A road and bridge pavement environment monitoring and early warning method and system

Abstract

The present invention relates to the field of environmental monitoring and early warning technology, and in particular to a method and system for monitoring and early warning of road and bridge pavement environment; the method includes deploying a collection device for collecting pavement environmental data, the pavement environmental data includes pavement ice and snow status data; and pre-processing the collected pavement environmental data; training an environmental monitoring and early warning model based on historical pavement environmental data to obtain a trained environmental monitoring and early warning model, and inputting the pre-processed environmental data into the environmental monitoring and early warning model for monitoring and early warning; comparing the results of the monitoring and early warning analysis with a preset threshold, and confirming whether an early warning needs to be issued based on the comparison result; and when ice and snow are detected on the pavement, marking the location and size in a visualization device. The system includes a data collection module, a model training module, and an early warning judgment module. The present invention forms a comprehensive and efficient road and bridge pavement environmental monitoring and early warning system.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring and early warning, and in particular to a method and system for monitoring and early warning of road and bridge pavement environments. Background Art

[0002] Environmental monitoring refers to the activities of environmental monitoring agencies to monitor and measure the environmental quality status. Environmental monitoring is to determine the environmental pollution status and the level of environmental quality by monitoring and measuring indicators that reflect environmental quality. The content of environmental monitoring mainly includes the monitoring of physical indicators, chemical indicators and ecosystems. Road environment monitoring not only focuses on air quality, but also involves traffic flow, road network travel conditions, etc., to achieve real-time calculation of traffic pollution emissions in key urban road networks, and to explore the intrinsic correlation between traffic pollution and air quality monitoring stations.

[0003] Prior art 1, application number CN202110950055.9 discloses a road environment monitoring and cleaning system, including: a road environment monitoring module, a control module, a cleaning module, a host computer module, and an alarm module; the control module is connected to the road environment monitoring module, the cleaning module, the host computer module and the alarm module respectively; the road environment monitoring module is used to collect road environment data and convert the road environment data into monitoring data; the control module is used to perform threshold processing on the monitoring data according to the parameter threshold, generate cleaning instructions and alarm instructions based on the threshold processing results or the host computer instructions, and also to transmit the monitoring data to the host computer module; the cleaning module cleans the road environment based on the cleaning instructions; the host computer module is used to monitor the monitoring data, save and obtain the host computer instructions; the alarm module alarms based on the alarm instructions. Although the road cleaning and alarm monitoring in the wireless transmission mode are realized, the road environment status can be monitored in real time. However, there is a lack of early warning of possible environmental problems.

[0004] Prior art 2, application number CN202110516395.0 discloses a road surface environment monitoring and early warning system, including a front-end detection and early warning device unit, a network transmission unit, and a network application terminal. The front-end detection and early warning device includes an outdoor LED screen display early warning unit, an environmental detection unit, and a video monitoring unit. The network transmission unit includes an optical fiber transceiver, a road junction machine, and a network application terminal includes a PC client and a mobile client. Although the environment detection unit can detect the environment and vehicle condition of the highway in real time, and the network transmission unit quickly transmits the collected data information to the network application terminal through the road junction machine for data display, the network application terminal monitors the environmental information and vehicle condition information of the highway in real time, and performs early warning analysis on the collected information, which reduces the occurrence of traffic accidents. However, the required components are complex in structure, the construction cost is high, and the maintenance cost is high when built on the highway.

[0005] Prior art three, application number CN202110467109.6 discloses a highway pavement monitoring device, including: a top cover and a bottom shell, the top cover and the bottom shell are assembled by means of a connector, a temperature monitoring component is installed in the bottom shell, and the input end of the temperature monitoring component passes through the top cover and contacts the external environment. Although it is possible to obtain data on road temperature changes in a timely manner, detect the temperature of the road surface in a timely manner, and effectively provide early warning and prevention to ensure the safety of highway driving, it is beneficial to improve the safety of road traffic. However, it can only measure the road surface temperature data, and direct contact with the road surface is easily damaged by passing vehicles, which reduces the service life.

[0006] At present, the existing technologies 1, 2 and 3 have the problems of small scope of environmental monitoring, less measured data, lack of early warning of possible environmental problems, and high construction and maintenance costs. Therefore, the present invention proposes a road and bridge pavement environmental monitoring and early warning method and system to solve the problems in the existing technologies. Summary of the invention

[0007] The main purpose of the present invention is to provide a road and bridge pavement environmental monitoring and early warning method and system to solve the problems in the prior art of small environmental monitoring scope, less measured data, lack of early warning of possible environmental problems, and high construction and maintenance costs.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A road and bridge pavement environment monitoring and early warning method, the road and bridge pavement environment monitoring and early warning method comprising:

[0010] Arrange collection equipment for collecting road environment data, the road environment data including road ice and snow state data; and pre-process the collected road environment data to obtain pre-processed road environment data;

[0011] The environmental monitoring and early warning model is trained based on the historical environmental data of the road surface to obtain the trained environmental monitoring and early warning model, and the pre-processed environmental data is input into the environmental monitoring and early warning model for monitoring and early warning to obtain the results of monitoring and early warning analysis;

[0012] The results of the monitoring and early warning analysis are compared with the preset thresholds, and the need to issue an early warning is determined based on the comparison results; at the same time, when ice and snow are detected on the road, their location and size are marked in the visualization device.

[0013] As a further improvement of the present invention, obtaining the pre-processed road surface environment data includes:

[0014] Obtain the road surface to be monitored, determine the target area to be measured according to the length and width of the road surface, and determine the measuring distance of the remote sensing road condition sensor. Install the remote sensing road condition sensor, point the lens directly at the road surface area to be measured, and collect road condition information.

[0015] The infrared transmitter of the remote sensing road condition sensor is aimed at the selected position on the road surface, and emits a spectral signal. The spectral signal is reflected from the measured location back to the receiver. The remote sensing road condition sensor can distinguish the reflection of specific wavelengths caused by ice and snow. The reflection characteristics of ice layer and water film to spectral signals are different, so as to judge the ice and snow status of the road surface. At the same time, it also receives the infrared energy radiated by the road surface, and then obtains the road surface temperature value.

[0016] The remote sensing road condition sensor analyzes the received spectral signals to obtain real-time data on road surface temperature and road surface status, denoises the real-time data, and outputs the results to the environmental monitoring and early warning model and visualization equipment.

[0017] As a further improvement of the present invention, the judgment of the ice and snow state of the road surface is realized, including:

[0018] Convert the time domain signal of the spectral signal to the frequency domain and identify the frequency components in the spectral signal; extract the characteristic wavelength related to the state of ice and snow by analyzing the spectral characteristics of the spectral signal; compare the extracted characteristic wavelength with a known standard spectral library, which contains the reflection characteristics of different materials;

[0019] A classification model is constructed using a decision tree. By training the intensity of some extracted characteristic wavelengths, a mapping relationship between the road conditions of ice and snow, dry road surface, and water film is established. The received wavelength data is input into the trained model classification type to determine the current road condition, including ice and snow coverage, dry, or water film.

[0020] Real-time road surface temperature data is obtained from remote sensing road condition sensors; the current measured road surface temperature is below 0 degrees, indicating that there may be ice on the road surface; a series of rules are set based on wavelength discrimination and temperature status.

[0021] As a further improvement of the present invention, a series of rules include:

[0022] If the wavelength characteristics show ice and the temperature is ≤ 0 degrees, the road surface is judged to be icy;

[0023] If the temperature is between 0 ≤ temperature ≤ 5 degrees, and the spectral data supports the presence of a water film, it is judged to be in a water film state;

[0024] If the spectral data matches the characteristic wavelength of a dry road surface, it is judged to be in a dry state.

[0025] As a further improvement of the present invention, a classification model is constructed using a decision tree, including:

[0026] Collect data related to road conditions, clean the data, and select characteristic wavelengths related to ice and snow, dry road conditions, and water film conditions;

[0027] The data set is divided into a training set and a test set; the root node of the decision tree corresponds to all input samples, and the information gain of each feature is calculated to select the optimal feature for division until the termination condition is met and the depth of the decision tree reaches the maximum limit; each path, from the root node to the leaf node, represents a set of decision rules, and the final leaf node represents the classification result of the sample;

[0028] The classification model is cross-validated to evaluate its accuracy and stability. The evaluation indicators include accuracy, recall rate and F1-score. The newly received spectral data is input into the trained decision tree model. The decision tree model judges the current road condition according to the newly input characteristic wavelength and outputs the prediction result, "ice and snow covered", "dry" or "water film state".

[0029] As a further improvement of the present invention, the environmental monitoring early warning model is constructed, including:

[0030] We chose to use an artificial neural network environmental monitoring and early warning model to conduct in-depth analysis of the preprocessed data and extract features that are useful for environmental monitoring and early warning;

[0031] Use historical road environmental data to train the environmental monitoring and early warning model, adjust the parameters of the environmental monitoring and early warning model so that the environmental monitoring and early warning model reflects the mapping relationship between parameter changes and environmental changes, and use the verification data set to verify and evaluate the trained environmental monitoring and early warning model;

[0032] Analyze the preprocessed data to monitor the current road environment conditions. When the monitored data changes, analyze the changed data to predict upcoming environmental problems and notify staff through visualization equipment or alarms.

[0033] As a further improvement of the present invention, the environmental monitoring and early warning model reflects the mapping relationship between parameter changes and environmental changes, including:

[0034] Input the training data into the environmental monitoring and early warning model, calculate the predicted value through forward propagation, and calculate the gradient according to the loss function; use the optimization algorithm to update the weights and biases of the network; repeat until the error of the environmental monitoring and early warning on the training set no longer decreases significantly or reaches the preset maximum number of training iterations;

[0035] After the training is completed, the validation data set is input into the trained environmental monitoring and early warning model to calculate the performance of the environmental monitoring and early warning model on the validation set. The key indicators include: prediction accuracy, mean absolute error or mean square error;

[0036] Use the test set involved in training or validation to evaluate the environmental monitoring and early warning model to ensure its predictive effect on new data; use feature importance analysis to evaluate the performance of the environmental monitoring and early warning model and explain the output results.

[0037] As a further improvement of the present invention, the monitoring and early warning analysis result is compared with a preset threshold value, including:

[0038] Obtain monitoring and early warning analysis results from the environmental monitoring and early warning model, including road ice and snow conditions and temperature;

[0039] By analyzing and extracting historical environmental monitoring data and accidents and slips, we can determine certain key indicators, and ice and snow may appear when the temperature is below a certain value;

[0040] Compare the monitoring and early warning analysis results with the preset thresholds to determine whether to issue an early warning.

[0041] As a further improvement of the present invention, the early warning trigger: if the conditions are met, the early warning state is entered, the state is recorded and the subsequent steps are prepared; when ice or snow is detected on the road surface, the state is marked as early warning, and the corresponding processing flow is executed;

[0042] After confirming whether an early warning is needed, perform the following operations on the visualization device at the same time:

[0043] Marking location and size: On the visual interface, the currently detected road environment problems are displayed through a map or graphically; if ice and snow are detected on a specific road section, the section will be highlighted in the interface and the scope of ice and snow coverage will be indicated.

[0044] To achieve the above object, the present invention also provides the following technical solutions:

[0045] A road and bridge pavement environment monitoring and early warning system, which is applied to the road and bridge pavement environment monitoring and early warning method, and the road and bridge pavement environment monitoring and early warning system comprises:

[0046] The data acquisition module is used to deploy acquisition equipment for collecting road environment data, where the road environment data includes road ice and snow status data; and to pre-process the collected road environment data to obtain pre-processed road environment data;

[0047] The model training module is used to train the environmental monitoring and early warning model based on the historical environmental data of the road surface, obtain the trained environmental monitoring and early warning model, input the pre-processed environmental data into the environmental monitoring and early warning model for monitoring and early warning, and obtain the results of monitoring and early warning analysis;

[0048] The early warning judgment module is used to compare the results of the monitoring and early warning analysis with the preset thresholds, and confirm whether a warning needs to be issued based on the comparison results; at the same time, when ice and snow are detected on the road surface, the location and size are marked in the visualization device.

[0049] The present invention acquires multiple parameter data of the road environment in real time through environmental data acquisition equipment deployed on the road surface. The collected raw data may contain noise, etc., and needs to be preprocessed through data cleaning, denoising and other methods to ensure the quality of data input into the model; the machine learning or deep learning model is trained using historical road environment data, and the preprocessed real-time environmental data is input into the trained model. The model predicts the current and future environmental conditions based on the input data, and the early warning information is displayed to the staff through the visualization device, and the alarm system is cooperated to ensure the timely transmission of information; according to the needs of road safety, reasonable thresholds are set for key parameters such as visibility, road temperature, wind speed, etc., and the real-time collected data is compared with the preset thresholds. For different degrees of early warning, the system adopts different notification methods. For minor abnormalities, it is only displayed on the visualization device. For serious problems, an alarm is used to ensure that the staff can receive the notification and take measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic diagram of the steps of an embodiment of the road and bridge pavement environment monitoring and early warning method of the present invention;

[0051] Figure 2 A schematic flow chart of steps for obtaining pre-processed pavement environment data in one embodiment of a method for monitoring and early warning the pavement environment of roads and bridges according to the present invention;

[0052] Figure 3 A schematic flow chart of the steps of constructing an environmental monitoring and early warning model in an embodiment of the method for environmental monitoring and early warning of roads and bridge pavements of the present invention;

[0053] Figure 4 A schematic flow chart of the steps of comparing the monitoring and early warning analysis results with a preset threshold value in one embodiment of the road and bridge pavement environment monitoring and early warning method of the present invention;

[0054] Figure 5 A schematic diagram of functional modules of an embodiment of a road and bridge pavement environment monitoring and early warning system of the present invention;

[0055] Figure 6It is a structural schematic diagram of an embodiment of an electronic device of the present invention;

[0056] Figure 7 It is a schematic structural diagram of an embodiment of the storage medium of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] The terms "first", "second" and "third" in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present invention are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0059] Reference to an "embodiment" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0060] like Figure 1 As shown, this embodiment provides an embodiment of a road and bridge pavement environment monitoring and early warning method. In this embodiment, the road and bridge pavement environment monitoring and early warning method specifically includes the following steps:

[0061] Step S1: deploying a collection device for collecting road environment data, the road environment data including road ice and snow state data; and preprocessing the collected road environment data to obtain preprocessed road environment data;

[0062] Step S2: training the environmental monitoring and early warning model based on the historical road environment data to obtain a trained environmental monitoring and early warning model, inputting the pre-processed environmental data into the environmental monitoring and early warning model for monitoring and early warning, and obtaining the results of monitoring and early warning analysis;

[0063] Step S3: Compare the results of the monitoring and early warning analysis with the preset threshold value, and confirm whether it is necessary to issue an early warning based on the comparison result; at the same time, when ice and snow are detected on the road surface, the location and size are marked in the visualization device.

[0064] Preferably, in step S1 of this embodiment, data collection and preprocessing are carried out by systematically arranging road environment monitoring equipment (such as sensors, cameras, etc.) to collect real-time data on the ice and snow status of the road surface; these devices can obtain a variety of environmental parameters, including temperature, humidity, precipitation, and real-time road conditions; the collected raw data are cleaned, noise-filtered, and standardized; the preprocessing steps may include operations such as filling missing values, eliminating outliers, and data smoothing to ensure the data quality of subsequent analysis is reliable. Significance achieved: High-quality data collection and preprocessing provide a reliable foundation for environmental monitoring and early warning systems, ensuring the accuracy and effectiveness of subsequent analysis; by integrating a variety of detection equipment, real-time monitoring of road environment status can be achieved, key data can be obtained in a timely manner, and the needs of dynamic analysis can be met; data preprocessing lays the foundation for subsequent model training and analysis, optimizes the analysis process, and improves model performance. Step S2: Environmental monitoring and early warning model training. Based on historical environmental data, the environmental monitoring and early warning model is trained using machine learning, statistical analysis or other modeling methods. The model can learn from historical data and identify patterns that lead to changes in road ice and snow conditions. The latest pre-processed environmental data is input into the trained model, and the model outputs monitoring and early warning analysis results in real time based on the input data, including whether there is a potential risk of ice and snow on the road surface. Significance: Through intelligent technologies such as machine learning, the model can automatically learn and improve, accurately predict future environmental changes, and improve monitoring efficiency. Real-time data input and analysis capabilities can improve the response speed to changes in the road environment, so that relevant departments can respond quickly and ensure traffic safety. A data-based risk management system is established to provide scientific support for road safety and reduce the risk of traffic accidents. Step S3: Early warning result comparison and visualization. The results of the monitoring and early warning analysis are compared with the preset safety threshold to determine whether the early warning mechanism is triggered. The specific location and size of ice and snow on the road surface are marked in real time on the visualization device to ensure that key information is delivered to decision makers and the public in a timely manner. Significance achieved: By comparing with the preset threshold, it helps to determine whether to issue an early warning, making the early warning mechanism more intelligent and accurate; the visualization method enhances the intuitiveness and comprehensibility of information transmission, allowing relevant personnel to quickly grasp the road risk situation and optimize the emergency response process; improve traffic safety: through timely warnings and visual information, it helps the majority of drivers take preventive measures in advance, reduce the incidence of traffic accidents caused by ice and snow on the road, and ensure public safety.

[0065] In summary, the steps of this embodiment form a comprehensive and efficient road and bridge pavement environment monitoring and early warning system. Each step lays a solid foundation for the optimization and enhancement of the system in terms of technical effects, improves the level of road safety management in strategic significance, and provides a scientific basis for decision-making, which is conducive to achieving the goals of intelligent transportation and safe travel.

[0066] Furthermore, if Figure 2 As shown, the process of obtaining the pre-processed road surface environment data in step S1 specifically includes the following steps:

[0067] Step S11: Acquire the road surface to be monitored, determine the target area to be measured according to the length and width of the road surface, and simultaneously determine the measuring distance of the remote sensing road surface condition sensor, install the remote sensing road surface condition sensor, and place the lens directly on the road surface area to be measured to collect road surface condition information;

[0068] Step S12: the infrared transmitter of the remote sensing road condition sensor is aimed at a selected position on the road surface, and a spectrum signal is emitted. The spectrum signal is reflected from the measured location back to the receiver. The remote sensing road condition sensor can distinguish the reflection of a specific wavelength caused by ice and snow. The ice layer and the water film have different reflection characteristics for the spectrum signal, thereby realizing the judgment of the ice and snow state of the road surface. At the same time, the infrared energy radiated by the road surface is received, and the road surface temperature value is obtained.

[0069] Step S13: The remote sensing road condition sensor analyzes the received spectral signal to obtain real-time data of road surface temperature and road surface condition, performs denoising on the real-time data, and outputs the result to the environmental monitoring and early warning model and visualization equipment.

[0070] Preferably, S11 of this embodiment obtains the road surface to be monitored and installs sensors. By measuring the length and width of the road surface to be monitored, the road surface area that needs to be monitored is clarified to ensure the pertinence and effectiveness of the monitoring; a remote sensing road surface condition sensor is installed, and its lens is aimed at the road surface area to be measured to ensure the accuracy and continuity of data collection. The significance achieved: Accurately determining the target area for measurement makes subsequent data collection more efficient and can focus on road sections where ice and snow may exist; through a systematic sensor deployment plan, a sustainable road surface environment monitoring system is established to lay the foundation for subsequent data processing; ensuring that the installed sensors can operate stably and cover the areas that need to be monitored, providing guarantees for obtaining high-quality real-time data. Step S12 spectral signal emission and road surface state judgment, the infrared transmitter of the remote sensing road surface condition sensor is aimed at the road surface, emitting spectral signals and receiving reflected signals, and identifying ice and snow and other road surface conditions by analyzing the reflected wavelength; the remote sensing road surface condition sensor simultaneously receives infrared energy radiated by the road surface and measures the road surface temperature to assist in judging the road surface environment state. Significance achieved: Through the emission of spectral signals and remote sensing technology, rich road surface status information, such as ice and snow status and temperature, can be collected, which provides a scientific basis for accurately judging road conditions; different materials and states (such as ice layer, water film, etc.) have different reflection characteristics for spectral signals, which enables the remote sensing road condition sensor to accurately identify the road surface status and improve the accuracy and reliability of judgment; real-time acquisition of road surface temperature and status data enables the monitoring system to quickly adapt to environmental changes and improve the response speed to traffic safety. Step S13 data analysis and denoising processing, the remote sensing road surface condition sensor analyzes the received spectral signal, monitors the road surface temperature and status in real time, calculates and outputs relevant data; denoises the acquired road surface status data to reduce data errors caused by environmental interference. Significance achieved: Through real-time analysis and data output, the current road surface status information can be quickly provided to provide support for subsequent warnings and decisions; denoising processing can improve the accuracy and reliability of data, help the performance of subsequent analysis models, and reduce the probability of false warnings; the cleaned data can be directly output to the environmental monitoring warning model and visualization equipment to ensure the timely transmission and use of information and optimize the efficiency of the overall monitoring system.

[0071] In summary, the steps of this embodiment are combined to form a comprehensive and efficient road environment data acquisition and preprocessing process. Each step ensures the real-time and reliability of road environment data through precise remote sensing road condition sensor positioning and accurate signal analysis, further improving the effectiveness of the environmental monitoring and early warning system. Such a system means enables traffic management departments to better deal with road ice and snow, improve road safety, and thus ensure the safety of public travel.

[0072] Furthermore, the process of determining the state of ice and snow on the road surface in step S12 specifically includes the following steps:

[0073] Step S121: converting the time domain signal of the spectral signal into the frequency domain, identifying the frequency components in the spectral signal; extracting characteristic wavelengths related to the state of ice and snow by analyzing the spectral features of the spectral signal; comparing the extracted characteristic wavelengths with a known standard spectral library, which contains the reflection characteristics of different materials (such as ice, water, dry road surface, etc.);

[0074] Step S122: construct a classification model using a decision tree, and establish a mapping relationship with road conditions such as ice and snow, dry road surface, and water film by training the intensity of the extracted partial characteristic wavelengths; input the received wavelength data into the trained model classification type to determine the current road condition, including ice and snow coverage, dry or with water film, etc.;

[0075] Step S123: Acquire real-time road surface temperature data from the remote sensing road surface condition sensor; the currently measured road surface temperature is lower than 0 degrees, indicating that there may be ice on the road surface; combine wavelength discrimination and temperature status to set a series of rules, including:

[0076] If the wavelength characteristics show ice and the temperature is ≤ 0 degrees, the road surface is judged to be "icy";

[0077] If the temperature is between 0 ≤ temperature ≤ 5 degrees, and the spectral data supports the presence of a water film, it is judged as a "water film state";

[0078] If the spectral data matches the characteristic wavelength of a dry road surface, it is judged as a "dry state".

[0079] Preferably, in step S121 of this embodiment, the spectral signal analysis and feature extraction converts the time domain signal into the frequency domain signal and identifies the frequency components in the spectral signal; this process can clearly display the frequency characteristics of the signal and help extract important spectral information; by analyzing the characteristics of the spectral signal, the characteristic wavelength related to the ice and snow state is extracted; the characteristic wavelength is a key indicator for distinguishing different road conditions, and its extraction directly affects the accuracy of subsequent discrimination; the extracted characteristic wavelength is compared with a known standard spectral library. The standard spectral library contains the reflection characteristics of materials such as ice, water, and dry road surface, allowing the model to accurately identify specific materials. Significance achieved: Through frequency domain conversion and characteristic wavelength extraction, the system can more accurately capture the relationship between the surface state and the spectral signal, thereby improving the effectiveness of subsequent analysis; the comparison basis provided by the standard spectral library enables the system to analyze the state changes of the road surface in real time, thereby responding to sudden changes in the road surface environment in a timely manner; the comparison process based on the standard spectral library makes the data analysis have a unified standard, which improves the scientificity and reliability of the analysis process. Step S122: Construction of decision tree classification model. Using the intensity of the extracted characteristic wavelength, a classification model is constructed through a decision tree algorithm. This model can identify and classify different road conditions (such as ice and snow, dryness, and water film). The model is trained through existing data sets so that the model can learn the relationship between the characteristics and the road conditions, and improve the accuracy of the judgment. The wavelength data to be classified is input into the trained decision tree model to facilitate real-time judgment of the current road condition. Significance achieved: The decision tree model has a clear classification mechanism, which can quickly find useful information in complex data, greatly improving the ability to judge the road condition in real time. Through continuous training and optimization, the model can better adapt to new environments and different road conditions, thereby improving the flexibility and reliability of the overall system. Through classification output, the model can provide data support for traffic management and help decision makers take corresponding actions, such as issuing safety warnings. Step S123: Acquisition of temperature data and determination of state rules. Real-time road temperature data is obtained from remote sensing road condition sensors to provide basic data for subsequent judgments. A series of judgment rules are set according to temperature and wavelength characteristics, so that the system can comprehensively consider temperature factors and accurately judge the road condition. Significance achieved: Through the integration of temperature and wavelength characteristic information, the system can provide more comprehensive road condition information and improve the accuracy of judgment; combined with temperature data and its threshold, it can quickly respond to changes in the road environment and promptly issue early warning signals for road icing and water film formation to ensure road safety; by accurately judging the road condition, it can effectively reduce the probability of traffic accidents caused by ice, snow or water film, and improve driving safety and road traffic efficiency.

[0080] In summary, the three steps of this embodiment complement each other and together constitute a complete process for judging the state of ice and snow on the road. Through the combination of signal analysis, feature extraction, model construction and real-time temperature data, the system can quickly and accurately identify the state of the road surface, thereby providing scientific and reliable data support for traffic management, ensuring driving safety while improving the efficiency of the social transportation system.

[0081] Furthermore, the process of constructing a classification model using a decision tree in step S122 specifically includes the following steps:

[0082] Step S1221: collecting data related to road conditions, cleaning the data, and selecting characteristic wavelengths related to ice and snow, dry road conditions, and water film conditions;

[0083] Step S1222: Divide the data set into a training set and a test set; the root node of the decision tree corresponds to all input samples, and the information gain of each feature is calculated to select the optimal feature for division until the termination condition is met and the depth of the decision tree reaches the maximum limit; each path, from the root node to the leaf node, represents a set of decision rules, and the final leaf node represents the classification result of the sample;

[0084] Step S1223: Cross-validate the classification model to evaluate its accuracy and stability. Evaluation indicators include accuracy, recall rate, and F1-score. Input the newly received spectral data into the trained decision tree model. The decision tree model determines the current road surface state based on the newly input characteristic wavelength and outputs the prediction result, "ice and snow covered", "dry" or "water film state".

[0085] Preferably, step S1221 of this embodiment, data collection and preprocessing, by effectively collecting data related to the road surface state (such as characteristic wavelengths and corresponding state labels), provides a basis and support for subsequent modeling; removes noise and outliers, processes missing data, and ensures the quality and accuracy of the data. This step ensures the reliability of subsequent analysis; selects characteristic wavelengths most relevant to states such as ice and snow, dry road surface and water film, so that the model focuses on the features that can best distinguish different states, thereby improving the interpretability and accuracy of the model. Significance achieved: Data is the cornerstone of machine learning models, and high-quality data can significantly improve the performance of the model; through feature selection, the complexity of the model is reduced, while also reducing computing time and resource consumption, and enhancing the sensitivity of the model to target classification. Step S1222, data set partitioning and decision tree model construction, divides the data into training sets and test sets to ensure the independence of model training and evaluation. This allows the generalization ability of the model in real-world applications to be evaluated; the best partitioning features are selected by calculating the information gain of the features, and the hierarchical structure of the decision tree is constructed, and the partition of each node in the feature space is further refined until the preset termination condition is met. Significance achieved: The decision tree shows the decision-making process in an intuitive way. The path from the root node to the leaf node can clearly show why a certain state is judged as a specific category; by dividing the training set and the test set, it is ensured that the model not only performs well on the training data, but also maintains good performance when facing new and unseen data. Step S1223 Model verification and real-time application, cross-validation evaluation of the model, multiple training and verification through different division methods, reduce the risk of overfitting, and improve the stability and reliability of the model; calculate indicators such as accuracy, recall rate and F1-score, and comprehensively evaluate the performance of the classification model. This step ensures a clear understanding of the various performance of the model. Significance achieved: Through the accuracy of state judgment (such as ice and snow state), a scientific basis is provided for traffic safety management and decision-making, reducing the risk of accidents caused by unclear road conditions; the model can make state judgments based on the rapidly changing spectral data received in real time, so that traffic management can respond quickly under changing weather conditions.

[0086] In summary, each step of this embodiment plays a vital role in the entire decision tree classification model construction process: step S1221 ensures the collection and processing of high-quality data, making the model's training data more reliable; step S1222 provides the model with clear decision rules, enhancing interpretability and generalization capabilities; step S1223 focuses on the verification and practical application of the model to ensure that the model can operate effectively in a real environment; through the coordination and implementation of these steps, accurate judgment of the road surface condition is ultimately achieved, providing data support and theoretical basis for traffic safety.

[0087] Furthermore, if Figure 3 As shown, the process of constructing the environmental monitoring early warning model in step S2 specifically includes the following steps:

[0088] Step S21: Select and use an artificial neural network environmental monitoring and early warning model to conduct in-depth analysis on the preprocessed data and extract features useful for environmental monitoring and early warning;

[0089] Among them, artificial neural network includes:

[0090] Input layer, used to input preprocessed data;

[0091] Hidden layer: set 2 to 4 hidden layers, and select 50 to 100 neurons in each layer; introduce self-attention mechanism in the hidden layer to achieve better information capture and processing by enhancing the representation ability of input data features; specific implementation steps:

[0092] Input preparation: In the preprocessing stage, the environmental monitoring data are first converted into a form suitable for model input;

[0093] Generate Query, Key, and Value: For the input of each layer (assuming the input is (X), the dimension is (n\timesd), where (n) is the number of samples and (d) is the feature dimension), the self-attention mechanism generates three sets of weight matrices:

[0094] Query matrix ((Q)): multiplied by the weight matrix (W Q ) to obtain: [Q = XW Q ]

[0095] Key matrix ((K)): obtained by multiplying the input (X) by the weight matrix (WK): [K = XW K ]

[0096] Value matrix ((V)): multiplied by the weight matrix (W V ) to obtain: [V = XW V ]

[0097] Calculate the attention weight: The attention score is formed by calculating the dot product between the query and the key, and scaled:

[0098] where d k is the dimension of the Key vector, and Softmax is processed to obtain the attention weight:

[0099] Apply the softmax function to the attention score to get the weight:

[0100] Attention weights = softmax(Attention s core)

[0101] Generate weighted output: Use the attention weight to weight the Value vector and generate the output of the self-attention layer: Output = Attention w eights·V

[0102] The output is a new feature representation dynamically generated based on the relationship of the input data;

[0103] Fusion to hidden layers: Passing the self-attention output as input to subsequent network layers, allowing the neural network to learn and process more complex relational structures;

[0104] Step S22: Use the historical road environment data to train the environmental monitoring and early warning model, adjust the parameters of the environmental monitoring and early warning model, so that the environmental monitoring and early warning model reflects the mapping relationship between parameter changes and environmental changes, and use the verification data set to verify and evaluate the trained environmental monitoring and early warning model;

[0105] Step S23: Analyze the preprocessed data to monitor the current road environment conditions. When the monitored data changes, analyze the changed data to predict the environmental problems that will occur, and notify the staff through visualization equipment or alarms.

[0106] Preferably, in step S21 of this embodiment, feature selection and preprocessing are performed. In this stage, the preprocessed data includes operations such as cleaning, normalization, and dimensionality reduction to improve the quality and availability of the data. The data is processed using an artificial neural network (ANN), and domain knowledge and data features are combined to extract features useful for environmental monitoring and early warning, such as pollutant concentration, temperature and humidity changes, wind speed, etc. The significance achieved is: by selecting relevant features and reducing redundant information, the neural network can focus more on important data features during the training process, which can improve the efficiency and accuracy of model learning; the extracted features provide a more accurate basis for the environmental monitoring model, so that the subsequent early warning judgment can be more timely and accurate, thereby effectively preventing potential environmental problems. Step S22 model training and verification, using historical road environment data to train the artificial neural network, adjusting the model parameters through the back propagation algorithm, so that the model learns the mapping relationship between input features and environmental changes; using the verification data set to test the trained model and evaluate the model performance to ensure the generalization ability and accuracy of the model. Significance achieved: Through the process of training and verification, ensuring that the model can accurately reflect current and future environmental changes is crucial for timely discovery and handling of environmental problems; once the accuracy of the model is guaranteed, timely response decisions can be made based on the output of the model to enhance the scientific nature of environmental management. Step S23 real-time monitoring and early warning, dynamic analysis of pre-processed real-time data, when the monitored environmental data changes, the model quickly feedback, and analyze the changed data; based on the monitored input data, the trained model is used to predict upcoming environmental problems, such as excessive pollution, extreme weather, etc., and generate early warning information; the monitoring results are displayed through a visualization platform, or the relevant staff are notified through the alarm system to ensure rapid response. Significance achieved: This real-time monitoring and early warning mechanism makes environmental management and response mechanisms more timely and effective, and helps reduce the risk of sudden environmental accidents; when environmental problems can be quickly monitored and warned, the public's sense of security and ability to respond to environmental changes will also be improved, which is more conducive to social stability.

[0107] In summary, the entire process of constructing the environmental monitoring and early warning model of this embodiment forms a complete set of effective early warning mechanisms through the steps of feature selection and preprocessing, model training and verification, and real-time monitoring and early warning. Each step plays a key role in the accuracy, reliability and real-time response capability of the model. The ultimate goal is to improve the scientificity and effectiveness of environmental management, so as to better protect the public and the ecological environment. Based on the characteristics of the collected data, an artificial neural network is selected as the early warning model. The artificial neural network has a strong nonlinear mapping ability and self-learning ability, and can simulate the mapping relationship between complex environmental parameters and environmental changes. By constructing a multi-layer neural network, a nonlinear conversion from input parameters to output results can be achieved; the model is trained using historical road environment data to minimize the prediction error. Through multiple iterative training, the mapping relationship between parameter changes and environmental changes can be gradually learned; when the monitored data changes, the system will conduct an in-depth analysis of the changed data to predict possible environmental problems. When potential problems are predicted, the staff will be notified immediately through visualization equipment or alarm systems.

[0108] This embodiment selects an artificial neural network as the early warning model, making full use of its powerful nonlinear mapping ability and sub-learning ability. It can simulate the mapping relationship between complex environmental parameters and environmental changes, and provide high-precision and high-reliability prediction results for environmental monitoring and early warning; the model is trained using historical road environment data so that the model can accurately reflect the mapping relationship between parameter changes and environmental changes, thereby improving the prediction accuracy and generalization ability; through real-time detection and early warning, the human, material and financial resources of environmental protection work can be arranged more reasonably, thereby improving the efficiency and benefits of environmental protection work.

[0109] Furthermore, the process of making the environmental monitoring and early warning model reflect the mapping relationship between parameter changes and environmental changes in step S22 specifically includes the following steps:

[0110] Step S221: input the training data into the environmental monitoring and early warning model, calculate the predicted value through forward propagation, and calculate the gradient according to the loss function; use the optimization algorithm to update the weight and bias of the network; repeat until the error of the environmental monitoring and early warning on the training set no longer decreases significantly or reaches the preset maximum number of training iterations;

[0111] Step S222: After the training is completed, the validation data set is input into the trained environmental monitoring and early warning model, and the performance of the environmental monitoring and early warning model on the validation set is calculated. The key indicators include: prediction accuracy, mean absolute error or mean square error;

[0112] Step S223: Use the test set involved in training or validation to evaluate the environmental monitoring and early warning model to ensure its prediction effect on new data; use feature importance analysis to evaluate the performance of the environmental monitoring and early warning model and explain the output results.

[0113] Preferably, in the training process of step S221 of this embodiment, the input data is processed by a neural network, and is passed from the input layer to the output layer layer by layer to generate a model prediction value; the gradient is calculated according to the difference (loss function) between the predicted value and the true value, and then the network parameters (weights and biases) are adjusted using the selected optimization algorithm (such as SGD, Adam, etc.) to make future predictions closer to the true value; this process is repeated, and through multiple training iterations (epochs), the error is ensured to be gradually reduced. The significance achieved: by continuously adjusting the model parameters, the environmental monitoring and early warning model gradually learns the potential change rules and characteristics in the data, and realizes the modeling of the complex relationship between input and output; during the training process, it is ensured that the model gradually converges, that is, it can obtain a lower prediction error on the training set, which indicates that the learning process of the model is effective; by monitoring the difference between the training error and the verification error, the parameter adjustment strategy is timely corrected to prevent the model from overfitting the training set, thereby improving the generalization ability of the model on new data. Step S222: Verification process, apply the trained model to an independent validation data set, calculate the performance indicators of the model on the set, such as prediction accuracy, mean absolute error (MAE), and mean square error (MSE); use these indicators to evaluate the performance of the model on unseen data. Significance achieved: Ensure that the trained model can effectively capture the features in the data and has the ability to make reasonable predictions in actual application scenarios; the performance indicators in the verification process provide a quantitative method to compare performance with other models to select the best solution; after evaluation, it can be determined whether further adjustments are needed to the model structure, hyperparameters, or training process to improve the performance of the model. Step S223: Final evaluation process, use an independent test set (this set is not involved in training or verification) to perform a final evaluation of the model to check its generalization ability to new data; use feature importance analysis techniques (such as SHAP, LIME, etc.) to evaluate the contribution of each feature in model prediction and analyze the interpretability of the model output. Significance achieved: This stage helps confirm the actual performance of the model in the implementation environment to ensure its reliability and effectiveness in practical applications; through feature importance analysis, it provides transparency in the model's decision-making, allowing users to understand the model's response and enhance trust in the model's output results; it is especially important in the field of environmental monitoring because decisions may affect public safety and health; by analyzing the importance of features, it is possible to find out which factors actually affect environmental conditions, thereby guiding future data collection, strategy formulation and model improvement.

[0114] In summary, through these three steps, the environmental monitoring and early warning model of this embodiment can not only accurately reflect the mapping relationship between parameter changes and environmental changes, but also ensure the reliability and interpretability of the model. Such a step-by-step training and evaluation process not only improves the prediction ability of the model, but also provides a solid foundation for its credibility in practical applications.

[0115] Furthermore, if Figure 4 As shown, the process of comparing the monitoring and early warning analysis result with the preset threshold in step S3 specifically includes the following steps:

[0116] Step S31: Obtain monitoring and early warning analysis results from the environmental monitoring and early warning model, including road ice and snow status (whether there is ice and snow), temperature;

[0117] Step S32: by analyzing and extracting historical environmental monitoring data and accidents and slips, etc., to determine certain key indicators, ice and snow may appear when the temperature is below a certain value;

[0118] Step S33: Compare the monitoring warning analysis result with the preset threshold value to determine whether to issue a warning;

[0119] Warning trigger: If the conditions are met, the vehicle enters the warning state, records the state, and prepares for subsequent steps (such as triggering an alarm, recording a log, etc.); when ice or snow is detected on the road, the state can be marked as "warning" and the corresponding processing flow can be executed;

[0120] After confirming whether an early warning is needed, perform the following operations on the visualization device at the same time:

[0121] Marking location and size: On the visual interface, the currently detected road environment problems (such as location, size, etc.) are displayed in a map or graphical form; for example, if ice and snow are detected on a specific road section, the section can be highlighted in the interface and the scope of ice and snow coverage can be noted.

[0122] Preferably, step S31 of this embodiment obtains the monitoring and early warning analysis results, and extracts the current monitoring data from the environmental monitoring and early warning model, including key indicators such as the road ice and snow status and temperature. The significance achieved: ensure that the acquired data is the latest, thereby providing an accurate basis for subsequent judgments; by bringing together important parameters (such as ice and snow status and temperature), support the subsequent decision-making process, especially when environmental conditions are constantly changing. Step S32 analyzes and determines key indicators, analyzes historical environmental monitoring data, accident records, and slip events to identify key indicators related to the road ice and snow status (such as specific temperature thresholds). The significance achieved: through data analysis and historical event research, ensure that the prediction basis of the early warning model is based on actual conditions and increase the credibility of the model; confirming key indicators helps to identify potential dangers in advance and facilitate timely measures to prevent accidents. Step S33 compares with the preset threshold, compares the obtained monitoring and early warning analysis results with the preset threshold determined by historical data, and determines whether the early warning conditions are met. Significance achieved: Through effective threshold comparison, the system can accurately identify when an early warning is needed to avoid false alarms or missed alarms; quickly respond to changes in environmental conditions and provide information to managers and users so that they can take appropriate measures in a timely manner. Status marking and recording: Once the conditions for issuing an early warning are met, the system will record the status and start the corresponding processing process (such as triggering an alarm, recording a log, etc.). Significance achieved: Formal marking as "early warning" means a rapid response to potential dangers, which can significantly improve road safety and reduce the risk of accidents; automatically record and process early warning information, reduce human intervention, and improve the operating efficiency and response speed of the system. The position and size of the annotations on the visualization device are displayed in the visualization interface through a map or graphic method, and the location and coverage of the detected ice and snow are clearly marked. Significance achieved: Visual display makes information easy to understand and quickly conveys road conditions and potential risks to managers, drivers and the public; visualized data provides intuitive navigation to help decision makers quickly judge and take action, such as choosing a safer driving route or implementing further cleaning measures.

[0123] In summary, through this series of steps, the monitoring and early warning analysis process of this embodiment not only enhances the ability to respond to environmental changes, but also improves the predictability and handling efficiency of potential dangers. The effective implementation of the entire process can provide important protection for road safety, thereby reducing traffic accidents and losses caused by environmental factors.

[0124] like Figure 5As shown, this embodiment also provides an embodiment of a road and bridge pavement environment monitoring and early warning system. In this embodiment, the road and bridge pavement environment monitoring and early warning system is applied to the road and bridge pavement environment monitoring and early warning method in the above embodiment. The road and bridge pavement environment monitoring and early warning system includes: a data acquisition module 1, a model training module 2, and an early warning judgment module 3;

[0125] Among them, the data acquisition module 1 is used to deploy acquisition equipment for collecting road environment data, and the road environment data includes road ice and snow status data; and the collected road environment data is preprocessed to obtain preprocessed road environment data; the model training module 2 is used to train the environmental monitoring and early warning model based on the historical road environment data to obtain the trained environmental monitoring and early warning model, and input the preprocessed environmental data into the environmental monitoring and early warning model for monitoring and early warning, and obtain the results of monitoring and early warning analysis; the early warning judgment module 3 is used to compare the results of the monitoring and early warning analysis with the preset threshold, and confirm whether a warning needs to be issued based on the comparison results; at the same time, when ice and snow are detected on the road, the location and size are marked in the visualization device.

[0126] Preferably, the data acquisition module 1 of this embodiment is equipped with acquisition equipment (such as sensors, cameras, etc.) to collect road environment data in real time, including temperature, humidity, ice and snow status, etc.; the collected data is cleaned and processed to eliminate noise, fill missing values, ensure the integrity and consistency of the data set, and thus obtain reliable pre-processed data. Significance achieved: Through continuous data acquisition, the system can reflect the environmental status of the road in real time and provide timely information for subsequent decision-making; the pre-processing link ensures data accuracy, reduces misjudgments caused by data quality problems, and improves the credibility of the overall analysis results. The model training module 2 is based on historical environmental data and trains a model that can identify and warn environmental changes through machine learning methods (such as regression analysis, neural networks, etc.); the real-time pre-processed environmental data is input into the trained model to generate corresponding monitoring and early warning analysis results. Significance achieved: The model grasps the law of environmental changes by learning historical data, improves the accuracy and adaptability of prediction, and helps to achieve adaptive early warning for different climates and road conditions; the trained model can identify potential risks and issue an alarm before the situation deteriorates, thereby enhancing the safety of the transportation system. The early warning judgment module 3 compares the results of the monitoring and early warning analysis (such as the ice and snow status and temperature of the road surface, etc.) with the preset safety threshold to determine whether the early warning condition is triggered; the monitored environmental problems are displayed in the visualization device, such as highlighting the affected road section and its specific location and area. Significance achieved: Accurate judgment and triggering mechanisms ensure that relevant parties (such as drivers, traffic management departments) are quickly notified when road conditions are dangerous, effectively reducing the risk of accidents; visual display makes data more readable, helping people quickly understand the current situation and make better decisions (such as choosing a driving route, implementing cleaning measures, etc.).

[0127] In summary, the three modules of this embodiment form a complete road and bridge pavement environmental monitoring and early warning system, each playing an indispensable role. The data acquisition module provides basic data, the model training module uses this data to build an intelligent early warning model, and the early warning judgment module ensures that it can respond to environmental changes in a timely manner. The effective operation of the overall system will significantly improve traffic safety and help cope with the challenges brought by severe weather conditions.

[0128] like Figure 6 As shown, this embodiment provides an embodiment of an electronic device. In this embodiment, the electronic device 4 includes a processor 41 and a memory 42 coupled to the processor 41.

[0129] The memory 42 stores program instructions for implementing the road and bridge pavement environment monitoring and early warning method of any of the above embodiments.

[0130] The processor 41 is used to execute program instructions stored in the memory 42 to perform road and bridge pavement environment monitoring and early warning.

[0131] The processor 41 may also be referred to as a CPU (Central Processing Unit). The processor 41 may be an integrated circuit chip having signal processing capabilities. The processor 41 may also 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, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0132] Further, Figure 7 The schematic diagram of the structure of the storage medium of an embodiment of the present application is that the storage medium 5 of the embodiment of the present application stores program instructions 51 that can implement all the above methods, wherein the program instructions 51 can be stored in the above storage medium in the form of a software product, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or terminal devices such as a computer, a server, a mobile phone, and a tablet.

[0133] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0134] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units. The above is only an implementation mode of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present invention.

[0135] The specific implementation methods of the invention are described in detail above, but they are only examples, and the invention is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the invention, therefore, the equalization, modification, improvement, etc. made without departing from the spirit and principle of the invention should be included in the scope of the invention.

Claims

1. A road and bridge pavement environment monitoring and early warning method, characterized in that: The road and bridge pavement environment monitoring and early warning method comprises: Arrange collection equipment for collecting road environment data, the road environment data including road ice and snow state data; and pre-process the collected road environment data to obtain pre-processed road environment data; The environmental monitoring and early warning model is trained based on the historical environmental data of the road surface to obtain the trained environmental monitoring and early warning model, and the pre-processed environmental data is input into the environmental monitoring and early warning model for monitoring and early warning to obtain the results of monitoring and early warning analysis; Compare the monitoring and early warning analysis results with the preset thresholds, and determine whether an early warning is needed based on the comparison results; at the same time, when ice and snow are detected on the road, the location and size are marked in the visualization device; Obtain preprocessed road environment data, including: Obtain the road surface to be monitored, determine the target area to be measured according to the length and width of the road surface, and determine the measuring distance of the remote sensing road condition sensor. Install the remote sensing road condition sensor, point the lens directly at the road surface area to be measured, and collect road condition information. The infrared transmitter of the remote sensing road condition sensor is aimed at the selected position on the road surface, and emits a spectral signal. The spectral signal is reflected from the measured location back to the receiver. The remote sensing road condition sensor can distinguish the reflection of specific wavelengths caused by ice and snow. The reflection characteristics of ice layer and water film to spectral signals are different, so as to judge the ice and snow status of the road surface. At the same time, it also receives the infrared energy radiated by the road surface, and then obtains the road surface temperature value. The remote sensing road condition sensor analyzes the received spectral signals to obtain real-time data of road surface temperature and road surface conditions, performs denoising on the real-time data, and outputs the results to the environmental monitoring and early warning model and visualization equipment; Realize the judgment of the ice and snow status of the road, including: Convert the time domain signal of the spectral signal to the frequency domain and identify the frequency components in the spectral signal; extract the characteristic wavelength related to the state of ice and snow by analyzing the spectral characteristics of the spectral signal; compare the extracted characteristic wavelength with a known standard spectral library, which contains the reflection characteristics of different materials; A classification model is constructed using a decision tree. By training the intensity of some extracted characteristic wavelengths, a mapping relationship between the road conditions of ice and snow, dry road surface, and water film is established. The received wavelength data is input into the trained model classification type to determine the current road condition, including ice and snow coverage, dry, or water film. Obtain real-time road surface temperature data from remote sensing road condition sensors; the current measured road surface temperature is below 0 degrees, indicating that there may be ice on the road surface; combine wavelength discrimination and temperature status to set a series of rules; Use decision trees to build classification models, including: Collect data related to road conditions, clean the data, and select characteristic wavelengths related to ice and snow, dry road conditions, and water film conditions; The data set is divided into a training set and a test set; the root node of the decision tree corresponds to all input samples, and the information gain of each feature is calculated to select the optimal feature for division until the termination condition is met and the depth of the decision tree reaches the maximum limit; each path, from the root node to the leaf node, represents a set of decision rules, and the final leaf node represents the classification result of the sample; The classification model is cross-validated to evaluate its accuracy and stability. The evaluation indicators include accuracy, recall and F1-score. The newly received spectral data is input into the trained decision tree model. The decision tree model determines the current road surface state based on the newly input characteristic wavelength and outputs the prediction result, "ice and snow cover", "dry" or "water film state". Environmental monitoring and early warning model construction, including: We chose to use an artificial neural network environmental monitoring and early warning model to conduct in-depth analysis of the preprocessed data and extract features that are useful for environmental monitoring and early warning; Use historical road environmental data to train the environmental monitoring and early warning model, adjust the parameters of the environmental monitoring and early warning model so that the environmental monitoring and early warning model reflects the mapping relationship between parameter changes and environmental changes, and use the verification data set to verify and evaluate the trained environmental monitoring and early warning model; Analyze the preprocessed data to monitor the current road environment conditions. When the monitored data changes, analyze the changed data to predict upcoming environmental problems and notify staff through visualization equipment or alarms.

2. The road and bridge pavement environment monitoring and early warning method according to claim 1, characterized in that: A set of rules includes: If the wavelength characteristics show ice and the temperature is ≤ 0 degrees, the road surface is judged to be icy; If the temperature is between 0 ≤ temperature ≤ 5 degrees, and the spectral data supports the presence of a water film, it is judged to be in a water film state; If the spectral data matches the characteristic wavelength of a dry road surface, it is judged to be in a dry state.

3. The road and bridge pavement environment monitoring and early warning method according to claim 1, characterized in that: The environmental monitoring and early warning model reflects the mapping relationship between parameter changes and environmental changes, including: Input the training data into the environmental monitoring and early warning model, calculate the predicted value through forward propagation, and calculate the gradient according to the loss function; use the optimization algorithm to update the weights and biases of the network; repeat until the error of the environmental monitoring and early warning on the training set no longer decreases significantly or reaches the preset maximum number of training iterations; After the training is completed, the validation data set is input into the trained environmental monitoring and early warning model to calculate the performance of the environmental monitoring and early warning model on the validation set. The key indicators include: prediction accuracy, mean absolute error or mean square error; Use the test set involved in training or validation to evaluate the environmental monitoring and early warning model to ensure its predictive effect on new data; use feature importance analysis to evaluate the performance of the environmental monitoring and early warning model and explain the output results.

4. The road and bridge pavement environment monitoring and early warning method as claimed in claim 1, characterized in that: Compare the results of monitoring and early warning analysis with preset thresholds, including: Obtain monitoring and early warning analysis results from the environmental monitoring and early warning model, including road ice and snow conditions and temperature; By analyzing and extracting historical environmental monitoring data and accidents and slips, we can determine certain key indicators, and ice and snow may appear when the temperature is below a certain value; Compare the monitoring and early warning analysis results with the preset thresholds to determine whether to issue an early warning.

5. The road and bridge pavement environment monitoring and early warning method as claimed in claim 4, characterized in that: Warning trigger: If the conditions are met, the warning state is entered, the state is recorded and the subsequent steps are prepared; when ice or snow is detected on the road surface, the state is marked as a warning and the corresponding processing flow is executed; After confirming whether an early warning is needed, perform the following operations on the visualization device at the same time: Marking location and size: On the visual interface, the currently detected road environment problems are displayed through a map or graphically; if ice and snow are detected on a specific road section, the section will be highlighted in the interface and the scope of ice and snow coverage will be indicated.

6. A road and bridge pavement environment monitoring and early warning system, which is applied to the road and bridge pavement environment monitoring and early warning method as claimed in any one of claims 1 to 5, characterized in that: The road and bridge pavement environment monitoring and early warning system comprises: The data acquisition module is used to deploy acquisition equipment for collecting road environment data, where the road environment data includes road ice and snow status data; and to pre-process the collected road environment data to obtain pre-processed road environment data; The model training module is used to train the environmental monitoring and early warning model based on the historical environmental data of the road surface, obtain the trained environmental monitoring and early warning model, input the pre-processed environmental data into the environmental monitoring and early warning model for monitoring and early warning, and obtain the results of monitoring and early warning analysis; The early warning judgment module is used to compare the results of the monitoring and early warning analysis with the preset thresholds, and confirm whether a warning needs to be issued based on the comparison results; at the same time, when ice and snow are detected on the road surface, the location and size are marked in the visualization device.

Images