Road disease classification management method, system and device based on big data platform

Abstract

The application provides a road disease classification management method, system and equipment based on a big data platform, and belongs to the technical field of road disease identification. The method acquires various multi-source road detection information from various in-transit inspection devices; generates disease statistical information corresponding to various road disease inspection routes according to the multi-source road detection information and a preset disease identification model; the disease statistical information at least includes a disease statistical table and a disease distribution graph; after receiving a preset multi-dimensional classification standard from a user terminal, the disease statistical information is classified to determine a corresponding disease classification result; wherein the preset multi-dimensional classification standard is obtained based on one or more dimensions of a disease type, a disease level, an influence range and a repair difficulty; the disease classification result is sent to a preset expert system to determine a corresponding disease maintenance scheme and stored to the big data platform, so as to send the disease maintenance scheme and corresponding maintenance resource allocation information to the user terminal.

Description

Technical Field

[0001] This application relates to the field of road defect identification technology, and in particular to a road defect classification and management method, system and equipment based on a big data platform. Background Technology

[0002] With the acceleration of urbanization and the continuous increase in traffic flow, the frequency and types of road defects are also increasing. Traditional road defect management methods mainly rely on manual inspections and simple data recording, which suffer from problems such as low efficiency, inaccurate information, and unscientific classification, making it difficult to meet the needs of modern road maintenance and management.

[0003] The development of big data technology has provided new ideas and methods for road defect management. By collecting, organizing, and analyzing large amounts of road defect data, it is possible to achieve scientific classification, accurate diagnosis, and efficient management of road defects to a certain extent. However, current road defect classification and management solutions based on big data platforms are still immature. Existing solutions have weak capabilities in processing multi-source road detection information, their classification rules are relatively rigid and cannot be flexibly adjusted dynamically, and intelligent decision-making relies on basic classification results, resulting in insufficient intelligence. Furthermore, a significant investment of manpower is still required, impacting the user experience. Summary of the Invention

[0004] To address the aforementioned issues, this application provides a road defect classification management method, system, and equipment based on a big data platform, which improves the flexibility and intelligence of road defect classification management, thereby enabling efficient road maintenance.

[0005] On the one hand, embodiments of this application provide a road defect classification and management method based on a big data platform, the method comprising:

[0006] Acquire multi-source road inspection information from various on-the-road inspection devices;

[0007] Based on the multi-source road detection information and the preset disease identification model, disease statistics information corresponding to each road disease inspection route is generated; the disease statistics information includes at least a disease statistics table and a disease distribution map;

[0008] After receiving the preset multi-dimensional classification criteria from the user terminal, the disease statistics information is classified to determine the corresponding disease classification result; wherein, the preset multi-dimensional classification criteria are obtained based on one or more dimensions of disease type, disease level, scope of impact and repair difficulty;

[0009] The disease classification results are sent to a preset expert system to determine the corresponding disease maintenance plan and store it in a big data platform so that the disease maintenance plan and its corresponding maintenance resource allocation information can be sent to the user terminal.

[0010] In one implementation of this application, before acquiring multi-source road detection information from various on-the-road inspection devices, the method further includes:

[0011] Determine the inspection release task corresponding to the on-the-road inspection equipment; wherein, the inspection release task includes at least the inspection task execution time and the road defect inspection route;

[0012] Based on the inspection task and historical inspection records, the data collection interval time corresponding to each section of the road defect inspection route is determined; wherein, the data collection interval time is directly proportional to the inspection interval duration.

[0013] The data acquisition interval is sent to the corresponding on-route inspection equipment so that the on-route inspection equipment controls the image acquisition equipment and ground penetrating radar to acquire the multi-source road detection information according to the data acquisition interval.

[0014] In one implementation of this application, based on the multi-source road detection information and the preset defect identification model, defect statistical information corresponding to each road defect inspection route is generated, specifically including:

[0015] The multi-source road detection information is input into the preset disease identification model, so that point cloud feature data is extracted by the point cloud processing module, image data is extracted by the image processing module, and the point cloud feature data and the image data are fused by the fusion layer to obtain the joint feature vector corresponding to the multi-source road detection information.

[0016] The joint feature vector is mapped to a preset dimension output space by the preset disease identification model to determine the corresponding disease type and disease level, so as to add the disease type and disease level to the disease statistics information.

[0017] In one implementation of this application, based on the multi-source road detection information and the preset defect identification model, defect statistical information corresponding to each road defect inspection route is generated, specifically including:

[0018] Statistical analysis of each of the disease types and their corresponding disease type distribution information; the disease type distribution information includes at least the number of first diseases corresponding to each of the disease types and the proportion of first diseases among each of the disease types;

[0019] Determine the location distribution information and disease level distribution information of each road defect; the disease level distribution information includes at least the second disease data corresponding to each disease level and the proportion of second diseases among each disease level.

[0020] Add the disease type distribution information, the location distribution information, and the disease level distribution information to the disease statistics table;

[0021] Based on the disease statistics table, a corresponding disease heat map, location distribution map, and disease trend map are generated, which serve as the disease distribution map. Based on the disease statistics table and the disease distribution map, the disease statistics information is determined and displayed on the user terminal.

[0022] In one implementation of this application, before classifying the disease statistics, the method further includes:

[0023] Based on the disease statistics and historical disease repair records, determine one or more combinations of classification dimensions;

[0024] The various classification dimension combinations are sent to the user terminal so that the user can select the classification dimension combination as the preset multi-dimensional classification standard through the user terminal.

[0025] In one implementation of this application, the disease statistics are classified to determine the corresponding disease classification results, specifically including:

[0026] According to the preset multi-dimensional classification standard, each road disease in the disease statistics information is assigned to the corresponding classification category;

[0027] Determine the location distribution information of each road defect within the same classification category;

[0028] Based on the location distribution information, the corresponding historical traffic data, and the preset conflict judgment model, it is determined whether there is a conflict between each of the road defects. If so, a conflict defect pair is generated. The preset conflict judgment model is established based on the distance of the road defect and the impact value of the defect on traffic flow.

[0029] Based on the conflicting disease pairs, the conflicting road diseases are divided into different sets of diseases to be maintained; wherein, there are no conflicts among the road diseases in the same set of diseases to be maintained.

[0030] The number of road defects in each set of defects to be maintained is determined, maintenance priorities are generated based on the number of road defects, and a sequence of defect sets corresponding to each set of defects to be maintained is generated according to the maintenance priorities, so as to obtain the defect classification result.

[0031] In one implementation of this application, after determining the number of road defects in each set of defects to be maintained, the method further includes:

[0032] The set of road defects whose number is less than a preset threshold is identified as the set to be merged.

[0033] Based on each of the road defects in the set to be merged, determine the set of non-conflicting defects to be maintained in other classification categories, and merge the set to be merged with the set of non-conflicting defects to be maintained to generate the updated set of defects to be maintained.

[0034] In one implementation of this application, the disease classification results are sent to a preset expert system to determine the corresponding disease maintenance plan and stored in a big data platform, specifically including:

[0035] The preset expert system queries the pre-built knowledge graph for matching nodes and edges corresponding to the disease classification results; wherein, the pre-built knowledge graph uses disease type, disease cause, maintenance measures, maintenance materials and road attributes as nodes, and the relationships between road disease entities as edges;

[0036] Based on the matching nodes, matching edges, and preset expert experience triples, reasoning is performed to obtain the disease maintenance plan corresponding to the disease classification result, and the association between the disease maintenance plan and the disease classification result is sent to the big data platform.

[0037] On the other hand, embodiments of this application also provide a road defect classification and management system based on a big data platform, the system comprising:

[0038] The acquisition module is used to acquire multi-source road detection information from various on-the-road inspection devices;

[0039] The generation module is used to generate statistical information on road defects corresponding to each road defect inspection route based on the multi-source road detection information and the preset defect identification model; the statistical information on defects includes at least a defect statistics table and a defect distribution map;

[0040] The classification module is used to classify the disease statistics information after receiving a preset multi-dimensional classification standard from the user terminal, so as to determine the corresponding disease classification result; wherein, the preset multi-dimensional classification standard is obtained based on one or more dimensions of disease type, disease level, scope of impact and repair difficulty;

[0041] The sending module is used to send the disease classification results to a preset expert system to determine the corresponding disease maintenance plan and store it in a big data platform so that the disease maintenance plan and its corresponding maintenance resource allocation information can be sent to the user terminal.

[0042] Furthermore, embodiments of this application also provide a road defect classification and management device based on a big data platform, the device comprising:

[0043] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a road defect classification and management method based on a big data platform as described above.

[0044] Compared with the prior art, the significant advantages of this application are as follows:

[0045] The above technical solution enables multi-source data collection and comprehensive road damage statistics. By combining user-personalized multi-dimensional classification standards, road damage can be flexibly classified, and then an expert system can be used to generate maintenance plans adapted to these personalized classifications. This process requires minimal human intervention for classification and intelligent decision-making, effectively improving the flexibility and intelligence of road damage classification management and enabling efficient road maintenance. Attached Figure Description

[0046] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0047] Figure 1 This is a flowchart illustrating a road defect classification and management method based on a big data platform, as described in this application.

[0048] Figure 2 This is a schematic diagram of the structure of a road disease classification and management system based on a big data platform in an embodiment of this application;

[0049] Figure 3 This is a schematic diagram of the structure of a road disease classification and management device based on a big data platform, as described in an embodiment of this application. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0051] This application provides a road defect classification management method, system, and device based on a big data platform to solve the technical problem that the current road defect classification management is not flexible and intelligent enough, which affects the user experience. This improves the flexibility and intelligence of road defect classification management, thereby enabling efficient road maintenance.

[0052] The various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0053] This application provides a method for road defect classification and management based on a big data platform, such as... Figure 1 As shown, the method may include steps S101-S104:

[0054] S101, the server obtains multi-source road detection information from various on-the-road inspection devices.

[0055] It should be noted that the server, as the executing entity of the road disease classification and management method based on the big data platform, is only an example. The executing entity is not limited to the server; for example, it can also be an edge computing device, etc. This application does not make any specific limitations in this regard.

[0056] In-transit inspection equipment includes, but is not limited to, inspection vehicles and drones. The in-transit inspection equipment is equipped with image acquisition devices such as high-definition industrial cameras, GPS / BeiDou high-precision positioning modules, and 4G / 5G communication modules. Ground-penetrating radar can also be installed on inspection vehicles. Data from inspection vehicles and drones can be integrated on a server as multi-source road detection information.

[0057] High-definition industrial cameras, mounted on inspection vehicles or drones, are used to capture high-resolution images of road surfaces. These images can identify defects such as cracks, potholes, and subsidence. The cameras typically have a resolution of over 10 megapixels, enabling them to capture minute features of these defects. The cameras acquire images of the road surface through timed or triggered shooting. These images are then transmitted in real-time to a big data platform via 4G / 5G communication modules for subsequent processing and analysis.

[0058] Ground-penetrating radar (GPR) is installed on inspection vehicles to detect defects in the internal structure of roads, such as the condition of the base and surface layers. GPR acquires information about road defects by emitting high-frequency electromagnetic waves and detecting reflected signals from the underground structure. As the vehicle moves, the GPR continuously emits and receives electromagnetic waves, generating cross-sectional images of the underground structure. These images are then transmitted in real-time to a big data platform via a 4G / 5G communication module for further processing and analysis.

[0059] GPS / BeiDou high-precision positioning modules are installed on inspection vehicles or drones to provide precise location information of defects. The positioning module acquires the latitude, longitude, and elevation information of the vehicle or drone in real time via satellite signals. During data acquisition, the positioning module simultaneously records the precise location information of each defect point. This location information is transmitted in real time to a big data platform via a 4G / 5G communication module and correlated with defect images and ground-penetrating radar data. The 4G / 5G communication module, also installed on the inspection vehicle or drone, transmits the collected defect data to the big data platform in real time. The communication module supports high-speed data transmission, ensuring data real-time performance and integrity. During data acquisition, the communication module transmits data collected by high-definition industrial cameras, ground-penetrating radar, and positioning modules to the big data platform in real time. During data transmission, the communication module ensures data stability and reliability, supporting rapid transmission of large-scale data.

[0060] The data storage module of the road defect classification management system based on a big data platform, as described in this application, consists of a database management system, a storage engine, a distributed storage management module, and a data encryption and security module. The database management system can be a platform such as Transwarp Data Hub (TDH) from Transwarp Technology. The storage engine can utilize NAND Flash to construct a storage array. NAND Flash offers high storage density and fast write speeds, making it suitable for large-capacity data storage. Simultaneously, to ensure data storage integrity, error detection / correction (EDC / ECC) algorithms need to be incorporated. Distributed storage management can employ Transwarp Distributed Data Management System (TDDMS) from Transwarp Technology, which provides unified distributed storage management, supports multiple storage engines, and enables efficient data storage and management. The data encryption and security module can utilize xSTORAGE encryption technology to provide robust protection for the data, ensuring data security during storage and transmission.

[0061] In this embodiment of the application, before obtaining the multi-source road detection information from each on-the-road inspection device, the method further includes:

[0062] Determine the inspection deployment tasks corresponding to the on-route inspection equipment. These tasks must include at least the inspection execution time and the road defect inspection route. Based on the deployed tasks and historical inspection records, determine the data acquisition interval for each segment of the road defect inspection route. The data acquisition interval is directly proportional to the inspection interval duration. Send the data acquisition interval to the corresponding on-route inspection equipment so that the on-route inspection equipment can control the image acquisition equipment and ground-penetrating radar to collect multi-source road detection information according to the data acquisition interval.

[0063] Before an inspection is performed, the server can issue an inspection deployment task, or the inspection device itself can proactively issue such a task; this application does not impose specific limitations on this. After the server determines the inspection deployment task for the inspection device, it can flexibly adjust and control the working time of the data acquisition-related equipment of the inspection device based on historical inspection records associated with the deployment task. This allows for the effective collection of multi-source road detection information while saving power consumption costs.

[0064] Specifically, the server extracts the inspection task execution time and road defect inspection route from the inspection task release. It then matches the historical inspection times for each road location along that route with historical inspection records. Based on the current task's execution time, the server calculates the inspection interval between the historical inspection time and the current task execution time. The data acquisition interval is calculated using the formula: t1 = k·t0, where t1 is the data acquisition interval, k is a preset coefficient within the interval (0, 1), and t0 is the inspection interval duration. The server can send the data acquisition interval and the corresponding road segment to the on-the-road inspection equipment to control the equipment to intermittently collect multi-source road detection information at the specified intervals. For example, the image acquisition equipment and ground-penetrating radar can be controlled to collect data at 500 milliseconds (ms). The preset coefficient can be set by the user according to the actual usage scenario; this application does not impose specific limitations on this.

[0065] S102, the server generates statistical information on road defects corresponding to each road defect inspection route based on multi-source road detection information and preset defect identification models.

[0066] The above-mentioned disease statistics information includes at least a disease statistics table and a disease distribution map.

[0067] This application uses a pre-defined defect identification model to process multi-source road detection information, thereby generating defect statistics that can be displayed to users or used for further processing. Specifically, after obtaining the multi-source road detection information, this application prioritizes data preprocessing, such as data cleaning and data filtering.

[0068] Data cleaning: Identifying and correcting errors and outliers in the data to ensure accuracy and consistency. This includes missing value handling: detecting missing values ​​and processing them using methods such as imputation, deletion, or interpolation, depending on the specific situation. Outlier handling: Identifying outliers in the data and correcting or deleting them according to business rules. Duplicate value handling: Detecting and deleting duplicate data records to ensure data uniqueness. Data filtering: Removing invalid and noisy data to improve data purity. This includes image denoising: Denoising acquired image data to improve image clarity and quality. Data screening: Filtering out valid data and removing irrelevant and noisy data according to preset rules.

[0069] In this embodiment of the application, the above-mentioned generation of disease statistical information corresponding to each road disease inspection route based on multi-source road detection information and a preset disease identification model specifically includes:

[0070] Multi-source road detection information is input into a preset road defect identification model. Point cloud feature data is extracted using a point cloud processing module, and image data is extracted using an image processing module. A fusion layer then fuses the point cloud feature data and image data to obtain a joint feature vector corresponding to the multi-source road detection information. The preset road defect identification model maps this joint feature vector to a preset-dimensional output space to determine the corresponding defect type and defect level, which are then added to the defect statistics.

[0071] In other words, the pre-defined disease identification model includes a point cloud processing module and an image processing module. These modules process the point cloud data and images contained in the multi-source road detection information, respectively, extracting point cloud feature data and image data and fusing them to obtain a joint feature vector. Specifically, the image processing module uses a YOLO submodule to divide the image into multiple grids. Each grid is responsible for detecting diseases within a specific area and outputting the disease category and location information. PointRend is used to accurately segment the disease area and extract detailed disease features. By combining coarse and fine segmentation, PointRend improves the accuracy and efficiency of segmentation.

[0072] The preset disease identification model is a machine learning model trained from several feature vector samples. After the joint feature vector is input into the preset disease identification model, the model processes the joint feature vector and maps it to an output space of a preset dimension, such as 10 dimensions. Then, it can output the disease type and disease level corresponding to the result. The disease level is used to characterize the severity of the disease.

[0073] In one embodiment of this application, the above-mentioned generation of disease statistics information corresponding to each road disease inspection route based on multi-source road detection information and a preset disease identification model specifically includes:

[0074] Collect statistics on the distribution information of each road disease type and its corresponding disease type. The disease type distribution information should at least include the number of primary diseases for each disease type and the proportion of primary diseases among different disease types. Determine the location distribution information and disease level distribution information for each road disease. The disease level distribution information should at least include the data on secondary diseases corresponding to each disease level and the proportion of secondary diseases among different disease levels. Add the disease type distribution information, location distribution information, and disease level distribution information to the disease statistics table. Based on the disease statistics table, generate corresponding disease heat maps, location distribution maps, and disease trend maps, which will serve as the disease distribution map. Based on the disease statistics table and the disease distribution map, determine the disease statistical information for display on the user terminal.

[0075] In other words, this application can classify and statistically analyze diseases based on their type, generating the quantity and proportion of each type of disease. Disease location statistics: Based on the location information of diseases, the distribution of diseases in different road sections is generated. Disease severity statistics: Based on the severity of diseases, statistics are classified and generated, generating the quantity and proportion of diseases of different severity levels. Based on the above-mentioned disease type distribution information, location distribution information, and disease severity distribution information, a disease statistics table is created that can display the disease type, quantity, location, and severity level. This disease statistics table is formatted like an Excel spreadsheet.

[0076] Simultaneously, a disease distribution map will be generated to visually display the distribution of diseases on roads, and a disease heat map will be generated, using color intensity to indicate disease density and severity. A disease location distribution map will be generated, displaying the specific location and extent of diseases through points, lines, and areas. A disease trend map will be generated, showing the changing trends of diseases over time, providing a basis for disease prediction and prevention. User terminals can be users corresponding to the big data platform, such as mobile phones, computers, and other devices used by road maintenance and management personnel; this application does not impose specific limitations on this.

[0077] S103. After receiving the preset multi-dimensional classification criteria from the user terminal, the server classifies the disease statistics information to determine the corresponding disease classification result.

[0078] The preset multi-dimensional classification standard is based on one or more dimensions, including disease type, disease level, scope of impact, and difficulty of repair.

[0079] In this embodiment of the application, before classifying diseases based on statistical information, the method further includes:

[0080] Based on disease statistics and historical disease repair records, one or more combinations of classification dimensions are determined. These combinations are then sent to the user terminal, allowing the user to select the appropriate combination as a preset multi-dimensional classification standard.

[0081] Specifically, after determining the statistical information of road damage, the server can combine it with historical damage repair records to obtain one or more combinations of classification dimensions. These historical damage repair records include pre-stored or user-defined repair-related records, which contain the patterns of classification dimension combinations selected for damage repair during historical repair processes. For example, in several historical damage repair processes, road damage may be repaired sequentially using a binary classification dimension combination of damage type and damage level, or sequentially using a triple classification dimension combination of damage type, damage level, and repair difficulty. In this case, the classification dimension combinations would include [damage type, damage level] and [damage type, damage level, repair difficulty]. Among them, disease type, disease level, scope of impact, and repair difficulty can also be used as preset multi-dimensional classification standards. Here, disease type, disease level, scope of impact, and repair difficulty are only stored as examples. The classification dimensions can also include other dimensions. Depending on the specific use case, users can customize and add them. Alternatively, new classification dimensions for road diseases can be obtained by continuously analyzing road diseases and then supplemented. The classification dimensions can also be optimized by combining data analysis and machine learning algorithms to further improve the accuracy and efficiency of classification. This application does not make specific limitations in this regard.

[0082] This application can send one or more classification dimension combinations applicable to the above-mentioned disease statistical information to the user terminal, and the user can operate the user terminal to select one of the classification dimension combinations as a preset multi-dimensional classification standard, thereby classifying the disease statistical information.

[0083] The above-mentioned types of road defects are categorized as follows: Based on their appearance and causes, defects are classified into cracks (such as longitudinal cracks, transverse cracks, and network cracks), potholes, settlement, and loosening. Defect severity is categorized into three levels: minor, moderate, and severe, based on the degree of severity. Impact range is categorized into localized and widespread impacts based on the extent to which the defect affects the road's functionality. Repair difficulty is categorized into three levels: easy to repair, moderately difficult, and highly difficult, based on the complexity of repair and the resources required.

[0084] In this embodiment of the application, the above-mentioned classification of disease statistical information to determine the corresponding disease classification results specifically includes:

[0085] Based on a pre-defined multi-dimensional classification standard, road defects in the defect statistics are assigned to corresponding classification categories. The location distribution information of each road defect within the same category is determined. Based on the location distribution information, corresponding historical traffic data, and a pre-defined conflict judgment model, it is determined whether conflicts exist among road defects; if so, conflicting defect pairs are generated. The pre-defined conflict judgment model is established based on the distance of road defects and their impact on traffic flow. Based on each conflicting defect pair, the conflicting road defects are assigned to different sets of defects requiring maintenance. Road defects within the same set of defects requiring maintenance do not conflict. The number of road defects in each set of defects requiring maintenance is determined to generate maintenance priorities. Based on the maintenance priorities, a sequence of defect sets corresponding to each set of defects requiring maintenance is generated, resulting in the defect classification results.

[0086] In other words, the server uses a preset multi-dimensional classification standard to first classify each road defect in the defect statistics. For example, if the preset multi-dimensional classification standard is defect type and defect level, then {cracks, minor}, {cracks, moderate}, {cracks, severe}, {potholes, minor}, etc., will be classified one by one to obtain the classification category of each road defect. Subsequently, the location distribution information will be extracted separately for each classification category, and combined with historical traffic data, a preset conflict judgment model will be run to verify whether road maintenance will cause conflicts affecting traffic. The preset conflict judgment model is as follows:

[0087]

[0088] Among them, w ij The result of the conflict judgment between road defect i and road defect j is 1, indicating that a conflict exists, and 0 indicates that no conflict exists; d ij d represents the distance between road defect i and road defect j. th F is the preset distance threshold. i F represents the traffic flow impact value of road defects i during historical periods of concentrated traffic. j F represents the traffic flow impact value of road defects i during historical periods of concentrated traffic. max This is the maximum traffic flow carrying capacity threshold for the road segment between road defect i and road defect j.

[0089] The aforementioned distance threshold and maximum traffic flow capacity threshold can be preset by the user; this application does not impose specific limitations on them. i and F j According to The calculation yields t. n t represents the average travel time of the location of road defects extracted from historical traffic data during peak traffic periods under normal conditions.d Q represents the average travel time during peak traffic periods in the history of the location of road defects. n This represents the traffic flow during peak periods of historical traffic congestion under normal circumstances.

[0090] Based on the above processing, multiple conflicting defect pairs can be obtained, i.e., binary pairs consisting of two road defects. Then, road defects of the same classification category are grouped, and road defects that do not conflict with each other are assigned to a set of defects requiring maintenance. Next, maintenance priorities are assigned to each set of defects requiring maintenance according to the number of road defects in the set; for example, the more road defects in a set, the higher the maintenance priority. According to the maintenance priority, the name of each set of defects requiring maintenance is added to the defect set sequence so that subsequent maintenance processing can be carried out according to this defect set sequence.

[0091] Furthermore, in one embodiment of this application, after determining the number of road defects in each set of defects to be maintained, the method further includes:

[0092] The set of road defects requiring maintenance whose number is less than a preset threshold is identified as the set to be merged. Based on each road defect in the set to be merged, sets of non-conflicting road defects requiring maintenance in other categories are identified. The set to be merged is then merged with the sets of non-conflicting road defects to generate an updated set of road defects requiring maintenance.

[0093] In other words, if the number of road defects in a set of defects to be maintained is relatively small, less than a pre-set threshold, then this set of defects to be maintained can be merged with other sets of defects to be maintained for centralized processing later. Specifically, the server performs the aforementioned conflict determination on each road defect in the set to be merged with road defects in other sets of defects to be maintained of different categories. If there are no conflicting road defects between the set to be merged and a set of defects to be maintained, then that set of defects to be maintained is considered a non-conflicting set of defects to be maintained, and the road defect merging process is performed.

[0094] The above steps are performed by the system's disease classification module, whose workflow is as follows:

[0095] Data Input: Receives preprocessed and feature-extracted road defect data from the data processing module. Feature Matching: Matches defect data to preset classification criteria based on defect characteristics (such as type, severity, and extent of impact). Classification Execution: Assigns defect data to appropriate classification categories according to the matched criteria. Result Output: Generates defect classification results and stores them in the big data platform for use by subsequent defect diagnosis and management decision-making modules.

[0096] S104, the server sends the disease classification results to the preset expert system to determine the corresponding disease maintenance plan and store it in the big data platform so that the disease maintenance plan and its corresponding maintenance resource allocation information can be sent to the user terminal.

[0097] In this embodiment of the application, the above-mentioned sending the disease classification results to a preset expert system to determine the corresponding disease maintenance plan and storing it in a big data platform specifically includes:

[0098] Through a pre-built expert system, matching nodes and edges corresponding to the road disease classification results are queried from a pre-constructed knowledge graph. The pre-constructed knowledge graph uses disease type, disease cause, maintenance measures, maintenance materials, and road attributes as nodes, and the relationships between road disease entities as edges. Based on the matching results, matching edges, and pre-built expert experience triples, reasoning is performed to obtain the corresponding disease maintenance plan, and the association between the disease maintenance plan and the disease classification results is sent to the big data platform.

[0099] In other words, this application utilizes an expert system for disease diagnosis. This function is the system's disease diagnosis module, which uses the expert system and knowledge base to accurately diagnose diseases, determine their specific causes, and identify solutions. This module improves the accuracy and efficiency of disease diagnosis through an automated and intelligent diagnostic process, providing scientific decision support for road maintenance management departments. The diagnostic basis of the disease diagnosis module includes the following aspects: Disease classification results: Classification information such as disease type, severity, scope of impact, and repair difficulty obtained from the disease classification module. Relevant data: Including detailed images, location information, and historical data of the disease. Expert system and knowledge base: Utilizing a pre-set expert system and knowledge base, combined with the disease classification results and relevant data, a comprehensive analysis and diagnosis are performed.

[0100] The workflow of the disease diagnosis module is as follows: Data Input: Receives disease classification results and related data from the disease classification module. Feature Matching: Matches pre-defined diagnostic rules and expert experience from the knowledge base based on disease feature information. Comprehensive Analysis: Performs comprehensive analysis by combining disease classification results and related data to determine the specific cause of the disease. Solution Generation: Generates detailed repair suggestions and maintenance plans based on the specific cause of the disease. Result Output: Generates a detailed disease diagnosis report, including disease causes and repair suggestions, and stores the results in the big data platform for use by the subsequent management decision-making module.

[0101] The pre-built expert system has a pre-constructed knowledge graph that clearly defines various entities related to road defects, such as defect types (cracks, potholes, etc.), causes of defects (foundation settlement, vehicle overloading, etc.), maintenance measures (crack sealing, repairs, etc.), maintenance materials (asphalt, cement, etc.), and road attributes (road grade, traffic volume, etc.), which serve as the knowledge graph interface. Relationships between entities are defined, such as "cause" (defect causes lead to defect types), "applicable" (maintenance measures are applicable to defect types), and "use" (maintenance measures use maintenance materials), which serve as edges in the knowledge graph. This knowledge graph allows for matching nodes and edges to various defects in the defect classification results. Then, pre-set expert experience triples are used for reasoning. These expert experience triples can be pre-defined by experts to derive maintenance solutions corresponding to the defect classification results. These maintenance solutions can be stored on a big data platform so that users can maintain road defects according to these solutions.

[0102] Meanwhile, the big data platform can also generate repair plans, maintenance budgets, and resource allocation schemes based on the classification and diagnosis results of road defects through its management decision-making module, providing scientific decision support for road maintenance management departments. This module, through a "one-screen" data display, allows for quick and accurate understanding of system operations, defects, vehicle maintenance and inspection information, and provides early warnings of road structural risks.

[0103] The disease diagnosis module's diagnostic criteria include the following aspects: Disease classification results: Classification information obtained from the disease classification module, such as disease type, severity, affected area, and repair difficulty. Relevant data: Including detailed images, location information, and historical data of the disease. Expert system and knowledge base: Utilizing a pre-set expert system and knowledge base, combined with the disease classification results and relevant data, a comprehensive analysis and diagnosis are performed.

[0104] The workflow of the disease diagnosis module is as follows: Data Input: Receives disease classification results and related data from the disease classification module. Feature Matching: Matches pre-defined diagnostic rules and expert experience from the knowledge base based on disease feature information. Comprehensive Analysis: Performs comprehensive analysis by combining disease classification results and related data to determine the specific cause of the disease. Solution Generation: Generates detailed repair suggestions and maintenance plans based on the specific cause of the disease. Result Output: Generates a detailed disease diagnosis report, including disease causes and repair suggestions, and stores the results in the big data platform for use by the subsequent management decision-making module.

[0105] The management decision-making module provides decision support, specifically including the following aspects:

[0106] Disease Repair Plan: Based on the classification and diagnosis of diseases, a detailed disease repair plan is generated, including repair time, repair methods, required materials and equipment, etc. Disease repair tasks are assigned to the corresponding maintenance teams, and a repair schedule is set. According to the needs of the repair tasks, the necessary materials and equipment are allocated to ensure the smooth progress of the repair work.

[0107] Maintenance Budget: Generate a maintenance budget, including estimates for various expenses such as damage repair, routine maintenance, and preventative maintenance. Based on the damage repair plan, estimate the cost of each repair task, including material costs, labor costs, and equipment usage fees. Integrate all costs to prepare a detailed maintenance budget and optimize costs.

[0108] Resource Allocation Plan: Based on the disease repair plan and maintenance budget, develop a resource allocation plan to ensure the rational use of resources. Assess existing resources, including personnel, equipment, and materials. Based on the disease repair plan, develop a resource allocation plan to ensure the rational allocation of resources in time and space.

[0109] The digital cockpit is an important component of the management decision-making module, providing intuitive decision support information through "one-screen" data display.

[0110] Data Display: A graphical interface displays system business, defects, vehicle maintenance and inspection information, providing comparative analysis of real-time and historical data. Defect Distribution Map: Shows the distribution of defects across different road sections, using color and icons to indicate the severity of defects. Defect Statistics Table: Displays information such as the type, quantity, and location of defects, providing multi-dimensional statistical analysis. Vehicle Information: Shows the location, status, and task execution status of inspection vehicles, providing real-time monitoring and dispatching functions.

[0111] Risk Warning: Real-time warnings are issued for road structure risks, reminding maintenance and management departments to take timely measures. The risk level of the road structure is assessed based on the severity and development trend of the defects. When high-risk defects are detected, timely warning information is issued to remind relevant departments to take appropriate measures.

[0112] The user interaction module is a crucial component of the road defect classification and management system based on a big data platform. It provides a user-friendly interface to facilitate data entry, querying, analysis, and decision-making for different user roles (such as road maintenance managers, inspectors, and equipment operators). Through its intuitive interface design and rich functionality, this module enhances user experience and improves work efficiency.

[0113] The user interface design aims to provide a simple, intuitive, and easy-to-use operating environment, ensuring that different user roles can quickly get started and efficiently complete tasks. This includes: a login interface for user authentication to ensure system security; user input fields for username and password; verification for successful user authentication (redirecting to the main interface on success, and displaying an error message on failure); a main interface providing entry points to the system's main functions and displaying key information; quick access to key functions such as data entry, query, analysis, and decision-making; a dashboard displaying key indicators and real-time data, such as the number of defects, repair progress, and vehicle status; a notification bar displaying system notifications and warnings; a data entry interface for easy user data entry, supporting multiple data types; a form design with clear and concise forms including fields for defect type, location, severity, and discovery time; image upload supporting high-resolution image uploads for recording detailed defect information; integrated GPS / BeiDou positioning for automatic defect location information; a submit button for submitting data, which the system then verifies and stores; and a data query interface providing powerful data query capabilities, supporting multi-dimensional queries. Query Criteria: Supports queries based on disease type, location, severity, and discovery time. Results Display: Displays query results in tables and maps, supporting pagination and export functions. Detailed Information: Click on a specific item in the query results to view detailed disease information and images. Data Analysis Interface: Provides disease data analysis functions, generating statistical reports and charts. Statistical Reports: Generates disease statistical tables, displaying information such as disease type, quantity, and location. Chart Display: Generates disease distribution maps, disease trend charts, etc., visually displaying the disease situation. Export Function: Supports exporting analysis results to Excel, PDF, and other formats. Decision Support Interface: Provides decision support information such as disease repair plans, maintenance budgets, and resource allocation schemes. Repair Plan: Displays disease repair plans, including repair time, methods, materials, and equipment. Maintenance Budget: Displays maintenance budgets, including detailed breakdowns of various costs. Resource Allocation: Displays resource allocation schemes, including the allocation of personnel, equipment, and materials.

[0114] The user interaction module supports a variety of features to improve user experience and operational efficiency.

[0115] Zoom In / Out Functionality: Supports zooming in and out of maps and images for easy viewing of details. Map Operations: Adjust the map's display scale using the mouse wheel or zoom in / out buttons. Image Operations: Adjust the image's display scale using the mouse wheel or zoom in / out buttons. Data Entry Interface: Provides a clean and concise data entry interface for quick data entry. Form Design: Clear form fields and input field prompts reduce user input errors. Auto-fill: Supports auto-fill for some fields, improving entry efficiency. Instant Validation: Instant validation of input fields ensures data accuracy. Convenient and Quick Operation: Offers shortcuts and intelligent prompts to reduce user steps. Shortcut Buttons: Provides shortcut buttons for frequently used functions, reducing menu navigation. Intelligent Prompts: Provides intelligent prompts for input fields and operation steps to help users complete tasks quickly. Batch Operations: Supports batch data entry, querying, and export to improve work efficiency.

[0116] The above technical solution enables multi-source data collection and comprehensive road damage statistics. By combining user-personalized multi-dimensional classification standards, road damage can be flexibly classified, and then an expert system can be used to generate maintenance plans adapted to these personalized classifications. This process requires minimal human intervention for classification and intelligent decision-making, effectively improving the flexibility and intelligence of road damage classification management and enabling efficient road maintenance.

[0117] Figure 2 A schematic diagram of a road defect classification and management system based on a big data platform is provided as an embodiment of this application, as shown below. Figure 2 As shown, the road defect classification and management system 200 based on a big data platform includes:

[0118] The acquisition module 201 acquires multi-source road inspection information from various on-route inspection devices. The generation module 202 generates statistical information on road defects corresponding to each road defect inspection route based on the multi-source road inspection information and a preset defect identification model. The statistical information includes at least a defect statistics table and a defect distribution map. The classification module 203, upon receiving a preset multi-dimensional classification standard from the user terminal, classifies the statistical information to determine the corresponding defect classification result. The preset multi-dimensional classification standard is based on one or more dimensions, including defect type, defect level, impact range, and repair difficulty. The sending module 204 sends the defect classification result to a preset expert system to determine the corresponding defect maintenance plan and stores it in a big data platform, so that the defect maintenance plan and its corresponding maintenance resource allocation information can be sent to the user terminal.

[0119] Figure 3 A schematic diagram of a road defect classification and management device based on a big data platform is provided as an embodiment of this application, as shown below. Figure 3 As shown, the device includes:

[0120] At least one processor; and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0121] Acquire multi-source road inspection information from various on-route inspection devices. Based on the multi-source road inspection information and a preset defect identification model, generate defect statistics corresponding to each road defect inspection route. The defect statistics include at least a defect statistics table and a defect distribution map. Upon receiving preset multi-dimensional classification criteria from the user terminal, classify the defect statistics to determine the corresponding defect classification results. The preset multi-dimensional classification criteria are based on one or more dimensions, including defect type, defect level, impact range, and repair difficulty. Send the defect classification results to a preset expert system to determine the corresponding defect maintenance plan and store it in the big data platform so that the defect maintenance plan and its corresponding maintenance resource allocation information can be sent to the user terminal.

[0122] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system and device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0123] The systems, devices, and methods provided in this application are one-to-one correspondences. Therefore, the systems and devices also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and devices will not be repeated here.

[0124] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0125] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1.A method for road disease classification management based on a big data platform, characterized in that, The method comprises: acquiring each multi-source road detection information from each in-transit inspection device; generating disease statistical information corresponding to each road disease inspection route according to the multi-source road detection information and a preset disease identification model; the disease statistical information at least includes a disease statistical table and a disease distribution map; after receiving a preset multi-dimensional classification standard from a user terminal, classifying the disease statistical information to determine the corresponding disease classification result; wherein the preset multi-dimensional classification standard is based on one or more dimensions of disease type, disease level, influence range and repair difficulty; sending the disease classification result to a preset expert system to determine the corresponding disease maintenance scheme and store it to a big data platform, so as to send the disease maintenance scheme and its corresponding maintenance resource allocation information to the user terminal; wherein, the disease classification result is determined by classifying the disease statistical information according to the preset multi-dimensional classification standard, which comprises: assigning each road disease in the disease statistical information to the corresponding classification category according to the preset multi-dimensional classification standard; determining the position distribution information of each road disease in the same classification category; determining whether each road disease has conflict according to the position distribution information, corresponding historical traffic data and a preset conflict judgment model, and if so, generating a conflict disease pair; the preset conflict judgment model is based on road disease distance and disease pair traffic flow influence value; according to each conflict disease pair, each road disease with conflict is divided into different maintenance disease sets; wherein each road disease in the same maintenance disease set has no conflict; determining the number of road diseases in each maintenance disease set to generate a maintenance priority according to the number of road diseases, and generating a disease set sequence corresponding to each maintenance disease set according to the maintenance priority, to obtain the disease classification result. 2.The road disease classification management method based on a big data platform according to claim 1, wherein, Before acquiring each multi-source road detection information from each in-transit inspection device, the method further comprises: determining the inspection release task corresponding to the in-transit inspection device; wherein the inspection release task at least includes an inspection task execution time and the road disease inspection route; determining the data collection interval time corresponding to each road section in the road disease inspection route according to the inspection release task and historical inspection records; wherein the data collection interval time and the inspection interval time are in a positive proportional relationship; sending the data collection interval time to the corresponding in-transit inspection device, so that the in-transit inspection device controls the image collection device and ground penetrating radar to collect the multi-source road detection information according to the data collection interval time. 3.The road disease classification management method based on a big data platform according to claim 2, characterized in that, According to the multi-source road detection information and the preset disease identification model, the disease statistical information corresponding to each road disease inspection route is generated, which comprises: Input the multi-source road detection information into the preset disease identification model to extract point cloud feature data through a point cloud processing module, extract image data through an image processing module, and fuse the point cloud feature data and the image data through a fusion layer to obtain a joint feature vector corresponding to the multi-source road detection information; Map the joint feature vector to an output space of a preset dimension through the preset disease identification model, determine a corresponding disease type and a disease level, and add the disease type and the disease level to the disease statistical information. 4.The road disease classification management method based on a big data platform according to claim 3, characterized in that, According to the multi-source road detection information and the preset disease identification model, generate disease statistical information corresponding to each road disease inspection route, specifically including: Statistical information of each disease type and its corresponding disease type distribution information; the disease type distribution information at least includes the first disease quantity corresponding to each disease type and the first disease proportion between each disease type; Determine the position distribution information and the disease level distribution information of each road disease; the disease level distribution information at least includes the second disease data corresponding to each disease level and the second disease proportion between each disease level; Add the disease type distribution information, the position distribution information and the disease level distribution information to the disease statistical table; According to the disease statistical table, generate a corresponding disease heat map, position distribution map and disease trend map, and use them as the disease distribution map, so as to determine the disease statistical information according to the disease statistical table and the disease distribution map, and display them on the user terminal. 5.The road disease classification management method based on a big data platform of claim 1, wherein, Before classifying the disease statistical information, the method further includes: According to the disease statistical information and the historical disease repair record, determine one or more classification dimension combinations; Send each classification dimension combination to the user terminal, so that the user selects a classification dimension combination as the preset multi-dimensional classification standard through the user terminal. 6.The road disease classification management method based on a big data platform of claim 1, wherein, After determining the number of road diseases in each of the to-be-maintained disease sets, the method further includes: Determine the to-be-maintained disease set with the number of road diseases less than a preset threshold as a to-be-merged set; Based on each road disease in the to-be-merged set, determine a non-conflict to-be-maintained disease set in other classification categories, and merge the to-be-merged set with the non-conflict to-be-maintained disease set to generate an updated to-be-maintained disease set. 7.The road disease classification management method based on a big data platform of claim 1, wherein, Send the disease classification result to a preset expert system to determine a corresponding disease maintenance scheme and store it to a big data platform, specifically including: Through the preset expert system, query each matching node and matching edge corresponding to the disease classification result from a pre-constructed knowledge graph; wherein, the pre-constructed knowledge graph takes disease type, disease cause, maintenance measure, maintenance material and road attribute as node, and takes the relationship between road disease entities as edge; According to reasoning of the matching nodes, the matching edges, and the preset expert experience triplets, the disease maintenance scheme corresponding to the disease classification result is obtained, and an association between the disease maintenance scheme and the disease classification result is sent to the big data platform. 8.A road disease classification management system based on a big data platform, characterized by, The system can perform the road disease classification management method based on the big data platform according to any one of claims 1-7; the system comprises: An acquisition module is configured to acquire the multi-source road detection information from each in-transit inspection device; A generation module is configured to generate disease statistical information corresponding to each road disease inspection route according to the multi-source road detection information and a preset disease identification model; the disease statistical information at least includes a disease statistical table and a disease distribution map; A classification module is configured to perform disease classification on the disease statistical information to determine a corresponding disease classification result after receiving a preset multi-dimensional classification standard from a user terminal; the preset multi-dimensional classification standard is obtained based on one or more dimensions of disease type, disease level, influence range, and repair difficulty; A sending module is configured to send the disease classification result to a preset expert system to determine a corresponding disease maintenance scheme and store the disease maintenance scheme to a big data platform, so as to send the disease maintenance scheme and corresponding maintenance resource allocation information to the user terminal. 9.A road disease classification management device based on a big data platform, characterized by, The device comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the road disease classification management method based on the big data platform according to any one of claims 1-7.

Images