selected to train and optimize the artificial intelligence model, the fused data of the abnormal road section to be adjusted is input into the trained artificial intelligence model, and the generator outputs several candidate route design adjustment plans; through the geographic information system, different candidate route design adjustment plans are simulated, evaluation results are generated, and the optimal route design adjustment plan is selected.

[0011] According to step S100, the design drawings include route plan drawings, longitudinal section drawings and design drawings of various structures. Through the drawing management interface software, a connection is established with the drawing repository of the design unit storing the design drawings, and the data is filtered and extracted according to the project stage, and the native data in CAD and BIM formats is converted into the XML structured description language form; the construction log includes weather conditions, attendance and division of labor of the construction team, operation time and failure of mechanical equipment, the amount of work completed on the day, progress of the construction process, and sudden problems and response measures; through the construction log entry application on the mobile terminal or computer terminal, a standardized template and data verification rules are built-in, and the background interface captures the construction log in real time; the sensor data is obtained by using the Internet of Things gateway device, and the sensor includes a vibration sensor for roadbed compaction monitoring, a temperature and humidity sensor for concrete pouring quality monitoring, and a stress and strain sensor for structural construction safety; through the open data interface of the enterprise's ERP system, the construction material procurement list is obtained by filtering according to the project; through the interface docking with the geographic information system, the geographic information system data of the corresponding geographical location area of ​​the highway project is obtained.

[0012] According to step S100, for design drawings, by comparing the drawing numbers, design stage identifiers, core layer names and content features of the design drawing files, similar geometric figures and annotation information are identified in combination with graphic recognition technology to filter out redundant drawing versions; for construction logs, field duplication is checked, and only the initial version is retained for several identical log records submitted by the same person for the same construction process on the same day; for sensor data, repeated collection data points caused by transmission failures or program abnormalities are eliminated based on the sensor numbers and collection time points;

[0013] In the construction material procurement list, set the value range for the numerical fields to screen for erroneous data; if the construction log lacks the completed work volume on some dates, the median filling method is used to complete it in combination with the construction logs of adjacent dates; for sensor data, during the period of collection interruption and missing, use the ARIMA model to fit the historical data trend forecast to fill the missing values; standardize data in different formats and unify the timestamp format.

[0014] According to step S200, the design drawings are imported into the AutoCAD Civil 3D environment, and the software automatically identifies different layers and object types in the drawings to extract the route design data of the expressway; through ArcGIS, the digital elevation model data and related vector layers covering the expressway project area in the geographic information system are imported into the workspace, and the elevation value, slope and slope direction of each unit are calculated through the spatial analysis tool of the geographic information system software to generate a continuous terrain slope and elevation thematic map, display the quantitative characteristics of the terrain and geomorphology, and extract the terrain and geomorphology data; extract the attributes of the vector layer, and associate the terrain and geomorphology data with the route design data;

[0015] A coordinate conversion tool is used to construct a conversion parameter model based on the common points measured in the two coordinate systems to which the route design data and the topographic data belong respectively. The Bursa seven-parameter model is used to convert different geodetic coordinate systems, involving translation, rotation and scaling parameters. The four-parameter model is used to adapt the local conversion of plane coordinates, and the coordinates described in the route design data are batch converted to a coordinate system consistent with the topographic data. According to the index of mileage and geographic orientation, each point in the route design is associated with the corresponding topographic point to achieve spatial matching.

[0016] According to step S200, based on the route design data and the terrain data after spatial matching, the elevation difference is calculated point by point, and the actual terrain elevation is subtracted from the route design elevation to obtain a series of difference sequences; the elevation difference threshold is set, and when the elevation difference of a certain point exceeds the corresponding threshold, it is marked; for the slope, the route design longitudinal slope is compared with the actual terrain slope, and the allowable range of the slope difference is formulated; when the slope comparison exceeds the threshold, it is marked;

[0017] Based on the vector layer and attribute data of the geographic information system, the terrain data are classified, and the fit between the route and the terrain is calculated in combination with the functional positioning of the expressway and the design speed. The relevant sections whose fit does not meet the threshold are marked; the marked abnormal sections are integrated to form an abnormal section data set.

[0018] According to step S300, the elevation difference threshold is set and slope difference threshold , for each abnormal road section i, count the number of points exceeding the elevation difference threshold , the number of points exceeding the slope difference threshold , and points will be deducted if the route direction does not match the terrain well. ; Abnormal severity The calculation formula is as follows:

[0019] ;

[0020] in, is the total number of points participating in the comparison of road section i, is the first weight coefficient, reflecting the contribution of elevation difference to the severity of anomaly. is the second weight coefficient, which reflects the contribution of the slope difference to the severity of the anomaly. is the third weight coefficient, reflecting the contribution of the fit between the line direction and the terrain to the severity of the anomaly;

[0021] Using the mileage data of route design, the length of abnormal section i By ending mileage station Subtract the starting mileage station It is obtained, expressed as , in kilometers; extract terrain slope, undulation and terrain fragmentation based on geographic information system; assume the average slope in the road section is The fluctuation is , terrain fragmentation index is , terrain complexity coefficient The calculation formula is:

[0022] ;

[0023] in, is the weight of the average slope, is the weight of the fluctuation, The weight of the terrain fragmentation index is set according to the impact of the terrain on the construction difficulty;

[0024] The Apriori algorithm scans the data features of abnormal road sections, counts the frequency of occurrence of each single item, compares it with the preset minimum support threshold min_sup, and selects frequent item sets; it performs self-connection and pruning operations on frequent item sets to iteratively generate higher-order frequent item sets;

[0025] Set the abnormal road segment dataset features as item sets , set the construction log structured data as item set ,Will and Integrate into transaction T; run the Apriori algorithm to calculate the support of the mined rules , the calculation formula is:

[0026] ;

[0027] in, is the number of transactions that simultaneously meet the characteristics of the abnormal road section dataset and the construction log, is the total number of transactions; confidence for:

[0028] ;

[0029] in, is the number of transactions that meet the characteristics of the abnormal road section dataset, and the confidence reflects the reliability of the rule; based on the support and confidence of the abnormal road section dataset characteristics and the construction log, the association pattern of the abnormal road section dataset characteristics and the construction log is obtained;

[0030] Run the Apriori algorithm to calculate the support and confidence of the abnormal road section dataset characteristics and the construction material procurement list; based on the support and confidence of the abnormal road section dataset characteristics and the construction material procurement list, obtain the association pattern of the abnormal road section dataset characteristics and the construction material procurement list.

[0031] According to step S300, based on the data characteristics of abnormal road sections and the association pattern of construction logs and construction material purchase lists, the K-Means clustering algorithm is used to determine the K value, randomly initialize K cluster centers, calculate the distance from each association pattern data point to the cluster center, and assign the data points to the cluster where the nearest cluster center is located based on the distance; after assignment, the new center of each cluster is recalculated, and this process is iterated repeatedly until the cluster center no longer changes and each cluster is a type of abnormal road section problem combination.

[0032] According to step S400, the sensor data and the abnormal road section problem combination are fused in the time and space dimension, and the real-time sensor data of the abnormal road section is fused with the road section problem combination according to the mileage and geographic coordinates of the road section; the timestamp is unified in time, and the collection time of the sensor data is aligned with the frequency time series of the problem occurrence in the abnormal road section problem combination; and the abnormal road section fusion data to be adjusted is obtained;

[0033] The generative adversarial network algorithm consists of a generator and a discriminator. The generator inputs the fused data of the abnormal section to be adjusted to generate a new route design adjustment plan, and outputs the plane coordinate string, longitudinal elevation sequence and bend and vertical curve parameters of the adjusted route to form several candidate route design adjustment plans; the training set and the verification set are divided, and the effective adjustment cases of past highway engineering projects that have been verified in practice are collected, including the corresponding abnormal section fused data and route design adjustment plans, for model training; the fused data of the abnormal section to be adjusted is input into the trained model, and the generator outputs several candidate route design adjustment plans.

[0034] In the stage of training the discriminator, the parameters of the generator are first fixed so that they do not change. The real route design adjustment schemes in the training set and the candidate route design adjustment schemes generated by the generator are input into the discriminator respectively. The goal of the discriminator is to judge whether the input data is real or generated as accurately as possible. For real data, it is hoped that the probability output by the discriminator is close to 1; for generated data, it is hoped that the output probability is close to 0. Therefore, the loss function of the discriminator measures the difference between the discriminator's discrimination result and the real label. The real data label is 1, and the generated data label is 0. The loss of the discriminator on the real data and the generated data is calculated, and the two are added to obtain the total discriminator loss. According to the calculated loss, the gradient of the loss function to the discriminator parameters is calculated through the back-propagation algorithm, and then the parameters of the discriminator are updated using the stochastic gradient descent method, so that the discriminator can better distinguish between real data and generated data.

[0035] During the generator training phase, the parameters of the discriminator are fixed. The fused data of the abnormal road section to be adjusted is input into the generator, and the generator generates several candidate route design adjustment plans. The goal of the generator is to generate candidate route design adjustment plans that can deceive the discriminator, that is, to let the discriminator judge the generated data as real data. Therefore, the generator's loss function is related to the output of the discriminator, and it is hoped that the higher the probability of the discriminator's output of the generated data, the better. The generator's loss function is constructed by the discriminator's output probability of the generated data, using -log (D (G (z))), where D is the discriminator, G is the generator, and z is the input fused data of the abnormal road section to be adjusted. Similarly, the back-propagation algorithm is used to calculate the gradient of the generator's loss function to the generator's parameters, and then the optimization algorithm is used to update the generator's parameters, so that the candidate route design adjustment plan generated by the generator is closer to the real effective adjustment plan, so that it can better deceive the discriminator.

[0036] Repeat the steps of training the discriminator and generator alternately for multiple rounds of training. As the training progresses, the generator continuously learns to generate more reasonable candidate route design adjustment plans, and the discriminator continuously improves its ability to distinguish between real data and generated data. After each training cycle or a certain number of rounds, the model is evaluated using the validation set. The fused data of the abnormal road sections to be adjusted in the validation set is input into the trained model, and the generator outputs the candidate route design adjustment plan. The performance of the model on unseen data is then evaluated through the discriminator and other evaluation indicators. Based on the evaluation results, the model's hyperparameters are adjusted to optimize the model's performance.

[0037] According to step S400, the candidate route design adjustment plans are placed on the real terrain one by one, and the degree of fit between the candidate route and the terrain is calculated; whether the route direction conforms to the contour trend is analyzed, and the slope analysis function of the geographic information system is used to compare the designed longitudinal slope with the actual terrain slope, and the proportion of the length of the section exceeding the normal slope difference is counted, and the higher the proportion, the lower the degree of fit between the candidate route and the terrain; the spatial analysis of the geographic information system is combined with the engineering measurement algorithm to estimate the construction volume of each plan; the optimal route design adjustment plan is selected based on the degree of fit between the candidate route and the terrain and the construction volume of the plan.

[0038] A data management system for engineering projects based on data analysis, comprising:

[0039] Data collection and preprocessing module: including: data collection unit and data preprocessing unit, wherein the data collection unit obtains project data in highway projects, including design drawings, construction logs, sensor data, construction material purchase lists and geographic information system data, and summarizes them through interface programs; the data preprocessing unit screens the project data for duplicate, erroneous or missing data, standardizes and converts data in different formats, and unifies the timestamp format;

[0040] Terrain and design data processing module: including: route design data extraction unit, terrain and geomorphology data extraction unit, spatial matching unit and abnormal road section marking unit; wherein, the route design data extraction unit extracts the route design data of the expressway from the design drawings through AutoCAD Civil 3D based on the project data, the terrain and geomorphology data extraction unit extracts the terrain and geomorphology data based on the geographic information system data, and the spatial matching unit uses the coordinate conversion method to spatially match the route design data and the terrain and geomorphology data; the abnormal road section marking unit compares the elevation and slope of each point of the route design with the terrain and geomorphology at the corresponding position point by point, analyzes the compatibility of the terrain type and the route direction, marks the abnormal road section, and forms the abnormal road section data set;

[0041] Abnormal road section data mining module: including: feature extraction unit, association rule mining unit and cluster analysis unit; the feature extraction unit extracts abnormal road section data features from the abnormal road section data set, including the severity of the abnormality, the length of the road section and the terrain complexity coefficient; the association rule mining unit uses the Apriori algorithm to obtain the association patterns of the abnormal road section data features for the construction log and the construction material purchase list; the cluster analysis unit uses the K-Means clustering algorithm to cluster the association patterns into different categories according to similar features to form an abnormal road section problem combination;

[0042] Route design adjustment module: includes: data fusion unit, generative adversarial network modeling and solution generation unit and solution simulation evaluation and optimization unit; among them, the data fusion unit fuses the sensor data with the abnormal road section problem combination to obtain the fused data of the abnormal road section to be adjusted; the generative adversarial network modeling and solution generation unit selects the generative adversarial network algorithm architecture to train and optimize the artificial intelligence model, inputs the fused data of the abnormal road section to be adjusted into the trained artificial intelligence model, and the generator outputs several candidate route design adjustment solutions; the solution simulation evaluation and optimization unit simulates different candidate route design adjustment solutions through the geographic information system, generates evaluation results, and selects the optimal route design adjustment solution.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention uses professional software to extract key design and terrain data, realizes spatial matching through coordinate conversion, compares elevation and slope point by point, and comprehensively evaluates the adaptability of terrain type and line direction, accurately marks abnormal sections that do not meet design expectations and are difficult to construct, and forms a data set.

[0045] 2. The present invention extracts features from the abnormal road section data set, uses the Apriori algorithm to mine the hidden patterns of the associated construction logs and purchase lists, and then uses K-Means clustering to sort out similar problem combinations. It clearly presents the internal logic chain of "complex terrain - low construction efficiency - tight material supply", penetrates the surface and directly hits the root of the problem, avoiding blind policy implementation and repeated trial and error.

[0046] 3. The present invention combines real-time sensor information with abnormal road section problems and inputs them into a generative adversarial network model. After training, the generator generates multiple candidate route design adjustment plans. GAN breaks through the traditional experience reliance and generates routes that take into account terrain adaptation, structural stability, and convenient construction under complex terrain and working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the steps of a data management method for an engineering project based on data analysis of the present invention;

[0048] Figure 2 It is a system structure diagram of an engineering project data management system based on data analysis of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Example: Figure 1-Figure 2 As shown, the present invention provides a technical solution:

[0051] According to one embodiment of the present invention, Figure 1 A step diagram of a method for managing engineering project data based on data analysis is shown in the figure. A method for managing engineering project data based on data analysis comprises the following steps:

[0052] S100. In highway engineering projects, obtain engineering project data, including design drawings, construction logs, sensor data, construction material purchase lists and geographic information system data, and summarize them through interface programs; screen engineering project data for duplicate, erroneous or missing data, standardize and convert data in different formats, and unify the timestamp format;

[0053] S200. Based on the engineering project data, the route design data of the expressway is extracted from the design drawings through AutoCAD Civil 3D, the terrain data is extracted based on the geographic information system data, and the route design data and the terrain data are spatially matched by using the coordinate conversion method; the elevation and slope of each point of the route design are compared with the terrain and landform at the corresponding position point by point, the compatibility analysis of the terrain type and the route direction is performed, and the abnormal sections are marked to form an abnormal section data set;

[0054] S300, extracting abnormal road section data features from the abnormal road section data set, including abnormal severity, road section length and terrain complexity coefficient, and using the Apriori algorithm to obtain the association patterns of the abnormal road section data features for the construction log and the construction material purchase list; using the K-Means clustering algorithm to cluster the association patterns into different categories according to similar features to form an abnormal road section problem combination;

[0055] S400, the sensor data is combined with the abnormal road section problem to obtain the fused data of the abnormal road section to be adjusted; the generative adversarial network algorithm architecture is selected to train and optimize the artificial intelligence model, the fused data of the abnormal road section to be adjusted is input into the trained artificial intelligence model, and the generator outputs several candidate route design adjustment plans; through the geographic information system, different candidate route design adjustment plans are simulated, evaluation results are generated, and the optimal route design adjustment plan is selected.

[0056] According to step S100, the design drawings include route plan drawings, longitudinal section drawings and design drawings of various structures. Through the drawing management interface software, a connection is established with the drawing repository of the design unit storing the design drawings, and the data is filtered and extracted according to the project stage, and the native data in CAD and BIM formats is converted into the XML structured description language form; the construction log includes weather conditions, attendance and division of labor of the construction team, operation time and failure of mechanical equipment, the amount of work completed on the day, progress of the construction process, and sudden problems and response measures; through the construction log entry application on the mobile terminal or computer terminal, a standardized template and data verification rules are built-in, and the background interface captures the construction log in real time; the sensor data is obtained by using the Internet of Things gateway device, and the sensor includes a vibration sensor for roadbed compaction monitoring, a temperature and humidity sensor for concrete pouring quality monitoring, and a stress and strain sensor for structural construction safety; through the open data interface of the enterprise's ERP system, the construction material procurement list is obtained by filtering according to the project; through the interface docking with the geographic information system, the geographic information system data of the corresponding geographical location area of ​​the highway project is obtained.

[0057] According to step S100, for design drawings, by comparing the drawing numbers, design stage identifiers, core layer names and content features of the design drawing files, similar geometric figures and annotation information are identified in combination with graphic recognition technology to filter out redundant drawing versions; for construction logs, field duplication is checked, and only the initial version is retained for several identical log records submitted by the same person for the same construction process on the same day; for sensor data, repeated collection data points caused by transmission failures or program abnormalities are eliminated based on the sensor numbers and collection time points;

[0058] In the construction material procurement list, set the value range for the numerical fields to screen for erroneous data; if the construction log lacks the completed work volume on some dates, the median filling method is used to complete it in combination with the construction logs of adjacent dates; for sensor data, during the period of collection interruption and missing, use the ARIMA model to fit the historical data trend forecast to fill the missing values; standardize data in different formats and unify the timestamp format.

[0059] According to step S200, the design drawings are imported into the AutoCAD Civil 3D environment, and the software automatically identifies different layers and object types in the drawings to extract the route design data of the expressway; through ArcGIS, the digital elevation model data and related vector layers covering the expressway project area in the geographic information system are imported into the workspace, and the elevation value, slope and slope direction of each unit are calculated through the spatial analysis tool of the geographic information system software to generate a continuous terrain slope and elevation thematic map, display the quantitative characteristics of the terrain and geomorphology, and extract the terrain and geomorphology data; extract the attributes of the vector layer, and associate the terrain and geomorphology data with the route design data;

[0060] A coordinate conversion tool is used to construct a conversion parameter model based on the common points measured in the two coordinate systems to which the route design data and the topographic data belong respectively. The Bursa seven-parameter model is used to convert different geodetic coordinate systems, involving translation, rotation and scaling parameters. The four-parameter model is used to adapt the local conversion of plane coordinates, and the coordinates described in the route design data are batch converted to a coordinate system consistent with the topographic data. According to the index of mileage and geographic orientation, each point in the route design is associated with the corresponding topographic point to achieve spatial matching.

[0061] In this embodiment, we focus on a highway project under construction in a mountainous area. The route is about 30 kilometers long, with large undulations, covering many peaks, valleys, and river crossing areas. The terrain is highly complex, and the route design and terrain adaptability are strictly required. Obtain the detailed design drawings of the highway project in the format of .dwg, covering the route plan design drawing, longitudinal section design drawing, and various structural design details (bridges, culverts, etc.). The drawings follow the local independent coordinate system, and the route design and annotation are based on this coordinate system. Obtain the project area geographic information system data from the local surveying and mapping department, where the digital elevation model (DEM) has a resolution of 30 meters × 30 meters and records terrain elevation information; the vector layer contains rich information such as land use type, water system distribution, contour lines, etc. The geographic coordinates use the national 2000 geodetic coordinate system, which is different from the design drawing coordinate system.

[0062] Import the design drawings into the AutoCAD Civil 3D environment, and the software automatically identifies multiple layers. After screening, it locks the key layers, such as the "route centerline" layer (containing the coordinates of the route plane control points), the "longitudinal section design" layer (recording the design elevation of each mileage pile number), and the "horizontal curve element" layer (detailed marking of the curve radius, transition curve parameters, etc.).

[0063] An example of extracting critical route design data is as follows (partial excerpts are shown):

[0064] Route starting point coordinates (local independent coordinate system): X = 12345.67, Y = 23456.78, elevation = 500.00 meters;

[0065] At the first bend: bend radius R = 600 meters, transition curve length Ls = 100 meters, center coordinates (local independent coordinate system) X = 13000.00, Y = 24000.00;

[0066] The design elevation at stake number K5 + 000 = 520.00 meters, the design elevation at stake number K10 + 000 = 550.00 meters, and so on. Multiple sets of data are extracted at certain mileage intervals to fully outline the vertical trend of the route.

[0067] The DEM data and related vector layers covering the project area were imported into the ArcGIS Pro workspace. The “Slope” tool under the “Terrain Analysis” toolset was used to set the analysis resolution to be consistent with the DEM (30 m × 30 m). The slope of each unit was calculated and a slope thematic map was generated. Statistics showed that the slope ranged from 0 to 45%. The slope of some steep slope areas (such as around mountain peaks) exceeded 30%, and the slope at the bottom of the valley was mostly between 5 and 10%.

[0068] Through the "Elevation" function of the "Surface Analysis" tool, the elevation values ​​of each terrain unit are obtained to generate an elevation thematic map. The lowest elevation in the area is about 300 meters (river valley) and the highest elevation is 800 meters (mountain top). At the same time, the attributes of the contour vector layer are extracted, such as the contour interval of 20 meters, which clearly shows the rhythm of terrain undulations.

[0069] The attributes of the water system vector layer are extracted to clarify the direction, width and other information of the river, and to locate sections of road with potential water damage risks; the land use type layer marks the distribution of farmland and woodland, identifies ecologically sensitive areas and construction-restricted areas, and reserves a rich terrain data foundation for subsequent analysis related to route design.

[0070] Five public points with accurate coordinates (such as surveying benchmark stakes and landmark control points) were selected in the project area. Accurate coordinates were recorded in both coordinate systems. Sample public point data (partial display):

[0071] Public point 1: X = 12500.00, Y = 23800.00 in the local independent coordinate system, X = 43210.00, Y = 54320.00 in the national 2000 coordinate system;

[0072] Public point 2: X = 13200.00, Y = 24300.00 in the local independent coordinate system, X = 44000.00, Y = 55000.00 in the national 2000 coordinate system, and so on.

[0073] The coordinate conversion tool provided by ArcGIS Pro was used to construct the Bursa seven-parameter model based on these common points (the parameters were strictly calculated and calibrated, the translation parameters in the X, Y, and Z directions were approximately [Xt = 30000.00, Yt = 30000.00, Zt = 500.00], the rotation parameters [Rx = 0.02°, Ry = 0.03°, Rz = 0.01°], and the scaling parameter S = 1.0005) for geodetic coordinate system conversion; for local fine adjustment of plane coordinates, a four-parameter model (translation parameters [Tx = 100.00, Ty = 150.00], rotation parameter R = 0.05°, scaling parameter K = 1.002) was used to batch convert the route design data coordinates to the National 2000 coordinate system.

[0074] According to the mileage pile number and geographic orientation index, each route design point is matched with the corresponding terrain point one by one to achieve precise spatial association. For example, the coordinates of the route pile number K8 + 000 design point after conversion accurately correspond to the corresponding terrain elevation point (latitude and longitude positioning) under the same spatial reference.

[0075] According to step S200, based on the route design data and the terrain data after spatial matching, the elevation difference is calculated point by point, and the actual terrain elevation is subtracted from the route design elevation to obtain a series of difference sequences; the elevation difference threshold is set, and when the elevation difference of a certain point exceeds the corresponding threshold, it is marked; for the slope, the route design longitudinal slope is compared with the actual terrain slope, and the allowable range of the slope difference is formulated; when the slope comparison exceeds the threshold, it is marked;

[0076] Based on the vector layer and attribute data of the geographic information system, the terrain data are classified, and the fit between the route and the terrain is calculated in combination with the functional positioning of the expressway and the design speed. The relevant sections whose fit does not meet the threshold are marked; the marked abnormal sections are integrated to form an abnormal section data set.

[0077] Based on the geographic information system vector layer and attribute data, the topography of the project area is classified as follows (the proportion of each type is calculated by area):

[0078] Mountainous areas: accounting for about 60%, characterized by undulating terrain, dense contour lines, and a slope generally between 15% and 45%. There are multiple peaks and valleys, and some areas have exposed rocks, which have a great impact on the difficulty of engineering construction. For example, in some mountain excavation sections, a large amount of blasting operations and slope protection measures are required.

[0079] Hills: accounting for about 30%, with a certain degree of undulation in the terrain, with slopes mostly between 5% and 15%, and relatively sparse contour lines. The land use type is mostly forest land and a small amount of farmland. During construction, it is necessary to consider the balance of filling and excavation as well as the occupation of farmland.

[0080] River valley: accounts for about 10%, with low terrain and rivers passing through it. The terrain on both sides is relatively flat, but there is a risk of flooding. It is necessary to consider the flood control design of bridges and embankments. For example, the width of the river reaches 50-100 meters in some sections, and it is necessary to design bridges with appropriate spans.

[0081] This expressway is a regional traffic artery, and its function is to connect multiple cities, bear a large traffic flow, and the design speed is 100km / h. According to the design speed requirements, the radius of the horizontal curve should generally not be less than 700 meters (limit value), the maximum longitudinal slope should not exceed 4%, and the vertical curve radius should not be less than 4500 meters for concave vertical curves and should not be less than 10000 meters for convex vertical curves.

[0082] For the K5-K8 kilometer section (located in the mountains), the route is perpendicular to the contour lines in many places, resulting in a large number of high-fill and deep-excavation projects, and the design slope frequently exceeds the reasonable slope range of the terrain (such as the slope difference example mentioned above). In addition, this section crosses multiple peaks, increasing the possibility of bridge and tunnel projects. Considering the construction difficulty, driving safety and comfort, the route of this section has a low degree of fit with the terrain, and the fit threshold is set to 0.6 (the full score is 1.0, which is obtained by weighted scoring of various indicators, and the specific weight is set according to engineering experience and specifications). The score of this section is only 0.3.

[0083] In the K10-K12 kilometer section (in the river valley area), although the terrain is relatively flat, the line did not fully utilize the existing terrain conditions to reasonably select the bridge location when crossing the river, resulting in the bridge span being too large and increasing the construction cost. At the same time, some embankment sections are close to the river, posing a risk of flooding. The fit score is about 0.4, which is lower than the threshold standard.

[0084] According to the set elevation difference threshold (±1.0 m in mountainous areas), the elevation differences calculated above were screened and marked, and it was found that the elevation differences of sections such as K4+000 - K4+040 exceeded the threshold. These sections may face a large number of filling or excavation adjustments in subsequent construction, and it is necessary to focus on their earthwork allocation and the impact on the surrounding environment. A total of about 3.5 kilometers of sections with elevation differences that do not meet the threshold were marked, distributed in 7 different sections.

[0085] Based on the allowable range of slope difference (±2.0%), sections such as K5+000 - K6+040 where the slope difference exceeds the threshold are marked. Driving safety and comfort on these sections may be affected, and vehicles may need to slow down or accelerate frequently during driving. The slope also needs to be adjusted during construction, involving measures such as re-excavation of the roadbed or filling reinforcement. The total length of such abnormal slope sections marked is about 4.2 kilometers, involving 5 sections.

[0086] By calculating the degree of fit between the route and the terrain and comparing it with the threshold (0.6), in addition to the K5-K8 km section and the K10-K12 km section mentioned above, there are several other sections with similar fit issues. The total length of the sections marked whose fit does not meet the threshold is approximately 9.8 kilometers, accounting for approximately 32.7% of the total route length (30 kilometers).

[0087] According to step S300, the elevation difference threshold is set to 1.0 meter and the slope difference threshold is set to 2.0%. Based on expert experience and comprehensive consideration of this project, the first weight coefficient is set to 0.4, the second weight coefficient is set to 0.3, and the third weight coefficient is set to 0.3.

[0088] For abnormal section i (pile number is K4+000-K4+040):

[0089] The total number of points involved in the comparison of this road section (calculated as one comparison point every 20 meters) is 20 points.

[0090] The number of points exceeding the elevation difference threshold (1.0 meter) is 12 points. The number of points exceeding the slope difference threshold (2.0%) is 8 points. Based on the previous evaluation of the fit between the route and the terrain, this section of road will be deducted 5 points for poor fit (out of 10 points, the lower the score, the worse the fit, and the higher the deduction).

[0091] For abnormal section j (pile number is K5+000-K6+040):

[0092] The total number of points involved in the comparison is 32 points. The number of points exceeding the elevation difference threshold is 18 points. The number of points exceeding the slope difference threshold is 15 points. If the line direction does not fit the terrain well, 8 points will be deducted.

[0093] The ending mileage post number of abnormal section i is K4+040, and the starting mileage post number is K4+000, so the length of the section is 0.4 kilometers.

[0094] The ending mileage post number of abnormal section j is K6+040, the starting mileage post number is K5+000, and the section length is 1.4 kilometers.

[0095] According to the influence of terrain on construction difficulty, the weight of average slope is set to 0.5, the weight of undulation is set to 0.3, and the weight of terrain fragmentation index is set to 0.2.

[0096] According to the analysis of geographic information system, the area where abnormal road section i is located has an average slope of 12%, an undulation of 5 meters, and a terrain fragmentation index of 0.4.

[0097] According to the analysis of geographic information system, the area where abnormal road section j is located has an average slope of 15%, an undulation of 8 meters, and a terrain fragmentation index of 0.6.

[0098] The Apriori algorithm scans the data features of abnormal road sections, counts the frequency of occurrence of each single item, compares it with the preset minimum support threshold min_sup=0.2, and selects frequent item sets; performs self-connection and pruning operations on frequent item sets to iteratively generate higher-order frequent item sets; finds that the frequency of occurrence of the item set "abnormal severity greater than 3 and terrain complexity coefficient greater than 8" meets the threshold requirement and becomes a frequent 1-item set. On this basis, self-connection and pruning operations are performed to iteratively generate higher-order frequent item sets (such as "abnormal severity greater than 3 and terrain complexity coefficient greater than 8 and construction efficiency less than 80%").

[0099] Set the abnormal road segment dataset features as item sets , set the construction log structured data as item set ,Will and Integrate into transaction T; the following are some transaction examples:

[0100] Transaction 1: The abnormal severity of the medium abnormal section (pile number K4+000-K4+040) is 6.2, the section length is 0.4 kilometers, and the terrain complexity coefficient is 7.8; The construction efficiency was 80%, the downtime on that day was 2 hours, and the type of construction problem was "difficulty in compacting the fill".

[0101] Transaction 2: The severity of the abnormality of the medium abnormal section (pile number K5+000-K6+040) is 8.5, the length of the section is 1.4 kilometers, and the terrain complexity coefficient is 10.2; The construction efficiency was 70%, the downtime on that day was 3 hours, and the type of construction problem was "insufficient slope stability".

[0102] Run the Apriori algorithm to calculate the support of the mined rules , the number of transactions that meet the conditions of "the severity of the abnormality is greater than 3 and the complexity coefficient of the terrain is greater than 8" is 20, and the number of transactions that meet the conditions of "the severity of the abnormality is greater than 3 and the complexity coefficient of the terrain is greater than 8" and "the downtime on the same day exceeds 2 hours" is 12, and the total number of transactions is 50. The calculation formula is:

[0103] ;

[0104] in, is the number of transactions that simultaneously meet the characteristics of the abnormal road section dataset and the construction log, is the total number of transactions; confidence for:

[0105] ;

[0106] in, is the number of transactions that meet the characteristics of the abnormal road section dataset, and the confidence reflects the reliability of the rule; based on the support and confidence of the abnormal road section dataset characteristics and the construction log, the association pattern of the abnormal road section dataset characteristics and the construction log is obtained;

[0107] This shows that the rule has a certain degree of reliability. That is, when the abnormal road section has an "abnormal severity greater than 3 and a terrain complexity coefficient greater than 8", there is a 60% probability that the work stoppage time on that day will exceed 2 hours. This reveals the potential correlation between the characteristics of the abnormal road section and the construction log records, which helps to predict possible problems in construction in advance and reasonably arrange construction plans and resource allocation.

[0108] The Apriori algorithm is also run, and after integration with the construction material procurement list item set, the rule "Road section length is greater than 1 km and the abnormality severity is greater than 3 → steel overconsumption ratio exceeds 10%" is mined. Its support and confidence are calculated in a similar way to the above method. This result reflects the connection between the abnormal road section characteristics and the construction material consumption situation, which can assist in construction material management and prepare material reserves and cost control measures in advance.

[0109] According to step S300, based on the data characteristics of abnormal road sections and the association pattern of construction logs and construction material purchase lists, the K-Means clustering algorithm is used to determine the K value, randomly initialize K cluster centers, calculate the distance from each association pattern data point to the cluster center, and assign the data points to the cluster where the nearest cluster center is located based on the distance; after assignment, the new center of each cluster is recalculated, and this process is iterated repeatedly until the cluster center no longer changes and each cluster is a type of abnormal road section problem combination.

[0110] According to step S400, the sensor data and the abnormal road section problem combination are fused in the time and space dimension, and the real-time sensor data of the abnormal road section is fused with the road section problem combination according to the mileage and geographic coordinates of the road section; the timestamp is unified in time, and the collection time of the sensor data is aligned with the frequency time series of the problem occurrence in the abnormal road section problem combination; and the abnormal road section fusion data to be adjusted is obtained;

[0111] The generative adversarial network algorithm consists of a generator and a discriminator. The generator inputs the fused data of the abnormal section to be adjusted to generate a new route design adjustment plan, and outputs the plane coordinate string, longitudinal elevation sequence and bend and vertical curve parameters of the adjusted route to form several candidate route design adjustment plans; the training set and the verification set are divided, and the effective adjustment cases of past highway engineering projects that have been verified in practice are collected, including the corresponding abnormal section fused data and route design adjustment plans, for model training; the fused data of the abnormal section to be adjusted is input into the trained model, and the generator outputs several candidate route design adjustment plans.

[0112] According to step S400, the candidate route design adjustment plans are placed on the real terrain one by one, and the degree of fit between the candidate route and the terrain is calculated; whether the route direction conforms to the contour trend is analyzed, and the slope analysis function of the geographic information system is used to compare the designed longitudinal slope with the actual terrain slope, and the proportion of the length of the section exceeding the normal slope difference is counted, and the higher the proportion, the lower the degree of fit between the candidate route and the terrain; the spatial analysis of the geographic information system is combined with the engineering measurement algorithm to estimate the construction volume of each plan; the optimal route design adjustment plan is selected based on the degree of fit between the candidate route and the terrain and the construction volume of the plan.

[0113] According to another embodiment of the present invention, Figure 2 As shown in the system structure diagram of a project data management system based on data analysis, a project data management system based on data analysis includes:

[0114] Data collection and preprocessing module: including: data collection unit and data preprocessing unit, wherein the data collection unit obtains project data in highway projects, including design drawings, construction logs, sensor data, construction material purchase lists and geographic information system data, and summarizes them through interface programs; the data preprocessing unit screens the project data for duplicate, erroneous or missing data, standardizes and converts data in different formats, and unifies the timestamp format;

[0115] Terrain and design data processing module: including: route design data extraction unit, terrain and geomorphology data extraction unit, spatial matching unit and abnormal road section marking unit; wherein, the route design data extraction unit extracts the route design data of the expressway from the design drawings through AutoCAD Civil 3D based on the project data, the terrain and geomorphology data extraction unit extracts the terrain and geomorphology data based on the geographic information system data, and the spatial matching unit uses the coordinate conversion method to spatially match the route design data and the terrain and geomorphology data; the abnormal road section marking unit compares the elevation and slope of each point of the route design with the terrain and geomorphology at the corresponding position point by point, analyzes the compatibility of the terrain type and the route direction, marks the abnormal road section, and forms the abnormal road section data set;

[0116] Abnormal road section data mining module: including: feature extraction unit, association rule mining unit and cluster analysis unit; the feature extraction unit extracts abnormal road section data features from the abnormal road section data set, including the severity of the abnormality, the length of the road section and the terrain complexity coefficient; the association rule mining unit uses the Apriori algorithm to obtain the association patterns of the abnormal road section data features for the construction log and the construction material purchase list; the cluster analysis unit uses the K-Means clustering algorithm to cluster the association patterns into different categories according to similar features to form an abnormal road section problem combination;

[0117] Route design adjustment module: includes: data fusion unit, generative adversarial network modeling and solution generation unit and solution simulation evaluation and optimization unit; among them, the data fusion unit fuses the sensor data with the abnormal road section problem combination to obtain the fused data of the abnormal road section to be adjusted; the generative adversarial network modeling and solution generation unit selects the generative adversarial network algorithm architecture to train and optimize the artificial intelligence model, inputs the fused data of the abnormal road section to be adjusted into the trained artificial intelligence model, and the generator outputs several candidate route design adjustment solutions; the solution simulation evaluation and optimization unit simulates different candidate route design adjustment solutions through the geographic information system, generates evaluation results, and selects the optimal route design adjustment solution.

[0118] In this embodiment:

[0119] Collect data from 20 successful adjustment cases of similar abnormal sections of mountain highways in the past, split 80% as training set and 20% as validation set. Each case contains corresponding fusion data and adjusted route design parameters. Build a GAN model, the generator uses a 4-layer fully connected neural network, the input layer receives fusion data with a dimension of 20, and after hidden layer mapping, the output layer outputs the adjusted route design parameters (plane coordinate string dimension 10, longitudinal section elevation sequence dimension 8, bend and vertical curve parameter dimension 5); the discriminator is a 3-layer convolutional neural network, which is used to judge the authenticity of the input scheme and output 0-1 judgment probability. Set the learning rate to 0.001 and iterate the training for 500 rounds.

[0120] In the early stage of training, the generator outputs random solutions and the discriminator has low accuracy. As the training progresses, the discriminator's accuracy in identifying the real solution exceeds 90%, and the solutions generated by the generator gradually become reasonable. Finally, 5 candidate route design adjustment solutions are generated. Examples of key parameters of some solutions are as follows:

[0121] Option 1: In the plane, the original curve radius of the K11+000 - K12+000 section will be widened from 500 meters to 800 meters, and the length of the transition curve will be increased by 30 meters. The adjusted route will better conform to the terrain trend of the valley and reduce the cutting of the mountain. In the longitudinal section, the average longitudinal slope of the K13+000 - K14+000 section will be reduced from 4.5% to 3.5%, reducing the difficulty of heavy-loaded vehicles climbing the slope and alleviating the stress concentration on the road surface.

[0122] Option 2: The K10 + 500 - K11 + 500 section is moved 50 meters westward to avoid the unstable area with high slopes. The new route passes through an area with a gentle slope. The longitudinal slope is optimized to a combination of multiple gentle slopes (maximum longitudinal slope 3.2%), and the corresponding vertical curve radius is increased to improve driving comfort and safety.

[0123] The five candidate plans were imported into ArcGIS Pro geographic information system, and simulation parameters were set in combination with the engineering analysis plug-in. Taking terrain adaptability into consideration, the balance threshold of excavation and filling was set according to the mountainous area standard (the difference between excavation and filling volume was less than 10% for optimality); in terms of construction difficulty, the engineering volume (earthwork volume, structural concrete volume) and construction period (combined with construction quota and mechanical efficiency) were estimated; in terms of environmental impact, indicators such as ecological forest land occupation and vegetation destruction were counted, and the terrain adaptability, construction difficulty, and environmental impact were assigned weights of 0.4, 0.4, and 0.2 to construct an evaluation system.

[0124] After simulation calculation, the quantitative scores of each scheme are as follows:

[0125] Option 1: Terrain adaptability score is 7 points (fill and cut balance rate is 8%), construction difficulty score is 6 points (earthwork volume is reduced by 8%, construction period is shortened by 5 days), environmental impact score is 8 points (15% less damage to forest land), and the comprehensive weighted score is 7*0.4+6*0.4+8*0.2=6.8 points.

[0126] Option 2: Terrain adaptability 8 points (fill and cut balance rate 5%), construction difficulty 7 points (earthwork volume reduced by 12%, construction period shortened by 8 days), environmental impact 7 points (12% less damage to forest land), comprehensive score 8*0.4+7*0.4+7*0.2=7.4 points.

[0127] Schemes 3-5: The comprehensive scores were 6.2, 6.5, and 6.0 respectively. After comparison, Scheme 2 had the highest comprehensive score and was selected as the optimal route design adjustment scheme, which takes into account terrain fit, construction efficiency, and eco-friendliness. The design drawings will be updated accordingly, and the construction plan will be compiled to promote the optimization and implementation of the project.

[0128] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A data management method for engineering projects based on data analysis, characterized in that: The method comprises the following steps: S100. In highway engineering projects, obtain engineering project data, including design drawings, construction logs, sensor data, construction material purchase lists and geographic information system data, and summarize them through interface programs; screen engineering project data for duplicate, erroneous or missing data, standardize and convert data in different formats, and unify the timestamp format; S200. Based on the engineering project data, the route design data of the expressway is extracted from the design drawings through AutoCAD Civil 3D, the terrain data is extracted based on the geographic information system data, and the route design data and the terrain data are spatially matched by using the coordinate conversion method; the elevation and slope of each point of the route design are compared with the terrain and landform at the corresponding position point by point, the compatibility analysis of the terrain type and the route direction is performed, and the abnormal sections are marked to form an abnormal section data set; S300, extracting abnormal road section data features from the abnormal road section data set, including abnormal severity, road section length and terrain complexity coefficient, and using the Apriori algorithm to obtain the association patterns of the abnormal road section data features for the construction log and the construction material purchase list; using the K-Means clustering algorithm to cluster the association patterns into different categories according to similar features to form an abnormal road section problem combination; S400, the sensor data is combined with the abnormal road section problem to obtain the fused data of the abnormal road section to be adjusted; the generative adversarial network algorithm architecture is selected to train and optimize the artificial intelligence model, the fused data of the abnormal road section to be adjusted is input into the trained artificial intelligence model, and the generator outputs several candidate route design adjustment plans; through the geographic information system, different candidate route design adjustment plans are simulated, evaluation results are generated, and the optimal route design adjustment plan is selected.

2. The engineering project data management method based on data analysis according to claim 1, characterized in that: According to step S100, the design drawings include route plan drawings, longitudinal section drawings and design drawings of various structures. Through the drawing management interface software, a connection is established with the drawing repository of the design unit storing the design drawings, and the data is filtered and extracted according to the project stage, and the native data in CAD and BIM formats is converted into the XML structured description language form; the construction log includes weather conditions, attendance and division of labor of the construction team, operation time and failure of mechanical equipment, the amount of work completed on the day, progress of the construction process, and sudden problems and response measures; through the construction log entry application on the mobile terminal or computer terminal, a standardized template and data verification rules are built-in, and the background interface captures the construction log in real time; the sensor data is obtained by using the Internet of Things gateway device, and the sensor includes a vibration sensor for roadbed compaction monitoring, a temperature and humidity sensor for concrete pouring quality monitoring, and a stress and strain sensor for structural construction safety; through the open data interface of the enterprise's ERP system, the construction material procurement list is obtained by filtering according to the project; through the interface docking with the geographic information system, the geographic information system data of the corresponding geographical location area of ​​the highway project is obtained.

3. The engineering project data management method based on data analysis according to claim 2 is characterized in that: According to step S100, for design drawings, by comparing the drawing numbers, design stage identifiers, core layer names and content features of the design drawing files, similar geometric figures and annotation information are identified in combination with graphic recognition technology to filter out redundant drawing versions; for construction logs, field duplication is checked, and only the initial version is retained for several identical log records submitted by the same person for the same construction process on the same day; for sensor data, repeated collection data points caused by transmission failures or program abnormalities are eliminated based on the sensor numbers and collection time points; In the construction material procurement list, set the value range for the numerical fields to screen for erroneous data; if the construction log lacks the completed work volume on some dates, the median filling method is used to complete it in combination with the construction logs of adjacent dates; for sensor data, during the period of collection interruption and missing, use the ARIMA model to fit the historical data trend forecast to fill the missing values; standardize data in different formats and unify the timestamp format.

4. The engineering project data management method based on data analysis according to claim 1, characterized in that: According to step S200, the design drawings are imported into the AutoCAD Civil 3D environment, and the software automatically identifies different layers and object types in the drawings to extract the route design data of the expressway; through ArcGIS, the digital elevation model data and related vector layers covering the expressway project area in the geographic information system are imported into the workspace, and the elevation value, slope and slope direction of each unit are calculated through the spatial analysis tool of the geographic information system software to generate a continuous terrain slope and elevation thematic map, display the quantitative characteristics of the terrain and geomorphology, and extract the terrain and geomorphology data; extract the attributes of the vector layer, and associate the terrain and geomorphology data with the route design data; A coordinate conversion tool is used to construct a conversion parameter model based on the common points measured in the two coordinate systems to which the route design data and the topographic data belong respectively. The Bursa seven-parameter model is used to convert different geodetic coordinate systems, involving translation, rotation and scaling parameters. The four-parameter model is used to adapt the local conversion of plane coordinates, and the coordinates described in the route design data are batch converted to a coordinate system consistent with the topographic data. According to the index of mileage and geographic orientation, each point in the route design is associated with the corresponding topographic point to achieve spatial matching.

5. The engineering project data management method based on data analysis according to claim 4 is characterized in that: According to step S200, based on the route design data and the terrain data after spatial matching, the elevation difference is calculated point by point, and the route design elevation is subtracted from the actual terrain elevation to obtain a series of difference sequences; Set the elevation difference threshold. When the elevation difference of a point exceeds the corresponding threshold, mark it. For slope, compare the designed longitudinal slope of the route with the actual slope of the terrain, and set the allowable range of slope difference; when the slope comparison exceeds the threshold, mark it; Based on the vector layer and attribute data of the geographic information system, the terrain data is classified, and the degree of fit between the route and the terrain is calculated in combination with the functional positioning and design speed of the expressway, and the relevant sections whose fit does not meet the threshold are marked; The marked abnormal road sections are integrated to form an abnormal road section dataset.

6. The engineering project data management method based on data analysis according to claim 1, characterized in that: According to step S300, the elevation difference threshold is set and slope difference threshold , for each abnormal road section i, count the number of points exceeding the elevation difference threshold , the number of points exceeding the slope difference threshold , and points will be deducted if the route direction does not match the terrain well. ; Abnormal severity The calculation formula is as follows: ； in, is the total number of points participating in the comparison of road section i, is the first weight coefficient, reflecting the contribution of elevation difference to the severity of the anomaly. is the second weight coefficient, which reflects the contribution of the slope difference to the severity of the anomaly. is the third weight coefficient, reflecting the contribution of the fit between the line direction and the terrain to the severity of the anomaly; Using the mileage data of route design, the length of abnormal section i By ending mileage station Subtract the starting mileage station It is obtained, expressed as , in kilometers; extract terrain slope, undulation and terrain fragmentation based on geographic information system; assume the average slope in the road section is The fluctuation is , terrain fragmentation index is , terrain complexity coefficient The calculation formula is: ； in, is the weight of the average slope, is the weight of the fluctuation, The weight of the terrain fragmentation index is set according to the impact of the terrain on the construction difficulty; The Apriori algorithm scans the data features of abnormal road sections, counts the frequency of occurrence of each single item, compares it with the preset minimum support threshold min_sup, and selects frequent item sets; it performs self-connection and pruning operations on frequent item sets to iteratively generate higher-order frequent item sets; Set the abnormal road segment dataset features as item sets , set the construction log structured data as item sets ,Will and Integrate into transaction T; run the Apriori algorithm to calculate the support of the mined rules , the calculation formula is: ； in, is the number of transactions that simultaneously meet the characteristics of the abnormal road section dataset and the construction log, is the total number of transactions; confidence for: ； in, is the number of transactions that meet the characteristics of the abnormal road section dataset, and the confidence reflects the reliability of the rule; based on the support and confidence of the abnormal road section dataset characteristics and the construction log, the association pattern of the abnormal road section dataset characteristics and the construction log is obtained; Run the Apriori algorithm to calculate the support and confidence of the abnormal road section dataset characteristics and the construction material procurement list; based on the support and confidence of the abnormal road section dataset characteristics and the construction material procurement list, obtain the association pattern of the abnormal road section dataset characteristics and the construction material procurement list.

7. The method for engineering project data management based on data analysis according to claim 6, characterized in that: According to step S300, based on the data characteristics of the abnormal road section, the K-Means clustering algorithm is used to determine the K value for the association pattern of the construction log and the construction material purchase list, and K cluster centers are randomly initialized. The distance from each association pattern data point to the cluster center is calculated, and the data point is assigned to the cluster where the nearest cluster center is located according to the distance. After allocation, the new centers of each cluster are recalculated, and this process is iterated repeatedly until the cluster centers no longer change. Each cluster is a combination of abnormal road section problems.

8. The engineering project data management method based on data analysis according to claim 1 is characterized by: According to step S400, the sensor data and the abnormal road section problem combination are fused in the time and space dimension, and the real-time sensor data of the abnormal road section is fused with the road section problem combination according to the mileage and geographic coordinates of the road section; the timestamp is unified in time, and the collection time of the sensor data is aligned with the frequency time series of the problem occurrence in the abnormal road section problem combination; Obtaining fusion data of abnormal road sections to be adjusted; The generative adversarial network algorithm consists of a generator and a discriminator. The generator inputs the fused data of the abnormal section to be adjusted to generate a new route design adjustment plan, and outputs the plane coordinate string, longitudinal elevation sequence, and curve and vertical curve parameters of the adjusted route to form several candidate route design adjustment plans; the training set and the validation set are divided to collect the effective adjustment cases of past highway engineering projects that have been verified in practice, including the corresponding abnormal section fusion data and route design adjustment plans, for model training; The fused data of the abnormal road sections to be adjusted are input into the trained model, and the generator outputs several candidate route design adjustment plans.

9. The engineering project data management method based on data analysis according to claim 8 is characterized in that: According to step S400, the candidate route design adjustment plans are placed on the real terrain one by one, and the degree of fit between the candidate route and the terrain is calculated; whether the route direction conforms to the contour trend is analyzed, and the slope analysis function of the geographic information system is used to compare the designed longitudinal slope with the actual terrain slope, and the proportion of the length of the section exceeding the normal slope difference is counted, and the higher the proportion, the lower the degree of fit between the candidate route and the terrain; Use the spatial analysis of geographic information systems combined with engineering measurement algorithms to estimate the construction volume of each plan; The optimal route design adjustment plan is selected based on the degree of fit between the candidate routes and the terrain and the construction volume of the plan.

10. A data analysis-based engineering project data management system, using a data analysis-based engineering project data management method according to any one of claims 1 to 9, characterized in that: include: Data collection and preprocessing module: including: data collection unit and data preprocessing unit, wherein the data collection unit obtains project data in highway projects, including design drawings, construction logs, sensor data, construction material purchase lists and geographic information system data, and summarizes them through interface programs; the data preprocessing unit screens the project data for duplicate, erroneous or missing data, standardizes and converts data in different formats, and unifies the timestamp format; Terrain and design data processing module: including: route design data extraction unit, terrain and geomorphology data extraction unit, spatial matching unit and abnormal road section marking unit; wherein, the route design data extraction unit extracts the route design data of the expressway from the design drawings through AutoCAD Civil 3D based on the project data, the terrain and geomorphology data extraction unit extracts the terrain and geomorphology data based on the geographic information system data, and the spatial matching unit uses the coordinate conversion method to spatially match the route design data and the terrain and geomorphology data; the abnormal road section marking unit compares the elevation and slope of each point of the route design with the terrain and geomorphology at the corresponding position point by point, analyzes the compatibility of the terrain type and the route direction, marks the abnormal road section, and forms the abnormal road section data set; Abnormal road section data mining module: including: feature extraction unit, association rule mining unit and cluster analysis unit; the feature extraction unit extracts abnormal road section data features from the abnormal road section data set, including the severity of the abnormality, the length of the road section and the terrain complexity coefficient; the association rule mining unit uses the Apriori algorithm to obtain the association patterns of the abnormal road section data features for the construction log and the construction material purchase list; the cluster analysis unit uses the K-Means clustering algorithm to cluster the association patterns into different categories according to similar features to form an abnormal road section problem combination; Route design adjustment module: includes: data fusion unit, generative adversarial network modeling and solution generation unit and solution simulation evaluation and optimization unit; among them, the data fusion unit fuses the sensor data with the abnormal road section problem combination to obtain the fused data of the abnormal road section to be adjusted; the generative adversarial network modeling and solution generation unit selects the generative adversarial network algorithm architecture to train and optimize the artificial intelligence model, inputs the fused data of the abnormal road section to be adjusted into the trained artificial intelligence model, and the generator outputs several candidate route design adjustment solutions; the solution simulation evaluation and optimization unit simulates different candidate route design adjustment solutions through the geographic information system, generates evaluation results, and selects the optimal route design adjustment solution.

Images