Municipal road quality diagnosis method and system based on Internet of Things

Abstract

The invention provides a municipal road quality diagnosis method and system based on the Internet of Things, and relates to the technical field of the Internet of Things. The method comprises the steps that an intelligent sensor network collects multi-dimensional data including municipal road surface sensing data and crack images in real time; transmitting the collected municipal road surface sensing data to a cloud server; the cloud server calculates a road condition index by adopting a multi-source data fusion calculation method; automatically analyzing the crack image by using a deep learning model, and calculating the type and position parameters of the road surface crack; a real-time road quality comprehensive evaluation report is generated by integrating a data fusion calculation result and an image analysis calculation result; and automatically pushing the generated road quality comprehensive evaluation report to a municipal management platform, and automatically triggering to generate corresponding road maintenance early warning and maintenance suggestions. According to the invention, the road inspection and maintenance efficiency can be greatly improved, the maintenance cost is effectively reduced, and the intelligent and refined management target of urban roads is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of the Internet of Things, and particularly to a method and system for diagnosing the quality of municipal roads based on the Internet of Things. Background Art

[0002] With the acceleration of the urbanization process, the maintenance of road infrastructure has become one of the important tasks of urban management. At present, most traditional road quality inspections rely on manual patrol methods, which not only consume a large amount of manpower and time, have low patrol efficiency, but are also easily affected by the subjective experience of patrol personnel and environmental factors, making it difficult to ensure the consistency and accuracy of inspection results. In addition, the manual patrol cycle is long, and it is impossible to detect the subtle defects and hidden dangers on the road surface in real time, which may lead to the gradual deterioration of problems, further increasing the road maintenance cost and the risk of potential traffic accidents. On the other hand, some regions have begun to try to use simple mechanical equipment or drones for auxiliary inspections, but these methods are usually limited by problems such as high equipment costs, difficult maintenance, and limited coverage, and cannot be widely promoted and applied in real time and efficiently. Therefore, there is an urgent need for an efficient, real-time, and automated municipal road quality diagnosis technology to more comprehensively and timely understand the road conditions and improve the efficiency of road maintenance and management. Summary of the Invention

[0003] The present invention provides a method for diagnosing the quality of municipal roads based on the Internet of Things, which includes: S10. The intelligent sensor network real-time collects multi-dimensional data including the perception data of the municipal road surface and crack images; S20. The intelligent sensor network transmits the collected perception data of the municipal road surface to the cloud server; S30. The cloud server calculates the road condition index based on the perception data of the municipal road surface; S40. The cloud server automatically analyzes the crack images using a deep learning model and calculates the type and position parameters of the road surface cracks; S50. Combining the road condition index and the type and position parameters of the road surface cracks, a real-time comprehensive road quality evaluation report is generated; S60. The generated comprehensive road quality evaluation report is automatically pushed to the municipal management platform, and the corresponding road maintenance warning and repair suggestions are automatically triggered and generated.

[0004] A method for diagnosing the quality of municipal roads based on the Internet of Things as described above, wherein the layout process of the intelligent sensor network includes: Vibration sensors are arranged on the road surface and underground structures to real-time collect the vibration amplitude, frequency, and propagation path data of the road; Pressure sensors are arranged on the road surface or subgrade structure to detect the pressure changes and load distribution in real time when vehicles pass by; Temperature sensors and humidity sensors are arranged on the road surface and underground structure to collect the changes of ambient temperature and humidity in real time; High-definition image sensors are arranged on the road surface to automatically capture images of road cracks, potholes and surface irregular features at preset time intervals or trigger conditions.

[0005] A method for diagnosing the quality of municipal roads based on the Internet of Things as described above, wherein the data transmission process includes: Intelligent sensor nodes stably transmit the collected data to the cloud server in real time through the Internet of Things communication network; The time synchronization protocol is adopted during the transmission process to uniformly mark the time stamps of different sensor data; Sensor nodes are arranged in a hexagonal grid manner to ensure the integrity and comprehensiveness of road data collection.

[0006] A method for diagnosing the quality of municipal roads based on the Internet of Things as described above, wherein the data fusion calculation process includes: Perform unified data format conversion on the collected original data; Perform data normalization processing on the converted data using the Z-score normalization method; Adopt the sliding window anomaly detection algorithm to automatically identify and eliminate the outliers in the data; Perform multi-source data fusion on the data after eliminating anomalies through dynamic weight factors to calculate the road condition index.

[0007] A method for diagnosing the quality of municipal roads based on the Internet of Things as described above, wherein the automatic analysis process of crack images includes: Perform image enhancement preprocessing such as denoising, contrast enhancement and edge sharpening on the crack image data; Use a deep learning model constructed by integrating multi-scale convolutional networks, deep residual networks and attention mechanisms to automatically analyze crack images; Based on the trained deep learning model, automatically realize crack type classification and accurate calculation of position coordinates, and convert the image coordinates into actual road geographical coordinates.

[0008] A method for diagnosing the quality of municipal roads based on the Internet of Things as described above, wherein the generation of a real-time comprehensive evaluation report on road quality includes: Use the big data correlation analysis method to analyze the relationship between the overall road condition and crack defects; Calculate the comprehensive evaluation index of road quality based on the association analysis results, and generate a comprehensive evaluation report including the defect spatial distribution map, defect type, severity, and overall evaluation.

[0009] A method for diagnosing the quality of municipal roads based on the Internet of Things as described above, wherein the process of automatically triggering maintenance warnings and repair suggestions includes: Set the maintenance warning level and alarm threshold according to the evaluation index in the comprehensive evaluation report of road quality; When the evaluation index reaches or exceeds the preset alarm threshold, automatically generate maintenance warnings and repair suggestion plans corresponding to the level; Automatically push the generated maintenance warnings and repair suggestions to the municipal management platform and the terminal devices of relevant management personnel in real time.

[0010] The present invention also provides a system for diagnosing the quality of municipal roads based on the Internet of Things, which includes: An intelligent sensor network for real-time collection of road vibration, pressure, temperature, humidity, and crack image data; A data processing module for calculating the road condition index based on the surface perception data of the municipal road; An image analysis module for automatically analyzing the crack image data to identify the road defect type and location; A diagnostic comprehensive module for integrating the results of the data processing module and the image analysis module to generate a real-time comprehensive evaluation report of road quality; A warning module for pushing the generated comprehensive evaluation report of road quality to the municipal management platform and automatically generating maintenance warnings and repair suggestions The beneficial effects achieved by the present invention are as follows: By deploying an intelligent sensor network, the present invention can collect multi-dimensional road environment data in real time and adopt a multi-source data fusion calculation method to accurately calculate the road condition index; use a deep learning model to automatically and efficiently analyze crack images to accurately identify the defect type and location; further generate a real-time and accurate comprehensive evaluation report of road quality through a dual calculation method of data fusion and image analysis; at the same time, automatically push the evaluation report to the municipal management platform to intelligently trigger maintenance warnings and repair suggestions, which can greatly improve the efficiency of road inspection and maintenance, effectively reduce the maintenance cost, and achieve the goal of intelligent and refined management of urban roads. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0012] Figure 1 It is a flowchart of a method for diagnosing the quality of municipal roads based on the Internet of Things provided in Embodiment 1 of this application. Specific implementation manner

[0013] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the protection scope of the present invention.

[0014] Embodiment 1 As Figure 1 shown, Embodiment 1 of this application provides a method for diagnosing the quality of municipal roads based on the Internet of Things, including the following steps: S10. The intelligent sensor network collects multi-dimensional data including surface perception data and crack images of the municipal road in real time; A plurality of intelligent sensor nodes are arranged in different sections of the municipal road. The sensor nodes include vibration sensors, pressure sensors, temperature sensors, humidity sensors, and image sensors. The sensor nodes are communicatively connected to the municipal management platform to form a complete data collection and transmission link.

[0015] The vibration sensors are arranged on the road surface and in the underground structure, and can sense the vibration amplitude, vibration frequency, and vibration propagation path on the road surface in real time. When a vehicle passes by or the external environment changes (such as weather changes), the vibration sensors record the vibration data and transmit it to the cloud server through the Internet of Things communication module. By analyzing the changes in vibration amplitude and frequency, the occurrence of cracks or structural damages on the road surface can be identified.

[0016] The pressure sensors are installed on the road surface or in the roadbed structure, and are used to detect the pressure changes caused by vehicles during driving in real time. By recording the pressure data, the load distribution at different positions can be quantified, and the road wear caused by overloading or uneven loading can be identified.

[0017] The temperature and humidity sensors are deployed on the road surface and in the underground structure, and are used to sense the dynamic changes of the ambient temperature and humidity in real time. By recording the changes in temperature and humidity, important environmental parameters for the aging of road materials and the expansion of cracks are provided.

[0018] The image sensor uses a high-definition camera to automatically capture high-definition images of the road surface at preset time intervals or trigger conditions. The image sensor can preprocess the image data through the edge computing module to extract cracks, potholes, and surface irregular features on the road surface.

[0019] S20. The intelligent sensor network transmits the perceived data of the municipal road surface collected in step S10 to the cloud server; The multi-dimensional data such as vibration, pressure, temperature, humidity, and crack images collected by the intelligent sensor nodes are transmitted to the cloud server in real time and stably through the Internet of Things communication network During the data transmission process, the time synchronization protocol (such as the PTP protocol) is used to mark the time stamps of various sensor data to ensure the consistency of the time dimension of different sensor data. To improve the data collection coverage rate, the sensor nodes are arranged in a hexagonal grid pattern to ensure the comprehensiveness of road coverage and the integrity of the collected data.

[0020] S30. The cloud server calculates the road condition index based on the perceived data of the municipal road surface; After receiving the above data, the cloud server performs refined data preprocessing on the collected data to improve the data quality and ensure the accuracy of subsequent calculations. The data preprocessing process specifically includes: (1) Unified conversion of data formats: For the different formats and protocols of the data collected by each sensor, they are uniformly converted into a standardized data format for subsequent fusion analysis.

[0021] (2) Data normalization processing: The Z-score normalization method is used to perform normalization operations on various types of data to eliminate the differences between the data dimensions and ensure the comparability and fusion rationality between the data.

[0022] (3) Detection and processing of abnormal data: Through statistical analysis methods (such as the sliding window anomaly detection algorithm), the abnormal points and noise data in the data are detected and effectively removed to ensure the high reliability and integrity of the data.

[0023] Subsequently, the cloud server performs data fusion processing on the preprocessed multi-source heterogeneous data. To scientifically and accurately reflect the actual health status of the road, the cloud server adopts a multi-source data fusion calculation method to calculate the road condition index DSI. The specific calculation formula is defined as follows: , where DSI represents the road condition index, and the value range is (0,1). The closer the value is to 1, the better the road condition; the closer it is to 0, the worse the road condition; represents the vibration data collected by the i-th vibration sensor node at time t; represents the pressure data collected by the i-th pressure sensor node at time t; represents the temperature data collected by the i-th temperature sensor node at time t; represents the humidity data collected by the i-th humidity sensor node at time t; It represents the crack feature data extracted by the j-th image sensor node; n represents the total number of vibration, pressure, temperature, and humidity sensor nodes; m represents the total number of image sensor nodes; It represents the dynamic weight factor of each sensor data, which is automatically adjusted dynamically according to sensor performance, location, real-time environmental conditions, and historical data reliability; It represents the adjustment coefficient, which controls the sensitivity of the exponential function to the change of the overall fusion data; It is a small positive constant designed to prevent the denominator from being zero; It is a non-linear adjustment parameter, which controls the influence of the temperature and humidity combined data on the overall exponent, controls the influence weight of the image data, and adjusts the sensitivity of the non-linear fusion of the image data. Through this refined data fusion and exponential calculation method, the real state of the road can be comprehensively, accurately, and real-time diagnosed, effectively supporting the maintenance decision-making of the municipal road management department and the implementation of maintenance measures.

[0024] S40. The cloud server uses a deep learning model to automatically analyze the crack images and calculate the type and location parameters of the cracks on the road surface; After the cloud server completes the preprocessing of multi-dimensional data and the calculation of the road condition index, it further uses the trained deep learning model to automatically analyze the crack image data collected and preprocessed in step S10 to accurately calculate the type and location parameters of the cracks on the road surface.

[0025] In the specific implementation process, the cloud server first performs image enhancement preprocessing on the crack image data transmitted to the server, including operations such as image denoising, contrast enhancement, and crack edge sharpening, to improve the identifiability of crack features in the image.

[0026] Subsequently, the cloud server performs automatic crack analysis and processing based on a deep learning model that integrates a multi-scale convolutional network, a deep residual network (ResNet), and an attention mechanism (Attention mechanism). The specific structure of this model is: an input layer, which inputs the enhanced crack image data; a multi-scale convolutional layer, which uses convolutional kernels of various sizes to extract crack features; a residual network layer, which is used to overcome the problem of gradient disappearance in network training and improve the efficiency of deep network feature extraction; an attention mechanism layer, which automatically identifies and enhances the crack region features in the image; a feature fusion layer, which integrates image features from different scales and different levels; an output layer, which automatically realizes crack type classification and the calculation of crack position coordinates according to the fused features. The CNN model trained with a large amount of labeled training data can achieve automatic crack classification, and the identification of crack types includes but is not limited to categories such as linear cracks, reticular cracks, block cracks, and crisscross cracks. The classification calculation uses the model output features combined with the Softmax classification function to obtain the crack category.

[0027] Meanwhile, the crack position parameters are obtained through regression calculation using the features output by the CNN model. The cloud server comprehensively considers the multi-dimensional features in the image space for calculation to accurately obtain the two-dimensional coordinate position of the crack in the image pixel coordinate system. The specific formula is as follows: , where represents the calculated coordinate position of the crack center; and represent the abscissa and ordinate positions of the crack in the image coordinate system; represents the feature weight function in the horizontal and vertical directions of the k-th feature fusion layer under the scale parameter, reflecting the dynamics of the feature change with the scale; represents the local features in the horizontal and vertical directions extracted by the k-th attention mechanism layer under the scale parameter; represents the global image feature energy extracted by the k-th residual network layer under the scale parameter; represents the feature contributions of the fused multi-scale image detail features in the horizontal and vertical coordinate directions; K represents the total number of network layers involved in feature fusion in the CNN network; J represents the number of scales involved in cross-scale feature fusion; represents the scale parameter, with a value range of which represents the continuous fusion of multi-scale features; represents a very small positive constant to avoid the denominator being zero during the calculation process. The formula uses an integral form to achieve the non-linear dynamic fusion of multi-scale features, and the integral operation reflects the feature integration ability of the model in the continuous scale space.

[0028] Finally, the pixel coordinate position of the crack image is accurately calculated through the above formula , and the cloud server further maps and converts it into the actual road geographical space coordinate position. At the same time, after integrating the crack type classification result and the position coordinate data, they are transmitted to the next step for further fusing with the road condition index to form a real-time comprehensive road quality evaluation report, providing effective and real-time road maintenance warnings and repair suggestions for the municipal management platform.

[0029] S50. Generate a real-time comprehensive road quality evaluation report by integrating the data fusion calculation results of step S30 and the image analysis calculation results of step S40; After the cloud server completes the calculation of the road condition index and the analysis and calculation of the road surface crack type recognition and position parameters, it further conducts comprehensive data fusion analysis on the results obtained from the above two steps to generate a real-time, accurate, and visual comprehensive road quality evaluation report.

[0030] In specific implementation, the cloud server first normalizes the obtained road condition index data and defect data such as crack types and locations to unify the data dimensions and ensure effective comparability and unity between the two types of data.

[0031] Subsequently, the cloud server deeply analyzes the road condition index to clarify the overall road state reflected by the index, including comprehensive information such as road bearing capacity, flatness, and structural integrity. At the same time, the cloud server further deeply analyzes the crack type and location data, details the quantity, density, severity, spatial distribution characteristics, and specific geographical location information of each type of crack, and establishes a crack defect database to form a complete defect information map.

[0032] Next, the cloud server uses the big data correlation analysis method to establish the internal correlation between the overall road condition and local crack defects, specifically including analyzing the relationship between the spatial aggregation of crack defects and the overall road health condition, the degree of difference in the impact of different types of cracks on the overall road safety and service performance, and the dynamic relationship between the spatial location distribution characteristics of cracks and the road condition index.

[0033] Furthermore, the cloud server deeply integrates the comprehensive analysis results to generate a comprehensive road quality evaluation index. The specific calculation formula is as follows: , where CRQI represents the comprehensive road quality evaluation index, and the value range is (0, 1). The closer the index is to 1, the better the overall road condition; DSI represents the road condition index, representing the overall macroscopic state of the road; M represents the total number of crack defects detected on the road surface; represents the type weight coefficient of the i-th crack. Different types of cracks have different degrees of influence on road quality, so the weight coefficient is dynamically adjusted; represents the length parameter of the i-th crack, which is used to quantitatively evaluate the spatial scale of the crack distribution on the road surface; represents the depth parameter of the i-th crack, reflecting the severity of the crack; respectively represent the longitude and latitude spatial coordinates of the location of the i-th crack, which are used to reflect the specific spatial location of the crack on the road; represents the location density parameter of the i-th crack on the road surface; is the road crack sensitivity adjustment coefficient, which is used to control the sensitivity of the impact of crack defects on the overall road quality. This index comprehensively considers the mutual influence between the overall road health condition index and the distribution of road crack defects, and dynamically adjusts the sensitivity of the overall index to cracks of different types, different locations, and different severities through a specific algorithm, so that the comprehensive evaluation index can accurately reflect the real road quality condition.

[0034] Finally, based on the above comprehensive analysis, the cloud server automatically generates a real-time comprehensive evaluation report on road quality. The report details the overall road health status index, crack type statistics, crack defect distribution map, typical crack defect pictures, specific latitude and longitude coordinates of the crack defect locations, as well as a clear evaluation of the current overall quality of the road in a combination of text and graphics. The report content is clearly laid out, with intuitive information display, making it easy for the municipal road management department to directly grasp the current specific conditions of the road.

[0035] The above real-time comprehensive evaluation report on road quality generated by the cloud server is automatically sent to the next step after completion for further generating maintenance warnings and repair suggestions to accurately guide and optimize the decision-making of municipal road maintenance.

[0036] S60. Automatically push the comprehensive evaluation report on road quality generated in step S50 to the municipal management platform and automatically trigger the generation of corresponding road maintenance warnings and repair suggestions.

[0037] After completing the generation of the comprehensive evaluation report on road quality, the cloud server further automatically executes the push of the report and the automatic generation of maintenance warnings and repair suggestions to achieve the intelligentization and high efficiency of municipal road quality management.

[0038] In the specific implementation process, the cloud server first establishes a secure and stable data communication link with the municipal management platform based on the comprehensive evaluation report on road quality. Using the preset communication protocol and data interface, the cloud server automatically and real-time pushes the real-time comprehensive evaluation report on road quality to the municipal management platform.

[0039] After receiving the report, the municipal management platform will parse and display the report content in real time, including the comprehensive evaluation index of road quality (CRQI), detailed classification of crack types, specific coordinates of crack locations, distribution map of defect areas, and the evaluation conclusion of the overall road condition. The report is clearly and intuitively displayed in the user terminal of the municipal management platform in a combination of text and graphics, ensuring that managers can quickly, comprehensively, and accurately grasp the current specific conditions of the road.

[0040] At the same time, the cloud server automatically executes the road maintenance warning mechanism based on the CRQI index calculated in the comprehensive evaluation report on road quality and the specific characteristics of crack defects. Specifically, the cloud server has a built-in set of road quality warning rules and threshold systems. When parameters such as the comprehensive evaluation index of road quality or the severity of cracks exceed the preset thresholds, it automatically triggers a road maintenance warning and automatically generates road maintenance warning information according to the warning level classification (such as minor, medium, severe, and urgent).

[0041] In addition, based on the comprehensive evaluation results of defect types, defect severity, and road quality, the cloud server automatically matches and invokes the built-in road maintenance suggestion rule base, and generates a road maintenance suggestion plan through intelligent reasoning. The content includes specific maintenance measures (such as surface crack filling, pavement resurfacing, and structural strengthening), the urgency of maintenance, and the best time for recommended maintenance implementation, etc.

[0042] The generated road maintenance warning and maintenance suggestion information are then automatically pushed to the municipal management platform, clearly presented on the platform interface, and sent to the terminals of municipal road management personnel in a timely manner, so that relevant management personnel can receive the road maintenance warning and maintenance suggestions in a timely manner and make maintenance decisions quickly.

[0043] Embodiment 2 Embodiment 2 of the present application provides an Internet of Things-based municipal road quality diagnosis system, including: An intelligent sensor network for real-time collection of road vibration, pressure, temperature, humidity, and crack image data; The intelligent sensor network includes multiple intelligent sensor nodes, which are distributed in different sections of the municipal road to be monitored. Each intelligent sensor node is built with multiple types of sensors, specifically including vibration sensors, pressure sensors, temperature sensors, humidity sensors, and image sensors, as well as units such as wireless data transmission modules, edge computing modules, and power supply modules supporting them.

[0044] Specifically, the vibration sensor is used to detect the vibration information of the road surface and structure in real time. The vibration information includes the vibration amplitude, vibration frequency, vibration duration, and vibration wave propagation path when a vehicle passes, etc., to reflect the road structure state and potential damage; The pressure sensor is used to collect the load information borne by the road in real time, including the pressure magnitude, change law, and distribution of the road surface when a vehicle passes, so as to evaluate the bearing capacity of the road surface and the stability of the roadbed structure; The temperature and humidity sensors are used to sense the temperature and humidity changes of the road surface and the surrounding environment of the roadbed in real time, providing an environmental basis for diagnosing problems such as the aging of road materials, crack generation, and expansion; The image sensor uses a high-resolution camera to automatically capture and collect road surface crack image data regularly or according to trigger conditions such as vehicle passing and abnormal road vibration, and capture abnormal conditions such as cracks, potholes, and deformations on the road surface.

[0045] All sensor nodes in the intelligent sensor network are integrated with wireless data transmission modules (including but not limited to wireless transmission methods such as NB-IoT, LoRa, and 5G communication) to transmit the collected vibration, pressure, temperature, humidity, and crack image data to the data processing module of the cloud server in a real-time and stable manner.

[0046] In addition, an edge computing module is integrated inside each intelligent sensor node to perform preliminary processing on the collected raw data, including data denoising, preliminary feature extraction of images, etc., so as to reduce data redundancy, improve data quality and transmission efficiency. Each sensor node is equipped with a power supply module, which supports a power supply method combining solar energy and batteries to ensure the long-term stable operation of the system under different environmental conditions.

[0047] The intelligent sensor network as a whole is deployed in a distributed manner, specifically using a hexagonal honeycomb grid method to ensure that there are no blind spots in the sensor coverage area and the information collection is full and comprehensive. At the same time, the network nodes use time synchronization technology (such as the PTP protocol) to mark data synchronization to ensure the consistency of the collected data in the time dimension, which is convenient for subsequent modules to perform accurate data fusion and comprehensive analysis.

[0048] A data processing module, which is used to perform fusion processing on the collected multi-dimensional data and calculate the road condition index; The data processing module is communicatively connected to the intelligent sensor network, and is used to receive in real time multi-dimensional data such as road vibration, pressure, temperature, humidity and crack images collected by the intelligent sensor network, and perform preprocessing, cleaning and fusion analysis on the received data, and then accurately calculate the road condition index, providing important data support for subsequent road quality diagnosis.

[0049] In the specific implementation process, the data processing module first receives and stores the raw data transmitted in real time from the intelligent sensor network, and constructs an efficient data management system in the cloud server to ensure the real-time and integrity of the data.

[0050] The data processing module further performs data preprocessing on the received raw data, including links such as unified conversion of data formats, data normalization, identification and processing of data outliers, etc., to ensure data quality and consistency. Among them, the unified conversion of data formats is used to standardize the formats of different sensor data, making subsequent data fusion analysis more convenient; the data normalization process converts data such as vibration, pressure, temperature, humidity and crack image features to a unified numerical interval to eliminate the differences between data dimensions; the identification and processing of data outliers uses statistical methods and sliding window algorithms to automatically identify and remove abnormal data or perform interpolation repair, improving data quality and reducing noise interference.

[0051] After the preprocessing is completed, the data processing module further performs data fusion analysis. The data fusion analysis uses a self-designed multi-source heterogeneous data fusion calculation method, including a dynamic weight allocation mechanism and a multi-dimensional data non-linear fusion algorithm, comprehensively considering the internal relationships between various types of data to fully reflect the weights and correlation characteristics of the influence of various types of sensor data on the road state.

[0052] Based on data fusion analysis, the data processing module accurately calculates the Road Condition Index (DSI). Specifically, the calculation of the Road Condition Index integrates data from multiple dimensions, such as the amplitude, frequency, and pressure distribution of road vibrations (reflecting road load capacity and structural stability), temperature and humidity data (reflecting material aging and environmental impacts), and crack feature information from preliminary image analysis (reflecting surface damage). The calculated result of the Road Condition Index DSI is a value between 0 and 1. The higher the index value, the better the overall road condition; the lower the index value, the worse the overall road condition and the more urgent the need for maintenance.

[0053] The data processing module transmits the calculated Road Condition Index DSI to the diagnostic integration module in real time for further comprehensive analysis with the detailed crack type and location data provided by the image analysis module, ultimately forming a comprehensive road quality evaluation report to provide accurate and efficient data support for road quality diagnosis.

[0054] The image analysis module is used to automatically analyze crack image data and identify the types and locations of road defects; The image analysis module is communicatively connected to the data processing module and receives the road crack image data collected by the intelligent sensor network and preprocessed by the data processing module, specifically for automatically analyzing the crack image data to achieve accurate classification of road surface defect types and precise calculation of location parameters.

[0055] In specific implementation, the image analysis module first performs further image feature enhancement and refinement on the received crack image data, including denoising filtering, contrast adaptive enhancement, edge feature sharpening, and crack region feature extraction of the crack image to obtain high-quality crack images and ensure the accuracy and stability of subsequent analysis.

[0056] Subsequently, the image analysis module constructs and deploys a self-designed Convolutional Neural Network (CNN) deep learning model to automatically identify road crack types and precisely calculate crack location parameters. The structure of the CNN deep learning model mainly includes an input layer, a multi-scale convolutional feature extraction layer, a deep residual network layer, an attention mechanism layer, a feature fusion layer, and an output layer. Among them: the input layer is used to receive the preprocessed high-quality crack images; the multi-scale convolutional feature extraction layer automatically captures the spatial detail features of cracks using convolutional kernels of different scales; the deep residual network layer is used to enhance the deep feature extraction ability of the model, avoid gradient disappearance, and improve the crack feature learning efficiency; the attention mechanism layer automatically highlights the crack feature regions, reduces background interference, and improves the accuracy of crack region detection; the feature fusion layer comprehensively fuses crack features from different scales and different depth network layers; the output layer automatically completes the classification and identification of crack types and the precise calculation of crack location parameters based on the fused feature data.

[0057] After being trained with a large amount of labeled training data, the CNN deep learning model has the ability to accurately identify road crack types autonomously. The specific crack types include, but are not limited to, linear cracks, reticular cracks, block cracks, and crisscross cracks. At the same time, the CNN model can also automatically and accurately calculate the specific position coordinates of the cracks in the image and then convert them into actual road geospatial coordinates.

[0058] The results of crack type classification are represented by standard classification labels and automatically generate crack classification statistical information. The crack position parameters are accurately output in the form of specific longitude and latitude coordinates and the relative position in the road space to meet the refined requirements of subsequent road maintenance and repair work.

[0059] The crack type and position parameter data automatically generated by the image analysis module are transmitted to the diagnosis integration module in real time to be comprehensively processed with the road condition index calculated by the data processing module, ultimately providing accurate and comprehensive visual defect data support for generating a real-time comprehensive road quality evaluation report.

[0060] The diagnosis integration module is used to integrate the results of the data processing module and the image analysis module to generate a real-time comprehensive road quality evaluation report. The diagnosis integration module is communicatively connected to the data processing module and the image analysis module respectively, and is used to receive in real time the road condition index (DSI) calculated by the data processing module and the crack type and position parameters automatically analyzed by the image analysis module, and further conduct in-depth fusion analysis on the above two types of data to generate a real-time, accurate and comprehensive road quality evaluation report.

[0061] In specific implementation, the diagnosis integration module first integrates and uniformly processes the road condition index (DSI) received from the data processing module and the crack type and position data received from the image analysis module to ensure that the two types of data have a unified dimension and standard format before comprehensive calculation, so as to improve the accuracy of fusion calculation.

[0062] Subsequently, based on the self-designed comprehensive data analysis model, the diagnosis integration module systematically evaluates the correlation characteristics between the overall road health condition index and the crack type, quantity, density, severity, and spatial position distribution. Specific analyses include: analysis of the correlation between the road condition index and crack density, length, and depth; analysis of the differences in the impact of different crack types on the overall road state; analysis of the relationship between the spatial distribution pattern of cracks and the overall road health condition to identify key road risk areas.

[0063] Through the above comprehensive correlation analysis, the diagnostic comprehensive module automatically calculates and generates a comprehensive road quality evaluation index (CRQI), which intuitively reflects the current overall road quality with a unified and clear quantitative index. The calculation of the comprehensive index comprehensively considers the overall bearing capacity of the road, structural stability, and the specific impact of crack defects on road safety, forming a stable and accurate quantitative evaluation result.

[0064] After completing the comprehensive data analysis, the diagnostic comprehensive module automatically generates a real-time comprehensive road quality evaluation report. This report specifically includes, but is not limited to, the following content: the comprehensive road quality evaluation index (CRQI), clearly indicating the current overall road quality; the specific value of the road condition index (DSI) and the description of the overall road health status; the detailed classification information of road crack types, including linear, reticular, block-shaped, and crisscross cracks, etc.; the statistical data of the specific quantity, density, length, and severity of crack defects; the spatial distribution map of crack defects, clearly showing the specific locations (latitude and longitude) of each crack; high-definition image examples of typical crack defect areas, intuitively showing the specific damage conditions of the road; the overall road evaluation conclusion, putting forward targeted road maintenance priority suggestions.

[0065] The above real-time comprehensive road quality evaluation report generated by the diagnostic comprehensive module is transmitted to the warning module in a combination of text and graphics, intuitive and easy-to-understand manner in real time, to further automatically trigger the generation of road maintenance warnings and repair suggestions, guiding the municipal road management department to efficiently carry out maintenance and conservation work.

[0066] The warning module is used to push the generated comprehensive road quality evaluation report to the municipal management platform and automatically generate maintenance warnings and repair suggestions.

[0067] The warning module is communicatively connected to the diagnostic comprehensive module, and is used to receive in real time the comprehensive road quality evaluation report automatically generated by the diagnostic comprehensive module, and automatically execute the push of the report and the generation of road maintenance warnings and repair suggestions, so as to realize the intelligentization, automation, and precision of municipal road management.

[0068] In the specific implementation process, the warning module first establishes a safe and stable data communication channel with the municipal management platform, adopts a standard data transmission protocol, and automatically pushes the generated comprehensive road quality evaluation report to the municipal management platform in real time through an encrypted communication method, ensuring that the management department can obtain the latest road condition information in a timely and intuitive manner.

[0069] After receiving the report, the municipal management platform will display the report content in real time, including the comprehensive road quality evaluation index (CRQI), the road condition index (DSI), the classification statistics of crack types, the detailed information of defect locations, the high-definition images of defect areas, and the overall evaluation conclusion, etc., enabling the management personnel to intuitively and quickly grasp the actual road conditions.

[0070] Subsequently, based on the built-in road quality maintenance warning rule library, the warning module automatically analyzes the comprehensive road evaluation indicators and specific crack defect information provided in the evaluation report. When the comprehensive evaluation indicator or the severity of the defect exceeds the system's preset threshold or triggering condition, the warning module automatically triggers the maintenance warning mechanism and automatically generates road maintenance warning information of the corresponding level (such as minor, medium, severe, urgent, etc.) according to the preset level, and pushes it to the municipal management platform and the terminals of relevant management personnel in real time to remind the management personnel to pay attention to the potential risks of the road in a timely manner.

[0071] In addition, the warning module further automatically matches the built-in maintenance suggestion rule library according to the specific defect characteristics, severity and location parameters, and automatically generates a scientific and reasonable road maintenance suggestion plan through an intelligent reasoning algorithm. The maintenance suggestion plan specifically includes, but is not limited to: clear suggestions for specific maintenance measures, such as crack filling, pothole repair, local road repair, road resurfacing or road structure strengthening, etc.; division of maintenance priority levels and urgency levels, such as priority ranking of recommended maintenance time; description of the recommended best time window and resource requirements for specific maintenance implementation; provision of maintenance historical data of the corresponding area as a reference to improve the scientific nature of maintenance decisions.

[0072] The generated maintenance warning information and maintenance suggestion plan are automatically pushed to the municipal management platform and presented on the platform interface in real time to notify the road maintenance management personnel in a timely manner to ensure that road defects can be quickly responded to and properly handled.

[0073] Corresponding to the above embodiment, an embodiment of the present invention provides a computer storage medium, including: at least one memory and at least one processor; The memory is used to store one or more program instructions; The processor is used to run one or more program instructions to execute an Internet of Things-based municipal road quality diagnosis method.

[0074] Corresponding to the above embodiment, an embodiment of the present invention provides a computer-readable storage medium. The computer storage medium contains one or more program instructions, and the one or more program instructions are used to be executed by a processor to execute an Internet of Things-based municipal road quality diagnosis method.

[0075] The disclosed embodiment of the present invention provides a computer-readable storage medium. Computer program instructions are stored in the computer-readable storage medium. When the computer program instructions run on a computer, the computer is enabled to execute the above-mentioned Internet of Things-based municipal road quality diagnosis method.

[0076] In an embodiment of the present invention, the processor may be an integrated circuit chip with signal processing capabilities. The processor may be a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0077] It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present invention may be directly embodied as being executed by a hardware decoding processor, or may be executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The processor reads the information in the storage medium and combines its hardware to complete the steps of the above method.

[0078] The storage medium may be a memory, for example, it may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories.

[0079] Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory.

[0080] The volatile memory may be a Random Access Memory (RAM) which serves as an external cache. By way of example but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).

[0081] The storage media described in the embodiments of the present invention are intended to include, but not be limited to, these and any other suitable types of memory.

[0082] Those skilled in the art should be aware that, in one or more of the above examples, the functions described in the present invention can be implemented by a combination of hardware and software. When applying software, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. The computer-readable medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage media can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0083] The specific embodiments described above further elaborate on the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the present invention should be included within the protection scope of the present invention.

Claims

1. A municipal road quality diagnosis method based on the Internet of Things, characterized in that: The following steps are involved: S10, intelligent sensor network collects multi-dimensional data including municipal road surface perception data and crack images in real time; S20, the intelligent sensor network transmits the collected municipal road surface perception data to the cloud server; S30, the cloud server calculates the road condition index based on the municipal road surface perception data; S40, the cloud server automatically analyzes the crack image using a deep learning model to calculate the type and location parameters of the cracks on the road surface; S50, combining the road condition index and the type and location parameters of road surface cracks to generate a real-time comprehensive evaluation report on road quality; S60: Automatically push the generated comprehensive evaluation report on road quality to the municipal management platform, and automatically trigger the generation of corresponding road maintenance warnings and repair suggestions.

2. The municipal road quality diagnosis method based on the Internet of Things according to claim 1 is characterized in that: The deployment process of smart sensor networks includes: Vibration sensors are deployed on the road surface and underground structures to collect real-time data on the vibration amplitude, frequency and propagation path of the road; Pressure sensors are placed on the road surface or in the roadbed structure to detect pressure changes and load distribution in real time when vehicles pass by; Temperature sensors and humidity sensors are installed on the road surface and underground structures to collect real-time changes in ambient temperature and humidity; High-definition image sensors are placed on the road surface to automatically capture images of road cracks, potholes and surface irregularities at preset time intervals or trigger conditions.

3. The municipal road quality diagnosis method based on the Internet of Things according to claim 1 is characterized in that: The data transfer process includes: Intelligent sensor nodes transmit collected data to cloud servers in real time and stably through the IoT communication network; During the transmission process, a time synchronization protocol is used to uniformly timestamp different sensor data; Sensor nodes are arranged in a hexagonal grid to ensure the integrity and comprehensive coverage of road data collection.

4. The municipal road quality diagnosis method based on the Internet of Things according to claim 1 is characterized in that: The data fusion calculation process includes: Perform unified data format conversion on the collected raw data; The transformed data were normalized using the Z-score normalization method; Use sliding window anomaly detection algorithm to automatically identify and remove outliers in the data; After removing the abnormal data, multi-source data is fused through dynamic weight factors to calculate the road condition index.

5. The municipal road quality diagnosis method based on the Internet of Things according to claim 1 is characterized in that: The automatic crack image analysis process includes: Image enhancement preprocessing of crack image data including denoising, contrast enhancement and edge sharpening; Automatically analyze crack images using a deep learning model that integrates multi-scale convolutional networks, deep residual networks, and attention mechanisms; Based on the trained deep learning model, the crack type classification and location coordinates are automatically calculated, and the image coordinates are converted into actual road geographic coordinates.

6. The municipal road quality diagnosis method based on the Internet of Things according to claim 1 is characterized in that: Real-time road quality comprehensive evaluation report generation includes: Use big data correlation analysis methods to analyze the relationship between the overall road condition and crack defects; The comprehensive evaluation index of road quality is calculated based on the results of correlation analysis, and a comprehensive evaluation report including defect spatial distribution map, defect type, severity and overall evaluation is generated.

7. The municipal road quality diagnosis method based on the Internet of Things according to claim 1 is characterized in that: The maintenance warning and repair suggestion automatic triggering process includes: Set maintenance warning levels and alarm thresholds based on the evaluation indicators in the comprehensive road quality evaluation report; When the evaluation index reaches or exceeds the preset alarm threshold, the corresponding level of maintenance warning and repair suggestion plan will be automatically generated; The generated maintenance warnings and repair suggestions are automatically pushed to the municipal management platform and the terminal devices of relevant managers in real time.

8. A municipal road quality diagnosis system based on the Internet of Things, characterized in that: include: A smart sensor network for real-time collection of road vibration, pressure, temperature, humidity and crack image data; A data processing module for calculating a road condition index based on municipal road surface sensing data; Image analysis module, used to automatically analyze crack image data and identify road defect types and locations; The diagnostic synthesis module is used to integrate the results of the data processing module and the image analysis module to generate a real-time comprehensive evaluation report on road quality; The early warning module is used to push the generated comprehensive evaluation report on road quality to the municipal management platform and automatically generate maintenance warnings and repair suggestions.

Images