Intelligent evaluation method and system for highway construction quality integrating internet of things analysis

Abstract

The application discloses a highway construction quality intelligent evaluation method and system integrated with internet of things analysis, relates to the technical field of quality evaluation, and comprises the following steps: deploying an internet of things equipment cluster, and creating a general template of a multi-device data source; collecting a highway construction multi-modal data set in real time, performing integrated preprocessing, and obtaining an available highway construction multi-modal data set; collecting a highway construction historical quality data set, performing correlation mining and evaluation training, building a highway construction quality evaluation channel library, performing road section channel matching and construction quality evaluation on the available highway construction multi-modal data set, and determining a target highway construction quality result. The application solves the technical problems of scattered highway construction quality evaluation data sources, difficulty in fusion analysis, and insufficient accuracy and real-time performance of evaluation results in the prior art, and achieves the technical effects of realizing multi-source data fusion utilization and improving construction quality evaluation accuracy and real-time performance.

Description

Technical Field

[0001] This invention relates to the field of quality assessment technology, and more specifically to a method and system for intelligent assessment of highway construction quality integrating Internet of Things (IoT) analysis. Background Technology

[0002] In the process of highway construction quality assessment, analysis typically relies on manual inspection records, sampling test data, and information collected by some independent monitoring equipment. This data is scattered across different systems or management stages, with inconsistent formats and standards, making effective integration and comprehensive utilization difficult. Quality judgments are often based on periodic summaries or experience-based analysis, lacking the ability to real-time correlate and process multi-dimensional data from the construction process. This results in delayed assessment results and limited accuracy, failing to meet the demands for refined and dynamic quality control in complex construction environments. Summary of the Invention

[0003] This application provides a method and system for intelligent assessment of highway construction quality that integrates Internet of Things (IoT) analysis, in order to address the technical problems in existing technologies where highway construction quality assessment data is scattered, difficult to integrate and analyze, and lacks accuracy and real-time performance of assessment results.

[0004] In view of the above problems, this application provides a method and system for intelligent assessment of highway construction quality that integrates Internet of Things (IoT) analysis.

[0005] The first aspect of this application provides a method for intelligent assessment of highway construction quality integrating Internet of Things (IoT) analytics, the method comprising:

[0006] An IoT device cluster is deployed in the target highway construction area. Based on the data characteristics of the IoT device cluster, a universal template for multi-device data sources is created. Multi-modal datasets of highway construction are collected in real time through the IoT device cluster. These datasets are then integrated and preprocessed using the universal template to obtain a usable multi-modal dataset of highway construction. Historical quality datasets of highway construction are collected. Based on these historical quality datasets, association mining and evaluation training are performed to build a highway construction quality assessment channel library, which includes construction quality assessment channels for multiple road segment types. Based on this library, road segment channel matching and construction quality assessment are performed on the usable multi-modal dataset of highway construction to determine the construction quality result of the target highway.

[0007] A second aspect of this application provides an intelligent assessment system for highway construction quality integrating Internet of Things (IoT) analytics, the system comprising:

[0008] The module includes a template creation module for deploying an IoT device cluster in the target highway construction area and creating a universal template for multiple device data sources based on the data characteristics of the IoT device cluster. A data acquisition module is used to collect multimodal datasets of highway construction in real time through the IoT device cluster and perform integrated preprocessing on the multimodal datasets based on the universal template to obtain usable multimodal datasets of highway construction. A channel construction module is used to collect historical quality datasets of highway construction and perform association mining and evaluation training based on these datasets to build a highway construction quality assessment channel library, which includes construction quality assessment channels for multiple road segment types. A quality result determination module is used to perform road segment channel matching and construction quality assessment on the usable multimodal datasets of highway construction based on the highway construction quality assessment channel library to determine the construction quality result of the target highway.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] This application deploys an IoT device cluster in the target highway construction area. Based on the data characteristics of the IoT device cluster, a universal template for multi-device data sources is created. The IoT device cluster collects multi-modal datasets of highway construction in real time. Based on the universal template for multi-device data sources, the multi-modal datasets of highway construction are integrated and preprocessed to obtain usable multi-modal datasets of highway construction. Historical quality datasets of highway construction are collected. Based on these historical quality datasets, association mining and evaluation training are performed to build a highway construction quality evaluation channel library. This library includes construction quality evaluation channels for multiple road segment types. Based on the highway construction quality evaluation channel library, road segment channel matching and construction quality evaluation are performed on the usable multi-modal datasets of highway construction to determine the construction quality result of the target highway. This invention solves the technical problems of scattered data sources, difficulty in fusion and analysis, and insufficient accuracy and real-time performance of highway construction quality evaluation results in the prior art. By constructing an IoT device cluster to collect multi-modal data and combining it with historical quality data to establish a construction quality evaluation channel library for intelligent matching and evaluation, it achieves the technical effect of realizing multi-source data fusion and utilization, and improving the accuracy and real-time performance of construction quality evaluation. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1A schematic diagram of the intelligent assessment method for highway construction quality integrating IoT analysis provided in this application embodiment;

[0013] Figure 2 A schematic diagram of the structure of the intelligent assessment system for highway construction quality integrating Internet of Things analysis provided in this application embodiment.

[0014] Explanation of reference numerals in the attached diagram: Template creation module 11, data acquisition module 12, channel construction module 13, quality result determination module 14. Detailed Implementation

[0015] This application provides an intelligent assessment method and system for highway construction quality that integrates Internet of Things (IoT) analytics. It addresses the technical problems in existing highway construction quality assessment technologies, such as fragmented data sources, difficulty in data fusion and analysis, and insufficient accuracy and real-time performance of assessment results. By constructing an IoT device cluster to collect multimodal data and combining it with historical quality data to establish a construction quality assessment channel library for intelligent matching and assessment, the application achieves the technical effect of multi-source data fusion and utilization, improving the accuracy and real-time performance of construction quality assessment.

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0017] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0018] Example 1, as Figure 1 As shown, this application provides an intelligent assessment method for highway construction quality integrating Internet of Things (IoT) analysis, the method comprising:

[0019] Step S100: Deploy an IoT device cluster in the target highway construction area, and create a multi-device data source general template based on the data characteristics of the IoT device cluster.

[0020] In this embodiment, when deploying an IoT device cluster in a target highway construction area, the topographic and boundary data of the target highway construction area are first acquired. Based on the topographic and boundary data, a 3D model is created to generate a 3D model of the highway construction area. Then, according to the construction stage requirements, a monitoring coverage analysis is performed on the 3D model of the highway construction area to determine the set of highway construction monitoring locations. Subsequently, IoT devices are selected based on the information of each location in the highway construction monitoring location set to obtain a set of monitoring IoT device types. Finally, based on the set of highway construction monitoring locations and the set of monitoring IoT device types, IoT devices are deployed in the target highway construction area to construct an IoT device cluster.

[0021] When creating a universal template for multi-device data sources based on the data characteristics of IoT device clusters, a multi-device cluster data application standard is constructed based on the data characteristics monitored by various types of devices in the IoT device cluster. Then, universal template fields are defined, including core fields and auxiliary fields. Auxiliary fields include data collection time, collection device number, and collection segment source. Subsequently, the core fields are parsed according to the multi-device cluster data application standard to obtain the multi-device universal template format specification, which includes data type, data unit, and data range. Finally, a universal template for multi-device data sources is created based on the universal template fields and the multi-device universal template format specification.

[0022] Furthermore, the method provided in the application embodiments, which involves deploying an IoT device cluster in the target highway construction area, further includes:

[0023] The process involves acquiring topographic and boundary data of the target highway construction area, performing 3D modeling based on this data, generating a 3D model of the highway construction area, conducting monitoring coverage analysis on the 3D model according to the construction stage requirements, and determining the set of monitoring locations for highway construction. IoT devices are then selected for each location within the set of monitoring locations to obtain a set of monitoring IoT device types. Finally, IoT devices are deployed in the target highway construction area based on the set of monitoring locations and the set of monitoring IoT device types to construct an IoT device cluster.

[0024] In this embodiment, the topographic and boundary data of the target highway construction area are first obtained. Specifically, topographic and boundary data corresponding to the target highway construction area are retrieved from a pre-stored surveying database or construction design data. The topographic and boundary data are then unified into a single plane coordinate system. The topographic data includes plane coordinate values ​​and corresponding elevation values, where the plane coordinate values ​​include X and Y coordinates, and the elevation value is the Z value. Each plane coordinate value (X, Y) is combined with its corresponding elevation value Z to directly form a three-dimensional coordinate point (X, Y, Z), where X represents the east-west position, Y represents the north-south position, and Z represents the altitude. These three-dimensional coordinate points are sequentially connected according to their spatial adjacency relationship to form a continuous surface grid structure. The boundary data is then used as an outer contour constraint. Data exceeding the boundary range is deleted, retaining only the three-dimensional coordinate points within the boundary range and their connection relationships, thereby generating a three-dimensional model of the highway construction area. This model visually represents the spatial morphology of the target highway construction area.

[0025] Next, when conducting a monitoring coverage analysis of the 3D model of the highway construction area according to the construction stage requirements, we first read the monitoring objects and monitoring areas specified in the construction stage requirements, and then map the monitoring objects to specific spatial locations in the 3D model of the highway construction area. Then, we check each 3D coordinate point and its corresponding structural region in the 3D model of the highway construction area one by one to determine whether the location is within the monitoring range specified in the construction stage requirements. We mark the 3D coordinate points within the monitoring range and merge adjacent 3D coordinate points belonging to the same monitoring area to form specific spatial locations. After marking and merging all monitoring areas, we summarize the highway construction monitoring location set.

[0026] Subsequently, when selecting IoT devices for each location within the highway construction monitoring cluster, the monitoring parameter type corresponding to each monitoring location is read. The monitoring parameter type is then compared item by item with the functional descriptions of each type of device in the IoT device cluster to filter out the device type capable of collecting the monitoring parameter. If multiple device types capable of collecting the monitoring parameter exist, their measurement ranges are compared to see if they cover the monitoring value range, and their sampling frequencies are compared to see if they meet the requirements of the construction phase, to determine the final matching device type. After completing the above matching for all monitoring locations, a monitoring IoT device type set is formed.

[0027] Finally, when deploying IoT devices in the target highway construction area based on the set of highway construction monitoring locations and the set of IoT device types, the corresponding devices are transported to the target highway construction area according to the correspondence between monitoring locations and device types. The devices are then installed and fixed at specific spatial locations within each monitoring location, ensuring that the installation position matches the corresponding 3D coordinate point in the 3D model. Each device is assigned a data acquisition device number, establishing a correspondence between the data acquisition device number and the monitoring location information. After installation, the devices undergo power-on checks and data acquisition tests to confirm that the acquired data can be transmitted normally to the data receiving end. Once all monitoring locations have been installed and tested, an IoT device cluster is constructed to achieve continuous data acquisition and transmission from the highway construction monitoring location set.

[0028] Furthermore, the method provided in the application embodiments, in creating a universal template for multi-device data sources, further includes:

[0029] Based on the data characteristics monitored by various types of devices in the IoT device cluster, a multi-device cluster data application standard is constructed; a general template field is defined, which includes core fields and auxiliary fields. The auxiliary fields include data collection time, collection device number, and collection segment source; the core fields are parsed according to the multi-device cluster data application standard to obtain a multi-device general template format specification, which includes data type, data unit, and data range; a multi-device data source general template is created based on the general template field and the multi-device general template format specification.

[0030] In this embodiment, when constructing a multi-device cluster data application standard based on the data characteristics monitored by various types of devices in an IoT device cluster, the data characteristics are first defined. Data characteristics refer to the data performance attributes exhibited by each type of device when outputting monitoring data, including data type, data unit, data range, data precision, data sampling frequency, and data encoding format. Among them, data type is used to distinguish between numerical or character representations, data unit is used to represent the unit of measurement corresponding to the monitored value, data range is used to limit the minimum and maximum allowable values ​​of the monitored value, data precision is used to represent the number of decimal places retained in the value, data sampling frequency is used to represent the number of data collections per unit time, and data encoding format is used to represent the structural form of the data during transmission. After defining the data characteristics, the output data of each type of device in the IoT device cluster is statistically analyzed and compared item by item to extract the differences in data characteristics of the same monitoring parameter in different devices. The same monitoring parameter is then standardized in terms of data type, data unit, data range, and data precision to form a multi-device cluster data application standard.

[0031] After establishing a multi-device cluster data application standard, a common template field is defined. The common template field includes core fields and auxiliary fields. The core fields are used to store the monitoring parameter values ​​after being unified according to the multi-device cluster data application standard. The auxiliary fields are used to record the time information and source information of the data. The auxiliary fields include the data collection time, the collection device number, and the source of the collection segment. The core fields and auxiliary fields are arranged in a fixed field order to form a unified field structure framework, so that each data contains complete numerical information and source information.

[0032] Subsequently, when parsing the core fields according to the multi-device cluster data application standard, each monitoring parameter in the core fields is validated. First, it is determined whether the data type conforms to the specified data type, and data that does not conform is converted. Second, it is determined whether the data unit conforms to the unified data unit, and data with discrepancies is converted. Third, the monitoring parameter value is compared with the specified data range, and data that exceeds the data range is marked or removed. After completing the above data type validation, unit conversion and range judgment, a multi-device general template format specification is obtained. The multi-device general template format specification includes data type, data unit and data range, and serves as a unified constraint rule for subsequent data writing.

[0033] Finally, based on the general template fields and the multi-device general template format specification, when creating a multi-device data source general template, the field structure and format specification are combined, each core field is bound to the corresponding data type, data unit and data range constraint rules, and a unified format and encoding rules are set for the data collection time, collection device number and collection road segment source. The above field structure and constraint rules are solidified into a standard data structure file, thereby creating a multi-device data source general template.

[0034] Step S200: Collect highway construction multimodal dataset in real time through the IoT device cluster, and perform integrated preprocessing on the highway construction multimodal dataset based on the multi-device data source general template to obtain a usable highway construction multimodal dataset.

[0035] In this embodiment, when collecting multimodal datasets of highway construction in real time through an IoT device cluster, the various types of monitoring devices in the IoT device cluster are first put into operation. Each monitoring device collects data from its corresponding construction object according to a preset sampling frequency. Different types of monitoring devices collect different categories of construction parameter data. The multimodal dataset of highway construction includes displacement data, vibration data, stress data, temperature data, humidity data, compaction data, thickness data, smoothness data, and image data. Displacement data is used to reflect changes in structural position, vibration data is used to reflect the vibration state during construction, stress data is used to reflect the stress state of the structure, and temperature and humidity data are used for... To reflect construction environmental conditions, compaction data is used to reflect the compaction quality of the subgrade or pavement, thickness data is used to reflect the thickness of the paving layer, smoothness data is used to reflect the surface quality of the pavement, and image data is used to reflect the appearance of the construction site. After completing data acquisition, each monitoring device records the acquired monitoring parameter values ​​along with the data acquisition time, acquisition device number, and source of the acquired road section, and uploads the resulting data records to the data receiving end. The data receiving end summarizes and organizes the data from different monitoring devices, classifies and stores them according to the data acquisition time and source of the acquired road section, and integrates different categories of data from the same time period and the same road section to form a multimodal dataset for highway construction.

[0036] Next, the multimodal dataset of highway construction is integrated and preprocessed based on a common template for multi-device data sources. This process begins by matching fields and data within the multimodal dataset according to the common template, mapping various monitoring parameters to the core and auxiliary fields in the template to generate a common template for multimodal data association. Subsequently, the data processing flow is decomposed based on this template, clarifying steps such as data type validation, data unit unification, and data range validation, thus constructing a multimodal data preprocessing program. Finally, the multimodal data preprocessing program is used to standardize the multimodal dataset, ensuring it meets the requirements for unified data type, unified data unit, and unified data range, resulting in a usable multimodal dataset for highway construction.

[0037] Furthermore, the method provided in the application embodiments, in order to obtain a usable multimodal dataset of highway construction, further includes:

[0038] The highway construction multimodal dataset is associated and mapped according to the general template of the multi-device data source to generate a general template for multimodal data association; the data preprocessing steps are parsed based on the general template for multimodal data association to construct a multimodal data preprocessing program; the highway construction multimodal dataset is standardized and preprocessed through the multimodal data preprocessing program to obtain a usable highway construction multimodal dataset.

[0039] In this embodiment, when mapping the multimodal dataset of highway construction according to the general template for multi-device data sources, the original data records in the multimodal dataset of highway construction are first read one by one. Each original data record is split into several data items, and the monitoring parameter name, monitoring parameter value, data acquisition time, acquisition device number, and acquisition section source are identified. Then, the monitoring parameter name is matched with the core fields in the general template for multi-device data sources, and the data acquisition time, acquisition device number, and acquisition section source are filled with the corresponding auxiliary fields. The data items are rearranged according to the preset field order in the general template for multi-device data sources to ensure that data records of different sources and types are consistent in field structure. Fields not defined in the general template for multi-device data sources are not retained, and fields defined in the template but missing in the original data are filled with null values. After the field structure is unified, a general template for multimodal data association is generated.

[0040] When parsing data preprocessing steps based on the general template for multimodal data association, the core and auxiliary fields in the general template are read field by field. The data type, data unit, and data range requirements for each field are specified in the general template for multi-device data sources. The processing steps required for each field are listed in sequence, including data type verification steps, data unit unification steps, and data range judgment steps. At the same time, time format checking steps and number format checking steps are listed for auxiliary fields. The processing steps corresponding to the above fields are organized into a complete data processing flow according to the execution order, so that the data preprocessing logic corresponds one-to-one with the field specifications in the general template for multi-device data sources, forming a multimodal data preprocessing program.

[0041] Finally, when standardizing the multimodal dataset of highway construction using a multimodal data preprocessing program, the data processing flow is executed sequentially for each data record in the multimodal data association general template. First, the data types in the core fields are checked to ensure they conform to the specified data types, and non-conforming data is converted to a different format. Then, the monitoring parameter values ​​in the core fields are converted to a unified data unit. Next, the converted values ​​are compared with the specified data range, and data exceeding the range is marked or removed. Simultaneously, the data acquisition time in the auxiliary fields is processed to a unified format, and the acquisition device number and acquisition section source are checked for completeness. After all fields have undergone the above processing, a dataset with consistent field structure, unified data types, unified data units, and a data range that meets the requirements is formed, resulting in a usable multimodal dataset of highway construction.

[0042] Step S300: Collect historical quality datasets of highway construction, perform association mining and evaluation training based on the historical quality datasets of highway construction, and build a highway construction quality evaluation channel library, which includes construction quality evaluation channels for multiple road sections.

[0043] In this embodiment of the application, when collecting the historical quality dataset of highway construction, the quality inspection records, acceptance records and quality assessment results of completed highway construction projects are retrieved from the historical database, and these data are sorted and summarized to form a unified historical quality dataset of highway construction.

[0044] Next, based on the historical quality dataset of highway construction, association mining and evaluation training are conducted to build a highway construction quality assessment channel library. First, the historical quality dataset of highway construction is classified according to the type of construction section. The quality inspection data corresponding to different types of construction sections are grouped to form a highway construction section type quality dataset. Then, a highway construction quality index system is established, which includes various quality indicators and their evaluation rules to reflect construction quality. Based on the highway construction quality index system, association mining and evaluation training are performed on the highway construction section type quality dataset to generate multi-section type construction quality assessment channels corresponding to different construction section types. Finally, the multi-section type construction quality assessment channels are uniformly numbered and structurally integrated, so that the construction quality assessment channels are organized and managed in parallel to form a highway construction quality assessment channel library.

[0045] Furthermore, the method provided in the application embodiments, in establishing a highway construction quality assessment database, also includes:

[0046] The historical quality dataset of highway construction is divided according to the type of construction section to obtain a quality dataset of highway construction section type; a highway construction quality index system is established, and the highway construction quality index system is used to perform association mining and evaluation training on the quality dataset of highway construction section type to generate a multi-section type construction quality evaluation channel; the multi-section type construction quality evaluation channels are integrated and identified in parallel to build a highway construction quality evaluation channel library.

[0047] In this embodiment, when processing the historical quality dataset of highway construction according to the type of construction section, firstly, each quality inspection record in the historical quality dataset of highway construction is read, and the construction section number, construction structure type, structural layer category, and corresponding quality inspection result fields are extracted. Then, the construction structure type field is matched against a preset classification table of construction section types. Records with the construction structure type of roadbed structure are assigned to the roadbed section data set, records with the construction structure type of asphalt pavement structure are assigned to the asphalt pavement section data set, and records with the construction structure type of cement concrete pavement structure are assigned to the cement concrete pavement section data set. After completing the field matching, the quality inspection records under the same construction section type are summarized and organized, and grouped and stored according to the construction section number to form the corresponding highway construction section type quality dataset.

[0048] Next, when establishing the highway construction quality index system, the first step is to statistically analyze and organize all the fields of the inspection items included in the historical quality dataset of highway construction, listing the inspection item names, data units, and numerical fields corresponding to each construction structure level. The inspection items are then categorized according to the construction structure level: compaction test data and deflection test data related to the subgrade are included in the subgrade quality index set; smoothness test data, thickness test data, and strength test data related to the surface layer are included in the pavement quality index set; and visual inspection data are included in the visual quality index set. For each quality index, the corresponding data field name, data unit, and allowable value range are clearly defined, and a correspondence between quality indicators and construction section types is established, forming a highway construction quality index system with a clear structure and well-defined field correspondences.

[0049] Subsequently, when using the highway construction quality index system to perform association mining and evaluation training on the highway construction section type quality dataset, firstly, according to the quality indicators determined in the highway construction quality index system, the corresponding fields in the highway construction section type quality dataset are subjected to index association mining to analyze the correlation between different quality indicators, forming a multi-section type quality index association dataset; then, a deep neural network is used to perform label evaluation training on the multi-section type quality index association dataset, and by learning the mapping relationship between quality index data and corresponding quality results, a multi-section type quality index evaluation channel set is constructed; finally, a weighted decision fusion is performed on the multi-section type quality index evaluation channel set to comprehensively integrate the evaluation results of different quality indicators and generate a multi-section type construction quality evaluation channel.

[0050] Finally, when integrating and identifying the construction quality assessment channels for multiple road segment types in parallel, each of the generated multi-road segment type construction quality assessment channels is first numbered, establishing a unique correspondence between each construction quality assessment channel and its corresponding construction segment type. Then, a correspondence table between construction segment types and construction quality assessment channels is established, enabling different construction segment types to quickly match the corresponding construction quality assessment channels through the table. Subsequently, all multi-road segment type construction quality assessment channels are centrally stored according to a unified structure, ensuring that each construction quality assessment channel exists in parallel and is independent of the others. Ultimately, a highway construction quality assessment channel database containing multiple construction quality assessment channels and their corresponding relationship information is formed.

[0051] Furthermore, the method provided in the application embodiments for generating multi-segment construction quality assessment channels also includes:

[0052] According to the highway construction quality index system, the quality dataset of highway construction section types is subjected to index association mining to obtain a multi-section type quality index association dataset; a deep neural network is used to perform label evaluation training on the multi-section type quality index association dataset to construct a multi-section type quality index evaluation channel set; a weighted decision fusion is performed on the multi-section type quality index evaluation channel set to generate a multi-section type construction quality evaluation channel.

[0053] In this embodiment, when performing index association mining on the quality dataset of highway construction section types according to the highway construction quality index system, the quality detection data in the quality dataset of highway construction section types are first classified according to the index classification structure in the highway construction quality index system to form a multi-section type quality index classification dataset. On this basis, according to the hierarchical relationship of the indicators in the highway construction quality index system, the multi-section type quality index classification dataset is subjected to index refinement cascade analysis, and the quality index data of different levels under the same construction section type are hierarchically expanded and structurally combined to generate a multi-section type quality index cascade dataset. Subsequently, the multi-section type quality index cascade dataset is subjected to association mining to analyze the correlation relationship and synergistic change characteristics between different quality indicators to obtain a multi-section type quality index association dataset.

[0054] Next, a deep neural network is used to train the multi-segment type quality indicator association dataset for label evaluation. When constructing the multi-segment type quality indicator evaluation channel set, the dataset is first constructed by concatenating the association features of multiple quality indicators corresponding to the same construction segment type and the same construction segment number into an input vector according to a fixed field order. The historical quality assessment results corresponding to that construction segment number are then used as labels for binding, ensuring that each training sample consists of an input vector and a label. Subsequently, missing value processing and numerical normalization are performed on all training samples to ensure that indicators with different dimensions in the input vector are within a unified numerical range. Then, the training set and validation set are divided according to a fixed ratio, and the input dimension of the deep neural network is set to the feature dimension of the input vector. The deep neural network is configured as a feedforward fully connected network structure, containing an input layer, hidden layers, and an output layer. The number of hidden layers is... The system is configured with 3 layers, with the number of hidden layer neurons set to 128, 64, and 32 respectively. The activation function for the hidden layers is ReLU. The output layer is configured to either classify or regress based on the label type, with Softmax used for classification and linear output for regression. The loss function is set to either cross-entropy loss or mean squared error loss, the optimizer is set to Adam, the learning rate is set to 0.001, the batch size is set to 64, and the number of training epochs is set to 100. In each training epoch, the training set is input into the deep neural network in batches for forward computation to obtain the output. Then, the loss is calculated based on the label and backpropagation is performed to update the network parameters. At the same time, the validation loss is calculated on the validation set, and the network parameters with the minimum validation loss are saved. After training, the deep neural network models obtained for different construction road segment types are solidified into corresponding quality indicator evaluation channels, forming a multi-segment type quality indicator evaluation channel set.

[0055] Finally, when performing weighted decision fusion on the quality indicator evaluation channel set for multiple road segment types, the validation performance of each quality indicator evaluation channel under the same construction road segment type is first calculated on the validation set. During classification output, the accuracy rate and error rate of the evaluation channel on the validation set are calculated. During regression output, the mean squared error of the evaluation channel on the validation set is calculated. Then, a weight set is determined based on the validation performance. During classification output, the accuracy rate of each evaluation channel is used as the initial weight, and all accuracy rates are normalized to make the sum of all weights equal to 1. During regression output, the reciprocal of the mean squared error of each evaluation channel is used as the initial weight, and the reciprocal value is normalized to make the sum of all weights equal to 1. Subsequently, the weight set is further calculated... The input vectors for the same construction section number are input into the quality indicator evaluation channels corresponding to the construction section type, respectively, to obtain the output results of each evaluation channel. Then, the output results are weighted according to the weight set. For the classification output scenario, the category probability vectors output by each channel are weighted and summed to obtain the fusion probability vector. For the regression output scenario, the numerical results output by each channel are weighted and summed to obtain the fusion numerical result. Finally, the weight set and weighting calculation rules are bound and solidified with the construction section type to form a multi-section type construction quality evaluation channel. This allows the multi-section type construction quality evaluation channel to output the final construction quality evaluation result of the corresponding construction section type after inputting the associated features of the quality indicators.

[0056] Furthermore, the method provided in the application embodiments, in obtaining the multi-segment type quality index association dataset, also includes:

[0057] According to the highway construction quality index system, the quality dataset of highway construction section types is classified to obtain a multi-section type quality index classification dataset; the index refinement cascade analysis is performed on the multi-section type quality index classification dataset to obtain a multi-section type quality index cascade dataset; the association mining is performed on the multi-section type quality index cascade dataset to obtain a multi-section type quality index association dataset.

[0058] In this embodiment, when classifying the quality dataset of highway construction section types according to the highway construction quality index system, each quality inspection record in the historical quality dataset of highway construction is first read. Each record contains the construction section number, construction structure type, quality inspection parameters, and their corresponding measurement values. Based on the classification rules in the highway construction quality index system, these data are grouped according to quality indicators. Specifically, quality indicators can be divided into subgrade quality indicators, pavement quality indicators, and appearance quality indicators, etc. For each record, by comparing the construction structure type with the preset classification rules, the quality data in that record is classified into the corresponding quality indicator set. For example, inspection data related to subgrade quality, such as compaction and deflection, are classified into the subgrade quality indicator set, while inspection data related to pavement quality, such as smoothness, thickness, and strength, are classified into the pavement quality indicator set. All quality inspection data, based on the construction section type and construction structure type, form multiple different quality datasets, ultimately resulting in a multi-section type quality index classification dataset.

[0059] Next, when performing a detailed cascade analysis of the multi-section type quality indicator classification dataset, the various quality indicators are first hierarchically divided according to the hierarchical relationship in the highway construction quality indicator system. For example, a primary indicator might be an overall quality assessment, secondary indicators include compaction degree and smoothness, and tertiary indicators are specific test values, such as a compaction degree of 95% for a certain roadbed section. Then, for each construction section type, each indicator in the multi-section type quality indicator classification dataset is further refined. First, secondary indicators are associated with primary indicators, and then tertiary indicators are cascaded with secondary indicators, forming a hierarchically expanding data structure. This process ensures that each quality indicator data is reflected in higher-level quality assessments, and finally, all levels of quality indicator information are integrated into a complete record, generating a multi-section type quality indicator cascade dataset.

[0060] Finally, association mining was performed on the cascaded dataset of quality indicators for multiple road segment types. This process began by first assessing the criticality of each quality indicator within the highway construction quality indicator system, calculating the criticality coefficient for each indicator. Next, based on the calculated criticality coefficients, the association hierarchy of each quality indicator within the dataset was determined, identifying which indicators play a dominant role in the overall assessment and which are secondary influencing factors. Then, based on the determined association hierarchy of the construction quality indicator data, the cascaded dataset of quality indicators for multiple road segment types was mined, ultimately yielding the associated dataset of quality indicators for multiple road segment types.

[0061] Furthermore, the method provided in the application embodiment, which involves performing association mining on the concatenated dataset of multi-segment type quality indicators to obtain a multi-segment type quality indicator association dataset, further includes:

[0062] A criticality assessment is performed on each indicator in the highway construction quality indicator system to obtain the criticality coefficient of the construction quality indicator; based on the criticality coefficient of the construction quality indicator, the data association level of the construction quality indicator is determined; based on the data association level of the construction quality indicator, association mining is performed on the cascaded dataset of the multi-segment type quality indicators to obtain the multi-segment type quality indicator association dataset.

[0063] In this embodiment, when conducting a criticality assessment of each indicator in the highway construction quality indicator system, all relevant quality indicators are first extracted from the system. These indicators include the compaction degree of the subgrade, the smoothness, thickness, and strength of the pavement. To evaluate the impact of each quality indicator on the construction quality assessment results, Pearson correlation analysis is used. Specifically, historical construction data is first collected, and the values ​​of each quality indicator are paired with the corresponding construction quality assessment results. Then, the Pearson correlation coefficient between them is calculated. The Pearson correlation coefficient ranges from -1 to 1; a value closer to 1 indicates a stronger linear relationship between the two variables, while a value closer to 0 indicates a weaker correlation. Through this analysis, the criticality coefficient of each quality indicator is obtained, which represents the influence of that indicator on the construction quality assessment results. For example, if the Pearson correlation coefficient between compaction degree and construction quality assessment results is 0.85, it indicates that compaction degree has a significant impact on construction quality assessment, and therefore the criticality coefficient of compaction degree is high; if the Pearson correlation coefficient between smoothness and construction quality assessment results is 0.45, then the criticality coefficient of smoothness is low. Ultimately, this calculation yields the criticality coefficient of each quality indicator.

[0064] When determining the data association level of construction quality indicators based on their criticality coefficients, the first step is to compare the calculated criticality coefficient of each quality indicator with a preset standard. Assuming a threshold standard is set, quality indicators with a criticality coefficient greater than 0.6 are considered high-level, those between 0.3 and 0.6 are considered mid-level, and those below 0.3 are considered low-level. Quality indicators with criticality coefficients greater than 0.6, such as compaction, will be assigned to high-level indicators, while those with lower criticality coefficients, such as smoothness, will be assigned to low-level indicators. This assignment forms the data association level for each quality indicator.

[0065] When performing association mining on a cascaded dataset of quality indicators for multiple road sections based on the hierarchical association of construction quality indicators, the quality indicators in the dataset are first sorted according to their association hierarchy to ensure that higher-level quality indicators are processed first, followed by lower-level ones. Next, the Apriori algorithm is used for association rule mining, setting minimum support and minimum confidence to identify frequent itemsets—combinations of quality indicators that appear together in multiple records. As the hierarchy increases, the amount of data for association mining gradually increases because higher-level quality indicators are associated with more data. Finally, a multi-road section type quality indicator association dataset is obtained through association mining. This dataset includes data related to each quality indicator hierarchy, covering monitoring data corresponding to different levels of quality indicators.

[0066] Step S400: Based on the highway construction quality assessment channel library, perform road segment channel matching and construction quality assessment on the available highway construction multimodal dataset to determine the target highway construction quality result.

[0067] In this embodiment, when matching road segment channels to the available multimodal dataset of highway construction based on the highway construction quality assessment channel library, the data of each road segment in the available multimodal dataset of highway construction is first matched with the construction quality assessment channels for each road segment type in the highway construction quality assessment channel library. Through this process, a multimodal road segment data matching channel set is obtained, which contains the matching relationships between the construction quality assessment channels for each road segment type and the corresponding road segment data. Then, the multimodal road segment data matching channel set is used to perform association mapping evaluation and ensemble analysis on the available multimodal dataset of highway construction. In this process, by analyzing the correlation between road segment data and assessment channels, data from different sources are integrated to ultimately determine the target highway construction quality assessment result.

[0068] Furthermore, the method provided in the application embodiments, in determining the construction quality result of the target highway, further includes:

[0069] Based on the highway construction quality assessment channel library, road segment channel matching is performed on the available highway construction multimodal dataset to obtain a multimodal road segment data matching channel set; the multimodal road segment data matching channel set is used to perform association mapping evaluation and integration analysis on the available highway construction multimodal dataset to determine the target highway construction quality result.

[0070] In this embodiment, when matching road segment channels based on the highway construction quality assessment channel library to the available multimodal highway construction dataset, the data for each road segment is first extracted from the highway construction dataset. This data includes multiple parameters such as structural displacement, vibration, temperature, humidity, and stress, reflecting various changes during the construction process. Next, the data for each road segment is matched with a suitable assessment channel using the highway construction quality assessment channel library, ensuring that each road segment's data is matched to the correct assessment channel according to its type (e.g., roadbed, asphalt pavement, cement concrete pavement, etc.). During the matching process, based on the data characteristics of each road segment, the corresponding assessment channel is found, forming a multimodal road segment data matching channel set.

[0071] After obtaining the multimodal road segment data matching channel set, the fit between each data point and its corresponding evaluation channel is checked through association mapping evaluation. First, it is verified whether the data meets the requirements of the evaluation channel. If the data match successfully, integrated analysis is performed. At this point, the evaluation results of multiple evaluation channels are weighted and fused, combining the evaluation results of different channels according to their respective weights to ensure that the importance of different types of data and evaluation channels in quality assessment is reasonably reflected. Finally, through this process, the construction quality result of the target highway is obtained.

[0072] In summary, the embodiments of this application have at least the following technical effects:

[0073] This application deploys an IoT device cluster in the target highway construction area. Based on the data characteristics of the IoT device cluster, a universal template for multi-device data sources is created. The IoT device cluster collects multi-modal datasets of highway construction in real time. Based on the universal template for multi-device data sources, the multi-modal datasets of highway construction are integrated and preprocessed to obtain usable multi-modal datasets of highway construction. Historical quality datasets of highway construction are collected. Based on these historical quality datasets, association mining and evaluation training are performed to build a highway construction quality evaluation channel library. This library includes construction quality evaluation channels for multiple road segment types. Based on the highway construction quality evaluation channel library, road segment channel matching and construction quality evaluation are performed on the usable multi-modal datasets of highway construction to determine the construction quality result of the target highway. This invention solves the technical problems of scattered data sources, difficulty in fusion and analysis, and insufficient accuracy and real-time performance of highway construction quality evaluation results in the prior art. By constructing an IoT device cluster to collect multi-modal data and combining it with historical quality data to establish a construction quality evaluation channel library for intelligent matching and evaluation, it achieves the technical effect of realizing multi-source data fusion and utilization, and improving the accuracy and real-time performance of construction quality evaluation.

[0074] Example 2, based on the same inventive concept as the intelligent assessment method for highway construction quality integrating IoT analysis in the foregoing examples, such as... Figure 2 As shown, this application provides an intelligent assessment system for highway construction quality integrating Internet of Things (IoT) analysis. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0075] The template creation module 11 is used to deploy an IoT device cluster in the target highway construction area and create a universal template for multiple device data sources based on the data characteristics of the IoT device cluster. The data acquisition module 12 is used to collect multimodal datasets of highway construction in real time through the IoT device cluster and perform integrated preprocessing on the multimodal datasets of highway construction based on the universal template for multiple device data sources to obtain usable multimodal datasets of highway construction. The channel construction module 13 is used to collect historical quality datasets of highway construction and perform association mining and evaluation training based on the historical quality datasets of highway construction to build a highway construction quality evaluation channel library, which includes construction quality evaluation channels for multiple road segments. The quality result determination module 14 is used to perform road segment channel matching and construction quality evaluation on the usable multimodal datasets of highway construction based on the highway construction quality evaluation channel library to determine the construction quality result of the target highway.

[0076] Furthermore, the system is also used to implement the following functions:

[0077] The process involves acquiring topographic and boundary data of the target highway construction area, performing 3D modeling based on this data, generating a 3D model of the highway construction area, conducting monitoring coverage analysis on the 3D model according to the construction stage requirements, and determining the set of monitoring locations for highway construction. IoT devices are then selected for each location within the set of monitoring locations to obtain a set of monitoring IoT device types. Finally, IoT devices are deployed in the target highway construction area based on the set of monitoring locations and the set of monitoring IoT device types to construct an IoT device cluster.

[0078] Furthermore, the system is also used to implement the following functions:

[0079] Based on the data characteristics monitored by various types of devices in the IoT device cluster, a multi-device cluster data application standard is constructed; a general template field is defined, which includes core fields and auxiliary fields. The auxiliary fields include data collection time, collection device number, and collection segment source; the core fields are parsed according to the multi-device cluster data application standard to obtain a multi-device general template format specification, which includes data type, data unit, and data range; a multi-device data source general template is created based on the general template field and the multi-device general template format specification.

[0080] Furthermore, the system is also used to implement the following functions:

[0081] The highway construction multimodal dataset is associated and mapped according to the general template of the multi-device data source to generate a general template for multimodal data association; the data preprocessing steps are parsed based on the general template for multimodal data association to construct a multimodal data preprocessing program; the highway construction multimodal dataset is standardized and preprocessed through the multimodal data preprocessing program to obtain a usable highway construction multimodal dataset.

[0082] Furthermore, the system is also used to implement the following functions:

[0083] The historical quality dataset of highway construction is divided according to the type of construction section to obtain a quality dataset of highway construction section type; a highway construction quality index system is established, and the highway construction quality index system is used to perform association mining and evaluation training on the quality dataset of highway construction section type to generate a multi-section type construction quality evaluation channel; the multi-section type construction quality evaluation channels are integrated and identified in parallel to build a highway construction quality evaluation channel library.

[0084] Furthermore, the system is also used to implement the following functions:

[0085] According to the highway construction quality index system, the quality dataset of highway construction section types is subjected to index association mining to obtain a multi-section type quality index association dataset; a deep neural network is used to perform label evaluation training on the multi-section type quality index association dataset to construct a multi-section type quality index evaluation channel set; a weighted decision fusion is performed on the multi-section type quality index evaluation channel set to generate a multi-section type construction quality evaluation channel.

[0086] Furthermore, the system is also used to implement the following functions:

[0087] According to the highway construction quality index system, the quality dataset of highway construction section types is classified to obtain a multi-section type quality index classification dataset; the index refinement cascade analysis is performed on the multi-section type quality index classification dataset to obtain a multi-section type quality index cascade dataset; the association mining is performed on the multi-section type quality index cascade dataset to obtain a multi-section type quality index association dataset.

[0088] Furthermore, the system is also used to implement the following functions:

[0089] A criticality assessment is performed on each indicator in the highway construction quality indicator system to obtain the criticality coefficient of the construction quality indicator; based on the criticality coefficient of the construction quality indicator, the data association level of the construction quality indicator is determined; based on the data association level of the construction quality indicator, association mining is performed on the cascaded dataset of the multi-segment type quality indicators to obtain the multi-segment type quality indicator association dataset.

[0090] Furthermore, the system is also used to implement the following functions:

[0091] Based on the highway construction quality assessment channel library, road segment channel matching is performed on the available highway construction multimodal dataset to obtain a multimodal road segment data matching channel set; the multimodal road segment data matching channel set is used to perform association mapping evaluation and integration analysis on the available highway construction multimodal dataset to determine the target highway construction quality result.

[0092] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for intelligent assessment of highway construction quality integrating Internet of Things (IoT) analysis, characterized in that: The method includes: Deploy an IoT device cluster in the target highway construction area, and create a common template for multiple device data sources based on the data characteristics of the IoT device cluster; The multimodal dataset of highway construction is collected in real time by the IoT device cluster, and the multimodal dataset of highway construction is integrated and preprocessed based on the common template of the multi-device data source to obtain a usable multimodal dataset of highway construction. Collect a historical quality dataset of highway construction, perform association mining and evaluation training based on the historical quality dataset of highway construction, and build a highway construction quality evaluation channel library, which includes construction quality evaluation channels for multiple road sections. Based on the highway construction quality assessment channel library, the available highway construction multimodal dataset is used to perform road segment channel matching and construction quality assessment to determine the target highway construction quality result.

2. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 1, characterized in that, Deploy a cluster of IoT devices in the target highway construction area, including: Obtain the topographic and boundary data of the target highway construction area, and perform three-dimensional modeling based on the topographic and boundary data to generate a three-dimensional model of the highway construction area. According to the needs of the construction stage, a monitoring coverage analysis was performed on the three-dimensional model of the highway construction area to determine the set of highway construction monitoring locations. IoT devices were selected for each location in the highway construction monitoring area to obtain a set of monitoring IoT device types; Based on the set of highway construction monitoring locations and the set of monitoring IoT device types, IoT devices are deployed in the target highway construction area to build an IoT device cluster.

3. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 1, characterized in that, Create a universal template for multi-device data sources, including: Based on the data characteristics monitored by various types of devices in the IoT device cluster, a multi-device cluster data application standard is constructed. Define general template fields, which include core fields and auxiliary fields. The auxiliary fields include data collection time, collection device number, and collection segment source. The core fields are parsed according to the multi-device cluster data application standard to obtain a multi-device general template format specification, which includes data type, data unit, and data range. Based on the general template fields and the multi-device general template format specification, a multi-device data source general template is created.

4. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 1, characterized in that, A usable multimodal dataset of highway construction was obtained, including: The highway construction multimodal dataset is associated and mapped according to the multi-device data source general template to generate a multimodal data association general template; Based on the general template for multimodal data association, the data preprocessing steps are analyzed to construct a multimodal data preprocessing program; The highway construction multimodal dataset is standardized and preprocessed using the multimodal data preprocessing program to obtain a usable highway construction multimodal dataset.

5. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 1, characterized in that, Establish a database for assessing highway construction quality, including: The historical quality dataset of highway construction is divided according to the type of construction section to obtain a quality dataset of highway construction section type. A highway construction quality index system is established, and the highway construction quality index system is used to perform correlation mining and evaluation training on the quality dataset of highway construction section types to generate a multi-section type construction quality evaluation channel. The construction quality assessment channels for the various road sections are integrated and identified in parallel to build a highway construction quality assessment channel database.

6. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 5, characterized in that, Generate construction quality assessment channels for multiple road segment types, including: According to the highway construction quality index system, the quality dataset of highway construction section types is subjected to index association mining to obtain a multi-section type quality index association dataset. A deep neural network is used to identify and evaluate the associated dataset of quality indicators for the multi-segment type, and a multi-segment type quality indicator evaluation channel set is constructed. The multi-segment type quality index evaluation channel set is weighted and fused to generate a multi-segment type construction quality evaluation channel.

7. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 6, characterized in that, The resulting dataset contains a correlation of quality indices for multiple road segment types, including: The highway construction section type quality dataset is classified according to the highway construction quality index system to obtain a multi-section type quality index classification dataset. The quality index classification datasets of the multi-segment type are subjected to index refinement and concatenation analysis to obtain the concatenated datasets of quality indices of the multi-segment type. The association mining of the concatenated dataset of quality indicators for multiple road segments is performed to obtain the associated dataset of quality indicators for multiple road segments.

8. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 7, characterized in that, The association mining of the concatenated dataset of quality indicators for multiple road segments is performed to obtain an association dataset of quality indicators for multiple road segments, including: A criticality assessment was conducted on each indicator in the highway construction quality indicator system to obtain the criticality coefficient of the construction quality indicator. Based on the criticality coefficients of the construction quality indicators, the data association levels of the construction quality indicators are determined; Based on the data association hierarchy of the construction quality indicators, the association mining of the multi-segment type quality indicator cascade dataset is performed to obtain the multi-segment type quality indicator association dataset.

9. The intelligent assessment method for highway construction quality integrating IoT analysis as described in claim 1, characterized in that, Determine the construction quality results of the target highway, including: Based on the highway construction quality assessment channel library, road segment channel matching is performed on the available highway construction multimodal dataset to obtain a multimodal road segment data matching channel set; The multimodal road segment data matching channel set is used to perform association mapping evaluation and integrated analysis on the available highway construction multimodal dataset to determine the construction quality results of the target highway.

10. A smart assessment system for highway construction quality integrating Internet of Things (IoT) analysis, characterized in that: The system is used to execute the intelligent assessment method for highway construction quality integrating IoT analysis as described in any one of claims 1-9, and the system includes: The template creation module is used to deploy an IoT device cluster in the target highway construction area and create a general template for multiple device data sources based on the data characteristics of the IoT device cluster. The data acquisition module is used to collect multimodal datasets of highway construction in real time through the IoT device cluster, and to perform integrated preprocessing on the multimodal datasets of highway construction based on the common template of the multi-device data source to obtain usable multimodal datasets of highway construction. The channel construction module is used to collect historical quality datasets of highway construction, perform correlation mining and evaluation training based on the historical quality datasets of highway construction, and build a highway construction quality evaluation channel library, which includes construction quality evaluation channels for multiple road sections. The quality result determination module is used to perform road segment channel matching and construction quality assessment on the available highway construction multimodal dataset based on the highway construction quality assessment channel library, and determine the construction quality result of the target highway.

Images