Method and system for evaluating the anti-skid performance of an ice surface anti-skid texture

Abstract

The application discloses an anti-skid performance evaluation method for ice body surface anti-skid texture structure, and comprises the following steps: S1, acquiring ice body surface micro-texture data and environmental condition data, the ice body surface micro-texture data comprises surface roughness, texture distribution density and texture depth, and the environmental condition data comprises environmental temperature, humidity, wind speed and vehicle passing frequency, and the application relates to the technical field of traffic safety. Through the acquisition of the ice body surface micro-texture and the environmental condition data, a double-model collaborative prediction mechanism is constructed, real-time data comparison and error correction are combined, the limitation that the existing technology is difficult to dynamically capture the influence of micro-texture changes on anti-skid performance is solved, real-time monitoring of the dynamic changes of the ice body surface micro-texture is realized, environmental factors and vehicle passing data are combined, double-model collaborative prediction and error correction are realized, and the real-time performance and accuracy of the anti-skid performance evaluation are improved.

Description

Technical Field

[0001] This invention relates to the field of traffic safety technology, specifically to a method and system for evaluating the anti-slip performance of anti-slip textured structures on ice surfaces. Background Technology

[0002] In icy and snowy weather, the anti-skid performance of roads and other icy surfaces is crucial for traffic safety. Currently, various methods exist for assessing the anti-skid performance of ice surfaces based on their texture, such as measuring the road surface friction coefficient and texture depth. However, existing technologies have limitations in assessing the dynamic changes in anti-skid performance caused by microstructural variations in ice surfaces. In the natural environment, the microstructure of ice surfaces changes over time due to factors such as temperature, humidity, and vehicle traffic. These subtle changes significantly impact the anti-skid performance of ice. Existing assessment methods have limitations in terms of real-time performance and accuracy, making it difficult to comprehensively capture the dynamic changes in anti-skid performance caused by these microstructural variations. Therefore, it is challenging to provide road users with real-time information on the anti-skid performance of ice surfaces.

[0003] In practical applications, existing technologies typically rely on static measurement methods, lacking the ability to continuously monitor dynamic changes. For example, traditional methods mainly focus on measuring macroscopic physical parameters, while rarely addressing the evolution of microscopic textures and their impact on anti-skid performance. Furthermore, the complexity and variability of environmental factors further increase the difficulty of real-time assessment, posing challenges to the accuracy and timeliness of data collection and analysis. This limitation may lead to discrepancies between anti-skid performance assessment results and actual conditions, thus affecting the timeliness of traffic safety warnings and decisions.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for evaluating the anti-slip performance of ice surface anti-slip texture structure. It features real-time monitoring of the dynamic changes in the micro-texture of the ice surface, combined with environmental factors and vehicle traffic data, and improves the real-time performance and accuracy of anti-slip performance evaluation through dual-model collaborative prediction and error correction.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the anti-slip performance of an anti-slip texture structure on an ice surface, comprising the following steps: S1. Acquire micro-texture data of the ice surface and environmental condition data. The micro-texture data of the ice surface includes surface roughness, texture distribution density, and texture depth. The environmental condition data includes ambient temperature, humidity, wind speed, and vehicle traffic frequency. S2. Classify the micro-texture data of the ice surface and environmental condition data to obtain classification results. Based on the classification results, establish a data partitioning model and deploy an anti-skid performance evaluation mirror system on a distributed computing platform. In the data partitioning model, preprocess and extract features from the corresponding micro-texture data of the ice surface and environmental condition data to extract surface roughness variation features, texture distribution density variation features, and dynamic environmental condition variation features. S3. Retrieve historical data of the ice surface, combine surface roughness variation characteristics, texture distribution density variation characteristics, and environmental condition dynamic variation characteristics to construct a first dynamic variation model of the anti-slip performance of the ice surface, and update the model in real time. In the anti-slip performance evaluation mirror system, use the micro-texture data of the ice surface and environmental condition data to train the support vector machine model to construct a second dynamic variation model of the anti-slip performance of the ice surface. In the anti-slip performance evaluation mirror system, update the second dynamic variation model of the anti-slip performance of the ice surface in real time. Compare the prediction results of the first dynamic variation model of the anti-slip performance of the ice surface with those of the second dynamic variation model of the anti-slip performance of the ice surface. If the difference in the prediction results is within the preset range, the first dynamic variation model of the anti-slip performance of the ice surface takes effect. S4. Substitute the real-time collected micro-texture data of the ice surface and environmental condition data into the dynamic change model of the anti-slip performance of the first ice surface to obtain the dynamic change result of the anti-slip performance of the ice surface. Collect the real-time anti-slip performance data of the ice surface, compare the real-time anti-slip performance data with the dynamic change result, and generate the anti-slip performance error. S5, analyze the anti-slip performance error, extract the time node information corresponding to the error, establish the correlation between the time node information and the anti-slip performance evaluation index of the ice surface, and generate optimization suggestions for the anti-slip performance of the ice surface.

[0007] Furthermore, this application also proposes that step S1 further includes the following steps: detecting the integrity of ice surface micro-texture data and environmental condition data, including surface roughness, texture distribution density, texture depth, ambient temperature, humidity, wind speed, and vehicle traffic frequency; generating an original data detection report, which contains qualified data and missing data information; comparing the data types of the original data detection report with historical data of the ice surface; if the data types are consistent, it is determined that the ice surface micro-texture data and environmental condition data match the historical data; if the data types are inconsistent, the difference data is identified and data type is marked, generating difference data category information; establishing new data detection rules based on the difference data category information; substituting the data corresponding to the new data detection rules and the data corresponding to the original data detection report into a preset digital twin simulation model to generate a dynamic evolution image of ice surface micro-texture; obtaining the key point coordinate information in the dynamic evolution image of ice surface micro-texture; verifying the accuracy of the key point coordinate information using the known basic point coordinates of the ice surface; if the key point coordinate information is verified, the ice surface micro-texture data and environmental condition data are marked as valid data.

[0008] Furthermore, this application also proposes that step S2 further includes the following steps: verifying whether the data conforms to a predetermined classification standard, which includes data category, environmental condition type, and timestamp label; performing preprocessing operations on the classified data, including data denoising, data normalization, and data format conversion; receiving anti-skid performance evaluation requirements, determining the type of features to be extracted based on the evaluation requirements, and extracting selected features from the preprocessed data using time series feature analysis; verifying the effectiveness of the extracted features using historical data or known anti-skid performance evaluation results, and adjusting the feature extraction strategy based on the verification results, including modifying the feature type and optimizing the feature extraction algorithm.

[0009] Furthermore, this application also proposes that step S3 further includes the following steps: simultaneously inputting the real-time collected micro-texture data of the ice surface and environmental condition data into the first dynamic change model of the anti-slip performance of the ice surface and the second dynamic change model of the anti-slip performance of the ice surface; the first dynamic change model of the anti-slip performance of the ice surface and the second dynamic change model of the anti-slip performance of the ice surface respectively predict based on the input data to generate corresponding dynamic change results of anti-slip performance, the dynamic change results of anti-slip performance including the first dynamic change result of anti-slip performance and the second dynamic change result of anti-slip performance; and processing the first dynamic change result of anti-slip performance and the second dynamic change result of anti-slip performance. A comparison is made, and the difference between the two is calculated. A difference index is defined to quantify the difference between the dynamic change results of the first and second anti-slip performance. The prediction errors of the real-time dynamic change models of the first and second ice surface anti-slip performance are calculated, and the prediction errors are compared with a preset error range. If the prediction error is within the preset error range, the dynamic change model of the first ice surface anti-slip performance is deemed effective. If the prediction error exceeds the preset range, the dynamic change models of the first and second ice surface anti-slip performance are optimized, and the above steps are repeated for verification.

[0010] Furthermore, this application also proposes that step S4 further includes the following steps: establishing a data processing grouping mechanism, establishing a grouping correspondence between the real-time anti-slip performance data of the ice surface and the dynamic change results of anti-slip performance and the data processing grouping mechanism, generating a data grouping task according to the grouping correspondence, and executing the data grouping task to perform parallel grouping processing of the real-time anti-slip performance data of the ice surface and the dynamic change results of anti-slip performance.

[0011] Furthermore, this application also proposes an anti-skid performance evaluation system for anti-skid textured structures on ice surfaces, which runs the aforementioned anti-skid performance evaluation method for anti-skid textured structures on ice surfaces. The system includes a central server, an environmental monitoring terminal, a road surface sensor network, and user terminal equipment. The central server establishes communication connections with the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment, respectively. The central server integrates and analyzes the data received from the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment, generates optimization suggestions for the anti-skid performance of the ice surface, and transmits them to the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment.

[0012] Furthermore, this application also proposes that the environmental monitoring terminal includes a temperature sensor, a humidity sensor, a wind speed sensor, and a vehicle traffic frequency detector. The temperature sensor uses thermistor technology to measure the ambient temperature, the humidity sensor measures humidity based on the capacitive humidity sensing principle, the wind speed sensor uses ultrasonic speed measurement technology to measure wind speed, and the vehicle traffic frequency detector uses geomagnetic induction technology to count the vehicle traffic frequency.

[0013] Furthermore, this application proposes that the road surface sensor network includes multiple distributed sensor nodes, each sensor node containing a surface roughness sensor, a texture distribution density sensor, and a texture depth sensor. The surface roughness sensor uses laser scanning technology to generate three-dimensional topographic data of the ice surface, the texture distribution density sensor calculates the texture distribution density based on image recognition technology, and the texture depth sensor uses an ultrasonic probe to measure the texture depth.

[0014] Furthermore, this application proposes that the central server includes a data integrity detection module, a data preprocessing module, a feature extraction module, a model building module, and an error analysis module. The data integrity detection module is used to detect data integrity and generate an original data detection report. The data preprocessing module is used to perform noise reduction, normalization, and format conversion operations on the data. The feature extraction module is used to extract surface roughness change features, texture distribution density change features, and dynamic environmental condition change features. The model building module is used to construct a dynamic change model of the anti-slip performance of the first ice surface and a dynamic change model of the anti-slip performance of the second ice surface. The error analysis module is used to analyze the anti-slip performance error and generate optimization suggestions.

[0015] Furthermore, this application also proposes that the user terminal device includes a smartphone, tablet computer, or dedicated display device, and that the user terminal device establishes a communication connection with a central server via the Internet to receive and display suggestions for optimizing the anti-slip performance of the ice surface.

[0016] Beneficial effects This invention acquires data on the micro-texture of the ice surface and environmental conditions, constructs a dual-model collaborative prediction mechanism, and combines real-time data comparison and error correction. This overcomes the limitations of existing technologies in dynamically capturing the impact of micro-texture changes on anti-skid performance. It has the ability to monitor the dynamic changes of the micro-texture of the ice surface in real time, and improves the real-time performance and accuracy of anti-skid performance assessment by combining environmental factors and vehicle traffic data and through dual-model collaborative prediction and error correction. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1 This invention provides a technical solution: a method for evaluating the anti-slip performance of an anti-slip texture structure on an ice surface, comprising the following steps: S1. Acquire micro-texture data of the ice surface and environmental condition data. The micro-texture data of the ice surface includes surface roughness, texture distribution density, and texture depth. The environmental condition data includes ambient temperature, humidity, wind speed, and vehicle traffic frequency. S2. Classify the micro-texture data of the ice surface and environmental condition data to obtain classification results. Based on the classification results, establish a data partitioning model and deploy an anti-skid performance evaluation mirror system on a distributed computing platform. In the data partitioning model, preprocess and extract features from the corresponding micro-texture data of the ice surface and environmental condition data to extract surface roughness variation features, texture distribution density variation features, and dynamic environmental condition variation features. S3. Retrieve historical data of the ice surface, combine surface roughness variation characteristics, texture distribution density variation characteristics, and environmental condition dynamic variation characteristics to construct a first dynamic variation model of the anti-slip performance of the ice surface, and update the model in real time. In the anti-slip performance evaluation mirror system, use the micro-texture data of the ice surface and environmental condition data to train the support vector machine model to construct a second dynamic variation model of the anti-slip performance of the ice surface. In the anti-slip performance evaluation mirror system, update the second dynamic variation model of the anti-slip performance of the ice surface in real time. Compare the prediction results of the first dynamic variation model of the anti-slip performance of the ice surface with those of the second dynamic variation model of the anti-slip performance of the ice surface. If the difference in the prediction results is within the preset range, the first dynamic variation model of the anti-slip performance of the ice surface takes effect. S4. Substitute the real-time collected micro-texture data of the ice surface and environmental condition data into the dynamic change model of the anti-slip performance of the first ice surface to obtain the dynamic change result of the anti-slip performance of the ice surface. Collect the real-time anti-slip performance data of the ice surface, compare the real-time anti-slip performance data with the dynamic change result, and generate the anti-slip performance error. S5, analyze the anti-slip performance error, extract the time node information corresponding to the error, establish the correlation between the time node information and the anti-slip performance evaluation index of the ice surface, and generate optimization suggestions for the anti-slip performance of the ice surface.

[0020] Among them, the micro-texture data of the ice surface includes surface roughness, texture distribution density, and texture depth. Specifically, it can be realized by using laser scanning technology, image recognition technology, and ultrasonic probe measurement technology to quantify the physical properties of the ice surface microstructure and capture the impact of dynamic changes in micro-texture on anti-skid performance.

[0021] The environmental condition data includes ambient temperature, humidity, wind speed, and vehicle traffic frequency. Specifically, it is achieved using a thermistor temperature sensor, a capacitive humidity sensor, an ultrasonic wind speed sensor, and a geomagnetic induction detector to monitor the cumulative effect of external environmental factors on the anti-slip performance of the ice surface in real time.

[0022] The data partitioning model categorizes data based on data type and environmental condition type, and is implemented using a parallel processing architecture of a distributed computing platform to improve the efficiency of data preprocessing and feature extraction, thus meeting the needs of processing large-scale real-time data.

[0023] Among them, the surface roughness variation characteristics, texture distribution density variation characteristics, and environmental condition dynamic variation characteristics are extracted using time series feature analysis algorithms to identify the evolution patterns of micro-textures and environmental factors over time, providing input parameters for dynamic model construction.

[0024] Among them, the dynamic change model of the anti-slip performance of the first ice surface is constructed based on historical data and is implemented by statistical regression or physical mechanism modeling methods to reflect the trend of anti-slip performance changes under long-term data accumulation.

[0025] Among them, the dynamic change model of the anti-slip performance of the second ice surface is based on support vector machine training and is specifically implemented using machine learning algorithms. It is used to dynamically optimize model parameters through real-time data to improve prediction accuracy.

[0026] The preset range of difference in prediction results is set by an error threshold, specifically defined by statistical confidence intervals or engineering experience values, to verify the consistency of predictions from different models and ensure the reliability of the evaluation results.

[0027] Among them, the anti-skid performance error is generated by comparing real-time data with model prediction results. Specifically, it is achieved by residual calculation or error backpropagation algorithm, which is used to quantify the model prediction deviation and guide the generation of subsequent optimization suggestions. The core innovation of this application lies in the dynamic acquisition of ice surface micro-texture data and environmental condition data, combined with a dual verification mechanism of a mechanistic model based on historical data and a machine learning model based on real-time data, to achieve real-time and accurate assessment of changes in anti-skid performance. This method utilizes a distributed computing platform to process multi-source heterogeneous data and continuously optimizes the assessment model through an error feedback mechanism, solving the technical challenge of existing technologies being unable to effectively capture dynamic changes in micro-texture and the superimposed influence of environmental factors.

[0028] The working process and principle of this application are as follows: First, microscopic texture data and environmental condition data of the ice surface are acquired. The microscopic texture data includes surface roughness, texture distribution density, and texture depth, while the environmental condition data includes ambient temperature, humidity, wind speed, and vehicle traffic frequency. These data are collected in real time through a dedicated sensor network.

[0029] Next, the acquired data is classified and processed to establish a data partitioning model. A mirror system for anti-slip performance evaluation is deployed on a distributed computing platform, and the data partitioning model is used to preprocess the data and extract features. Extracted features include surface roughness variation characteristics, texture distribution density variation characteristics, and dynamic environmental condition variation characteristics. This step achieves efficient data processing and extraction of key features.

[0030] Then, historical data of the ice surface is retrieved, and combined with the extracted features to construct a first dynamic change model of the ice surface's anti-skid performance, which is updated in real time. Simultaneously, a second dynamic change model of the ice surface's anti-skid performance is constructed using a support vector machine model and also updated in real time. The prediction results of the two models are compared; if the difference is within a preset range, the first model takes effect. This dual-model mechanism improves the accuracy and reliability of the predictions.

[0031] By substituting the real-time collected data into the effective model, the dynamic changes in anti-skid performance are obtained. Simultaneously, real-time anti-skid performance data is collected and compared with the dynamic changes to generate anti-skid performance error. This step achieves the comparative verification of model predictions and actual conditions.

[0032] Finally, the anti-skid performance error is analyzed, the corresponding time node information is extracted, and a correlation is established between this error and the anti-skid performance evaluation index to generate optimization suggestions. This step realizes a continuous optimization mechanism based on error analysis.

[0033] The entire process forms a closed-loop system, which achieves accurate evaluation and dynamic optimization of the anti-slip performance of the ice surface through real-time data acquisition, dual-model dynamic verification and error feedback optimization.

[0034] As a preferred embodiment, the solution of this application is specifically implemented as follows: During the data acquisition phase, a laser scanner was used to measure the surface roughness of the ice body with an accuracy of 0.01 mm. Texture distribution density was obtained by capturing surface images with a high-resolution camera and performing image processing, with a resolution of 1200 dpi. Texture depth was measured using an ultrasonic probe with an accuracy of 0.1 mm. Ambient temperature was measured using a PT100 platinum resistance temperature sensor with an accuracy of ±0.1℃. Humidity was measured using a capacitive humidity sensor with an accuracy of ±2%RH. Wind speed was measured using an ultrasonic anemometer with an accuracy of ±0.1 m / s. Vehicle traffic frequency was counted using a geomagnetic induction coil, updated every minute.

[0035] Data classification and processing employed the K-means clustering algorithm, dividing the data into three levels: high, medium, and low. The data partitioning model was established based on geographic location and timestamps, with each partition corresponding to an independent data processing unit.

[0036] Feature extraction employed wavelet transform to perform multi-scale analysis of surface roughness and texture distribution density. Dynamic environmental condition characteristics were extracted through time series analysis using an autoregressive integral moving average model.

[0037] The dynamic change model of the anti-skid performance of the first ice surface was constructed using a multiple linear regression method. The input variables were the extracted features, and the output variable was the predicted anti-skid performance index. The model updates its parameters every 5 minutes.

[0038] The second model for the dynamic change of the anti-skid performance of the ice surface uses a support vector machine algorithm, with a radial basis function chosen as the kernel function. The model is trained using cross-validation, retraining every 10 minutes.

[0039] The root mean square error was used as the metric to compare the prediction results of the two models, with a preset difference range of 5%.

[0040] Error analysis of anti-skid performance was performed using the Fourier transform method to extract the periodic characteristics of the error. The correlation between time node information and anti-skid performance evaluation indicators was calculated using the Pearson correlation coefficient.

[0041] The optimization suggestions are generated based on the decision tree algorithm. The inputs are error features and correlation coefficients, and the outputs are specific optimization measures and expected results.

[0042] Through the above-described scheme, this application achieves real-time and accurate assessment of the anti-skid performance of ice surfaces. Synchronous acquisition and processing of multi-source data eliminates data lag issues and improves the timeliness of the assessment. A dual-model collaborative verification mechanism enhances the reliability of prediction results and reduces the probability of misjudgment. A distributed data partitioning model improves data processing efficiency, enabling the system to handle large-scale data streams. Real-time error analysis and optimization suggestion generation mechanisms allow the system to continuously self-optimize and adapt to complex and changing environmental conditions. These improvements work together to significantly enhance the accuracy and practicality of ice surface anti-skid performance assessment, providing strong support for winter road safety management.

[0043] In some of the solutions described above in this application, when obtaining microscopic texture data of the ice surface and environmental condition data, there may be problems such as missing data or inconsistent data types, which may lead to deviations between the subsequent model prediction results and the actual anti-skid performance.

[0044] This application further proposes to detect the integrity of micro-texture data and environmental condition data on the surface of ice, generate an original data detection report, compare the data types of the report with historical data, mark the discrepancies if they are inconsistent, establish new data detection rules, substitute the discrepancies into a digital twin simulation model to generate dynamic evolution images, obtain the coordinate information of key points in the images, verify the accuracy of key points using known base point coordinates, and mark the data as valid data after verification.

[0045] The data integrity inspection includes surface roughness, texture distribution density, texture depth, ambient temperature, humidity, wind speed, and vehicle traffic frequency. The original data inspection report contains information on both qualified and missing data; data type comparison identifies format differences between historical and currently collected data. The information on the categories of differing data is used to establish new data inspection rules; for example, when the humidity sensor data format changes from percentage to absolute value, the data parsing algorithm needs to be adjusted. The digital twin simulation model simulates the evolution of ice surface texture through physical field coupling calculations, with dynamic evolution images displaying texture structure changes at a rate of 24 frames per second. Key point coordinate verification uses a three-dimensional coordinate matching algorithm; the known base point coordinates include at least 50 reference points distributed across different areas of the ice surface, with an allowable coordinate error range of ±0.5 mm.

[0046] Specifically, when new ice surface data is collected, the system first checks the completeness and format consistency of seven data items, including surface roughness and texture distribution density. For example, if vehicle traffic frequency data is missing, the detection report will mark that field as missing. The current data is then compared with standard data types in the historical database. If texture depth data changes from millimeter-level floating-point numbers to integer records, the system marks the difference field as an anomaly. For anomaly data, new detection rules including data format conversion rules are established, and the anomaly data is input into a digital twin model for simulation verification. During the simulation, the model calculates the stress distribution on the ice surface at a time step of 0.1 seconds, generating a 3D dynamic image containing changes in texture structure. By extracting the coordinates of key points at the 15th second of the image and comparing them spatially with pre-set reference points, the validity of the currently collected data is confirmed when the coordinate offset is less than a preset threshold. This process ensures the spatiotemporal consistency between the microscopic texture data input to the model and the environmental condition data, thereby improving the prediction accuracy of the subsequent anti-skid performance evaluation model.

[0047] As a preferred embodiment, the solution of this application is specifically implemented as follows: After acquiring microscopic texture data of the ice surface and environmental conditions, a data integrity check is performed. The check includes surface roughness, texture distribution density, texture depth, ambient temperature, humidity, wind speed, and vehicle traffic frequency. The check results generate a raw data check report, containing information on both valid and missing data.

[0048] Furthermore, the raw data detection report is compared with historical data on the ice surface. If the data types match, the microscopic texture data of the ice surface and the environmental condition data are determined to match the historical data. If the data types do not match, the discrepancies are identified and labeled with their data types, generating discrepancy category information.

[0049] New data detection rules are established based on the differences in data categories. The data corresponding to the new data detection rules and the data corresponding to the original data detection report are then substituted into a preset digital twin simulation model to generate a dynamic evolution image of the microscopic texture of the ice surface.

[0050] Therefore, the coordinate information of key points in the dynamic evolution image of the micro-texture of the ice surface is obtained. The accuracy of the key point coordinate information is verified using the known coordinates of the base points on the ice surface. If the key point coordinate information passes the verification, the ice surface micro-texture data and environmental condition data are marked as valid data.

[0051] Specifically, data integrity testing employs automated algorithms to verify each data item one by one. The original data testing report is presented in tabular form, clearly marking qualified and missing data. Data type comparison uses machine learning algorithms to automatically identify data type differences. New data testing rules are automatically generated based on the characteristics of the differing data, ensuring targeted testing. The digital twin simulation model uses high-precision 3D modeling technology to accurately simulate the microscopic texture changes on the ice surface. Multi-point calibration technology is used to verify key point coordinate information, improving verification accuracy.

[0052] Through the above technical solutions, this application achieves comprehensive detection and verification of microscopic texture data and environmental condition data of ice surface. Data integrity detection ensures the quality of raw data, and data type comparison improves the accuracy of data matching. The establishment of new data detection rules enhances the adaptability of detection, and the application of digital twin simulation models provides intuitive dynamic evolution images. Verification of key point coordinate information further guarantees the validity of the data. This series of steps significantly improves the reliability and accuracy of data processing, providing a high-quality data foundation for subsequent anti-skid performance evaluation.

[0053] In some of the solutions described above in this application, there may be issues such as inconsistent classification standards or insufficient data preprocessing during the data classification process, which may lead to a mismatch between subsequent feature extraction and anti-slip performance evaluation requirements, thus affecting the model training effect.

[0054] This application further proposes methods to verify whether the data conforms to predetermined classification criteria, which include data category, environmental condition type, and timestamp label. Preprocessing operations are performed on the classified data, including data denoising, data normalization, and data format conversion. The application receives anti-skid performance evaluation requests, determines the types of features to be extracted based on these requests, and uses time series feature analysis to extract selected features from the preprocessed data. The effectiveness of the extracted features is verified using historical data or known anti-skid performance evaluation results. Based on the verification results, the feature extraction strategy is adjusted, including modifying feature types and optimizing the feature extraction algorithm.

[0055] The predefined classification criteria distinguish surface texture data from environmental parameters by data category, environmental condition types correspond to dynamic parameters such as temperature and humidity, and timestamp labels are used to associate data acquisition time nodes. Data denoising uses a sliding window filtering algorithm to eliminate outliers, data normalization uses a maximum-minimum scaling method to unify the units, and data format conversion transforms unstructured image data into structured matrices. Time series feature analysis uses a sliding window statistical method to extract the mean and variance of texture density, and the dynamic change characteristics of environmental conditions are calculated using the difference method to calculate the temperature change rate. Feature validity verification uses the Pearson correlation coefficient to calculate the correlation between the feature and historical anti-slip performance data; if the correlation is below a threshold, the feature is discarded.

[0056] Specifically, the categorized data first undergoes format validation to ensure that the ambient temperature data and texture depth data have a unified timestamp label. During data denoising, pulse interference signals from the surface roughness sensor are eliminated by a median filter, and abnormal fluctuations in texture distribution density are removed using the interquartile range method. Normalization maps temperature data to the -1 to 1 range, and humidity data is converted to a 0 to 1 scale value. Format conversion converts the 3D point cloud data generated by laser scanning into a 512×512 pixel grayscale matrix. After assessing the requirements, if short-term anti-slip performance fluctuations need to be analyzed, the five-minute sliding variance feature of texture distribution density is extracted; if long-term trend prediction is required, the daily average rate of change of surface roughness is extracted. During feature validity validation, by comparing the correlation between the rate of change of texture depth and the anti-slip performance decay curve in historical data, it was found that the correlation coefficient between the rate of change of texture depth and anti-slip performance reaches 0.87, validating its effectiveness. When the correlation coefficient between wind speed variation features and anti-slip performance is lower than 0.3, the feature extraction strategy is adjusted, replacing the wind speed feature with the interaction feature between wind speed and temperature.

[0057] As a preferred embodiment, the solution of this application is specifically implemented as follows: Verify that the data conforms to predetermined classification criteria, which include data category, environmental condition type, and timestamp label. Data categories include surface roughness, texture distribution density, and texture depth. Environmental condition types include temperature, humidity, wind speed, and vehicle traffic frequency. Timestamp labels use a uniform time format, accurate to the second.

[0058] Preprocessing is performed on the categorized data. Median filtering is used for noise reduction to remove outliers. Min-max normalization is used to map values ​​to the 0-1 range. Data format conversion unifies data from different sources into JSON format.

[0059] Upon receiving a request for anti-slip performance evaluation, the types of features to be extracted are determined based on the evaluation requirements. Time series feature analysis is employed to extract selected features from the preprocessed data. Specifically, the sliding window method is used to extract time-series features such as the rate of change of surface roughness, the trend of texture distribution density variation, and the amplitude of environmental condition fluctuations.

[0060] The effectiveness of extracted features is validated using historical data or known anti-skid performance evaluation results. The correlation coefficient between the feature and the known evaluation results is calculated; a correlation coefficient greater than 0.8 is considered a valid feature. Based on the validation results, the feature extraction strategy is adjusted, including modifying feature types and optimizing the feature extraction algorithm. For example, if a feature has low correlation, it may be replaced with another potential feature or the feature extraction parameters may be adjusted.

[0061] Through the above technical solutions, this application achieves effective classification and preprocessing of ice surface micro-texture data and environmental condition data, improving data quality. Simultaneously, the feature extraction and verification process ensures the validity and relevance of the extracted features, providing a reliable data foundation for subsequent anti-skid performance evaluation. Furthermore, the dynamic adjustment mechanism of the feature extraction strategy enhances the model's adaptability, enabling it to be optimized for different scenarios and data characteristics, thereby improving the accuracy and reliability of anti-skid performance evaluation.

[0062] In some of the above-mentioned schemes in this application, a dynamic change model of the anti-skid performance of the first ice surface and a dynamic change model of the anti-skid performance of the second ice surface were constructed. However, inconsistencies in the prediction results may occur during the model prediction process, making it difficult to verify the effectiveness of the model and affecting the reliability of the evaluation results. This application further proposes to simultaneously input real-time collected micro-texture data of the ice surface and environmental condition data into a first dynamic change model and a second dynamic change model of the anti-slip performance of the ice surface. The first and second dynamic change models of the anti-slip performance of the ice surface are then used to predict the corresponding dynamic changes in anti-slip performance based on the input data, generating corresponding dynamic change results. The first and second dynamic change results of anti-slip performance are compared, and the difference between them is calculated. A difference index is defined to quantify the difference between the first and second dynamic change results of anti-slip performance, and the prediction error of the real-time first and second dynamic change models of the anti-slip performance of the ice surface is calculated and compared with a preset error range. If the prediction error is within the preset error range, the first dynamic change model of the anti-slip performance of the ice surface is deemed effective. If the prediction error exceeds the preset range, the first and second dynamic change models of the anti-slip performance of the ice surface are optimized, and the above steps are repeated for verification. In this model, real-time collected microscopic texture data of the ice surface and environmental condition data are input into two models in parallel to ensure data synchronization; the dynamic change results of anti-skid performance are generated by the internal algorithm of the model, including time series prediction values; the difference value is obtained by calculating the root mean square error or mean absolute error of the two result sequences; the difference index is set as an error threshold, such as an allowable range of 5% to 10%; when the prediction error exceeds the preset range, model optimization is achieved by adjusting the algorithm parameters or retraining the model. Specifically, after real-time data input, the two models independently compute based on their respective algorithms, generating a sequence of predicted results. The difference value quantifies the degree of deviation between the two sequences using mathematical methods, and the difference index serves as the basis for determining the consistency of the models. The preset error range is determined based on historical data or experimental calibration, for example, set at 8%. If the error is within the range, the first model is marked as valid; if it exceeds the range, a model optimization process is triggered, such as adjusting the model weights using gradient descent or increasing the amount of training data. The optimized model is then re-predicted and compared until the error meets the requirements. Through dynamic verification and adjustment, the accuracy and reliability of the model predictions are ensured, thereby improving the real-time performance and accuracy of anti-skid performance evaluation.

[0063] As a preferred embodiment, the solution of this application is specifically implemented as follows: Real-time collected microscopic texture data of the ice surface and environmental condition data are simultaneously input into the first and second dynamic change models of the ice surface anti-skid performance. The real-time collected data includes surface roughness, texture distribution density, texture depth, ambient temperature, humidity, wind speed, and vehicle traffic frequency.

[0064] The first and second models of dynamic change in the anti-skid performance of the ice surface are used to predict the corresponding dynamic changes in anti-skid performance based on the input data. Specifically, the dynamic changes in anti-skid performance include the first dynamic change result and the second dynamic change result.

[0065] Furthermore, the dynamic changes in the first and second anti-skid performance are compared, and the difference between them is calculated. For example, statistical methods such as root mean square error or mean absolute error can be used to calculate the difference.

[0066] Therefore, a difference index is defined to quantify the difference between the dynamic changes in the first and second anti-skid performance results. The difference index can be set as a percentage, such as 10%. Simultaneously, the prediction errors of the real-time dynamic change models of the first and second ice surface anti-skid performance models are calculated and compared with a preset error range. The preset error range can be set to 5%.

[0067] If the prediction error is within the preset error range, the dynamic change model of the anti-skid performance of the first ice surface is deemed effective. Conversely, if the prediction error exceeds the preset range, the dynamic change models of the anti-skid performance of the first and second ice surfaces are optimized. Optimization methods may include adjusting model parameters, increasing training data, or improving the model structure. After optimization, the above steps are repeated for verification until the prediction error falls within the preset range.

[0068] Through the above technical solution, this application achieves accurate assessment and dynamic monitoring of the anti-skid performance of ice surfaces. By simultaneously using two different models for prediction and comparing the results, the reliability and accuracy of the assessment are improved. Defining difference indicators and preset error ranges provides objective standards for judging the effectiveness of the models. When the prediction error exceeds the range, a mechanism of optimizing the model and repeated verification ensures the continuous improvement and adaptability of the assessment system. This method can timely capture changes in the microstructure of the ice surface, providing road users with more accurate and real-time anti-skid performance information, thereby effectively improving traffic safety.

[0069] In some of the solutions mentioned above in this application, the process of comparing and processing the real-time collected data on the anti-slip performance of the ice surface with the results of dynamic changes has an efficiency bottleneck. A single data processing flow is difficult to handle large-scale data input, resulting in a delay in the generation of anti-slip performance errors and affecting the timeliness of optimization suggestions.

[0070] This application further proposes to establish a data processing grouping mechanism, establish a grouping correspondence between the real-time anti-slip performance data and the dynamic change results of the anti-slip performance of the ice surface and the data processing grouping mechanism, generate data grouping tasks based on the grouping correspondence, and execute the data grouping tasks to perform parallel grouping processing of the real-time anti-slip performance data and the dynamic change results of the anti-slip performance of the ice surface.

[0071] The data processing grouping mechanism divides the data into multiple independent subsets according to preset grouping rules. Each subset contains real-time anti-slip performance data within a specific time window and its corresponding dynamic changes. Grouping relationships are established through data label matching, such as using timestamps or spatial locations as group identifiers. Data grouping tasks are distributed to different computing nodes using a distributed computing framework, with each node performing independent data comparison and error calculation. Parallel processing of the groups employs multi-threading technology, with each thread processing one set of data. The number of threads is dynamically adjusted based on computing resources; for example, the thread pool capacity is set to twice the number of currently available computing cores.

[0072] Specifically, the real-time anti-skid performance data and dynamic changes of the ice surface are first tagged with timestamps and location codes, and then divided into different data groups based on the tagging information. Each data group is encapsulated as an independent task unit and distributed to idle computing nodes via a message queue. The computing nodes perform error calculations on the received data groups, generate local error results, and return them to the central processing module. The central processing module integrates all local error results to generate a global anti-skid performance error report. Through this mechanism, the data processing throughput is increased to eight times that of the original single-threaded mode, and the error generation latency is reduced to the millisecond level, ensuring the real-time output of optimization suggestions.

[0073] As a preferred embodiment, the solution of this application is specifically implemented as follows: A data processing grouping mechanism is established to establish a grouping correspondence between the real-time anti-slip performance data and the dynamic change results of the anti-slip performance of the ice surface and the data processing grouping mechanism. Data grouping tasks are generated based on the grouping correspondence, and the data grouping tasks are executed to process the real-time anti-slip performance data and the dynamic change results of the anti-slip performance of the ice surface in parallel.

[0074] Specifically, the first step is to establish a data processing grouping mechanism. This mechanism divides data into different groups based on factors such as data type, collection time, and geographical location. For example, data can be grouped by hour as a time unit, or by 100-meter road segments as a spatial unit.

[0075] The real-time anti-skid performance data of the ice surface includes parameters such as the coefficient of friction and sliding distance, which are collected in real time through a road surface sensor network. The dynamic changes in anti-skid performance are theoretical values ​​derived from a prediction model.

[0076] Furthermore, a grouping correspondence is established between these two types of data and the data processing grouping mechanism. For example, measured data and prediction results from the same time period or the same road segment are assigned to the same group.

[0077] Therefore, data grouping tasks are generated based on the established grouping correspondence. Each task contains data processing instructions for a specific group, such as data comparison and error calculation.

[0078] Finally, a data grouping task is performed to process the real-time anti-slip performance data and dynamic changes in anti-slip performance of the ice surface in parallel. Distributed computing technology is used, with multiple processing units simultaneously processing different groups of data, improving data processing efficiency.

[0079] Through the above technical solution, this application achieves efficient processing of large amounts of real-time data and prediction results. By grouping data and processing it in parallel, the data processing speed is significantly improved, enabling the system to analyze the dynamic changes in the anti-skid performance of ice surfaces in a timely manner. This method enhances the real-time performance and accuracy of the system's assessment of the anti-skid performance of ice surfaces, providing road users with more timely and accurate safety information support. Simultaneously, the grouping processing mechanism also improves the system's scalability, enabling it to adapt to data processing needs of different scales.

[0080] In some of the solutions mentioned above in this application, the existing technology relies on static measurement methods and lacks the ability to continuously monitor the dynamic changes in the anti-slip performance of the ice surface. This results in insufficient accuracy and timeliness of data collection and analysis, making it difficult to integrate environmental conditions, micro-texture changes and vehicle traffic data in real time, thus affecting the timeliness of traffic safety warnings.

[0081] This application further proposes an anti-skid performance evaluation system for anti-skid textured structures on ice surfaces, including a central server, an environmental monitoring terminal, a road surface sensor network, and user terminal equipment. The central server establishes communication connections with the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment, respectively. The central server integrates and analyzes the data received from the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment, generates optimization suggestions for the anti-skid performance of the ice surface, and transmits them to the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment.

[0082] The environmental monitoring terminal includes temperature sensors, humidity sensors, wind speed sensors, and vehicle traffic frequency detectors. The temperature sensor uses thermistor technology to measure ambient temperature, the humidity sensor uses capacitive humidity sensing to measure humidity, the wind speed sensor uses ultrasonic speed measurement technology to measure wind speed, and the vehicle traffic frequency detector uses geomagnetic induction technology to count vehicle traffic frequency. The road surface sensor network consists of multiple distributed sensor nodes. Each sensor node integrates a surface roughness sensor, a texture distribution density sensor, and a texture depth sensor. The surface roughness sensor uses laser scanning technology to generate three-dimensional topographic data of the ice surface, the texture distribution density sensor calculates texture distribution density based on image recognition technology, and the texture depth sensor uses an ultrasonic probe to measure texture depth. The central server has built-in data integrity detection, data preprocessing, feature extraction, model building, and error analysis modules. User terminal devices connect to the central server via the Internet to receive and display optimization suggestions.

[0083] Specifically, environmental monitoring terminals collect data on ambient temperature, humidity, wind speed, and vehicle traffic frequency, which are transmitted to a central server via communication connections. Distributed sensor nodes in the road sensor network collect data on ice surface roughness, texture distribution density, and texture depth in real time and transmit this data to the central server via the network. The central server's data integrity detection module verifies the integrity of the received data and generates a raw data detection report. The data preprocessing module performs noise reduction, normalization, and format conversion on the data. The feature extraction module extracts surface roughness variation features, texture distribution density variation features, and dynamic environmental condition variation features from the preprocessed data. The model building module constructs a first dynamic variation model and a second dynamic variation model of the anti-skid performance of the ice surface based on the extracted features. The error analysis module compares the real-time anti-skid performance data with the model prediction results, generates anti-skid performance errors, analyzes the time node information corresponding to the errors, and finally generates optimization suggestions. The optimization suggestions are transmitted to user terminal devices for display via the Internet and simultaneously fed back to the environmental monitoring terminal and the road sensor network, achieving closed-loop data control.

[0084] As a preferred embodiment, the solution of this application is specifically implemented as follows: The anti-skid performance evaluation system for ice surface anti-skid textured structures includes a central server, an environmental monitoring terminal, a road surface sensor network, and user terminal equipment. The central server establishes communication connections with the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment.

[0085] The environmental monitoring terminal includes a temperature sensor, a humidity sensor, a wind speed sensor, and a vehicle traffic frequency detector. The temperature sensor uses a PT100 platinum resistance thermometer to measure ambient temperature, with a measurement range of -50℃ to 50℃ and an accuracy of ±0.1℃. The humidity sensor is a capacitive humidity sensor, with a measurement range of 0%RH to 100%RH and an accuracy of ±2%RH. The wind speed sensor is a three-cup anemometer, with a measurement range of 0 m / s to 60 m / s and an accuracy of ±0.3 m / s. The vehicle traffic frequency detector uses geomagnetic induction technology, with a detection range of 6m × 3m and a detection accuracy of 99%.

[0086] The road surface sensor network consists of multiple distributed sensor nodes, spaced 50m apart. Each sensor node includes a surface roughness sensor, a texture distribution density sensor, and a texture depth sensor. The surface roughness sensor uses laser triangulation, with a measurement range of 0.1μm to 1000μm and a resolution of 0.1μm. The texture distribution density sensor uses a CCD image acquisition device with a resolution of 1280×1024 pixels. The texture depth sensor uses an ultrasonic probe, with a measurement range of 0.1mm to 10mm and an accuracy of ±0.1mm.

[0087] The central server employs dual Intel Xeon processors, 128GB of memory, and 20TB of storage. It runs a Linux operating system and utilizes the Apache Hadoop distributed computing framework to process large-scale data. The data integrity verification module uses a checksum algorithm to verify data integrity. The data preprocessing module uses wavelet transform for data denoising and Min-Max normalization for data normalization. The feature extraction module uses principal component analysis to extract key features. The model building module uses a long short-term memory network to construct a dynamic model of the anti-skid performance of the first ice surface and a support vector machine to construct a dynamic model of the anti-skid performance of the second ice surface. The error analysis module uses root mean square error to evaluate model accuracy.

[0088] User terminal devices include smartphones, tablets, or dedicated display devices. These devices establish a communication connection with the central server via 4G / 5G networks, using HTTPS protocol for data transmission to ensure data security. A dedicated application is installed on the terminal device to receive and display suggestions for optimizing the anti-slip performance of the ice surface.

[0089] Through the above technical solution, this application achieves real-time evaluation and optimization suggestion generation of the anti-skid performance of ice surface anti-skid texture structure. The system integrates environmental monitoring, road surface sensing, data processing, and user interaction functions, improving the comprehensiveness and accuracy of data collection. The distributed computing architecture of the central server enhances the processing capability of large-scale data and supports real-time calculation of complex models. The dual-model comparison mechanism improves the reliability of the evaluation results. Real-time data transmission and visualization enable road users to obtain information on the anti-skid performance of ice surfaces in a timely manner, contributing to improved road safety.

[0090] In some of the solutions described above in this application, the environmental monitoring terminal needs to collect environmental condition data such as temperature, humidity, wind speed and vehicle traffic frequency in real time. However, traditional sensors may have problems such as insufficient measurement accuracy or poor environmental adaptability, which limits the accuracy of data collection and thus affects the reliability of the anti-skid performance evaluation model.

[0091] This application further proposes an environmental monitoring terminal including a temperature sensor, a humidity sensor, a wind speed sensor, and a vehicle traffic frequency detector. The temperature sensor uses thermistor technology to measure the ambient temperature, the humidity sensor measures humidity based on the capacitive humidity sensing principle, the wind speed sensor uses ultrasonic speed measurement technology to measure wind speed, and the vehicle traffic frequency detector uses geomagnetic induction technology to count the vehicle traffic frequency.

[0092] Among them, thermistor technology achieves temperature measurement by utilizing the characteristic that resistance changes with temperature, and its linear response characteristics can improve the stability of temperature data; capacitive humidity sensors utilize the principle that humidity changes cause changes in dielectric constant, reflecting humidity data through changes in capacitance, and have low drift characteristics; ultrasonic speed measurement technology calculates wind speed by the time difference between transmitting and receiving sound waves, avoiding interference from mechanical component wear; geomagnetic induction technology counts traffic frequency by detecting changes in the geomagnetic field when vehicles pass by, requiring no external power supply, and is suitable for complex road environments.

[0093] Specifically, the temperature sensor integrates a thermistor element. When the ambient temperature changes, the resistance value changes linearly, which is converted into a voltage signal output by the circuit, ensuring continuous temperature data acquisition. The humidity sensor uses a polymer film as the sensitive medium. Changes in ambient humidity cause a change in the dielectric constant of the film, resulting in a change in capacitance. After processing by the signal conditioning circuit, the humidity parameter is output. The wind speed sensor has a built-in ultrasonic transmitter and receiver. By calculating the time difference of sound wave propagation in the downwind and upwind directions, and combining it with a fluid dynamics model, the wind speed value is calculated, eliminating errors caused by mechanical friction. The vehicle traffic frequency detector is buried under the road surface. It uses a geomagnetic sensor to capture the geomagnetic field disturbances caused by the metal parts of vehicles. Through threshold judgment and counting algorithms, it counts the number of vehicle passages, achieving passive monitoring. The data output from the above sensors is integrated by a central server, providing high-precision input for the dynamic change model of anti-skid performance, improving the accuracy of the evaluation results.

[0094] As a preferred embodiment, the solution of this application is specifically implemented as follows: The environmental monitoring terminal includes a temperature sensor, a humidity sensor, a wind speed sensor, and a vehicle traffic frequency detector. The temperature sensor uses thermistor technology to measure ambient temperature. Specifically, the temperature sensor uses an NTC thermistor, whose resistance decreases as temperature increases; the ambient temperature is determined by measuring the change in resistance. The humidity sensor measures humidity based on the principle of capacitive humidity sensing. Further, the humidity sensor uses a polymer film as the medium; changes in humidity cause changes in capacitance, and the humidity is determined by measuring the capacitance. The wind speed sensor uses ultrasonic speed measurement technology to measure wind speed. The wind speed sensor includes two ultrasonic transducers, and wind speed is calculated by measuring the propagation time of ultrasonic waves in the air. The vehicle traffic frequency detector uses geomagnetic induction technology to count vehicle traffic frequency. Specifically, the vehicle traffic frequency detector buries a geomagnetic sensor under the road surface. When a vehicle passes, the geomagnetic field changes, thereby detecting the vehicle's passage and calculating the number of vehicles passing per unit time.

[0095] Through the above technical solutions, this application achieves comprehensive monitoring of the environmental conditions on the ice surface. The temperature sensor, using thermistor technology, provides accurate temperature data, aiding in the analysis of the impact of temperature changes on the ice surface's anti-skid performance. The humidity sensor, based on capacitive principles, accurately captures humidity changes, providing a basis for assessing the moisture level of the ice surface. The wind speed sensor, employing ultrasonic velocity measurement technology, reliably measures wind speed, helping to analyze the impact of wind speed on the ice surface structure. The vehicle traffic frequency detector, using geomagnetic induction technology, accurately counts vehicle traffic frequency, providing data support for assessing the impact of vehicle rolling on the ice surface's anti-skid performance. Therefore, the environmental monitoring terminal of this application can comprehensively and accurately collect data on environmental factors affecting the anti-skid performance of the ice surface, providing a reliable data foundation for subsequent anti-skid performance assessments.

[0096] In some of the solutions described above in this application, when the road surface sensor network collects micro-texture data of the ice surface, traditional sensor technology has difficulty in accurately capturing the dynamic changes in surface roughness, texture distribution density and depth, resulting in insufficient data accuracy and affecting the accuracy of anti-skid performance evaluation.

[0097] This application further proposes a road surface sensor network comprising multiple distributed sensor nodes. Each sensor node includes a surface roughness sensor, a texture distribution density sensor, and a texture depth sensor. The surface roughness sensor uses laser scanning technology to generate three-dimensional topographic data of the ice surface. The texture distribution density sensor calculates the texture distribution density based on image recognition technology. The texture depth sensor uses an ultrasonic probe to measure the texture depth.

[0098] The surface roughness sensor uses laser scanning technology to emit a high-precision laser beam and generates three-dimensional topographic data based on the displacement changes of the reflected light spot. This data can quantify the microscopic undulations of the ice surface. The texture distribution density sensor, equipped with an optical lens and image processing chip, continuously acquires images of the ice surface, identifies texture boundaries through edge detection algorithms, and counts the number of textures per unit area. The texture depth sensor uses an ultrasonic probe to emit pulse signals to the ice surface and calculates the texture depth value by receiving the time difference of the reflected echoes. Multiple sensor nodes are distributed to cover different areas of the ice surface, improving monitoring efficiency through parallel data acquisition.

[0099] Specifically, laser scanning technology avoids physical interference with the ice surface through non-contact measurement, and the three-dimensional morphology data can accurately reflect the dynamic changes in micro-roughness over time. Image recognition technology, combined with an adaptive illumination compensation algorithm, ensures the accuracy of texture distribution density calculation under different ambient brightness levels. The ultrasonic probe uses high-frequency signals to penetrate the ice layer, eliminating the influence of surface impurities on depth measurement through echo time difference. After data from multiple sensor nodes are integrated by a central server, a dynamic model of micro-texture covering the entire ice surface can be constructed, providing high-precision input parameters for anti-skid performance evaluation. For example, the laser scanning sampling frequency is set to 1000 times per second, the image recognition resolution reaches 0.1 mm / pixel, and the measurement error of the ultrasonic probe is controlled within ±0.05 mm.

[0100] As a preferred embodiment, the solution of this application is implemented as follows: The road surface sensor network comprises twelve distributed sensor nodes, evenly deployed in the sharp bends of urban roads. Each sensor node integrates a laser scanning unit, an image acquisition unit, and an ultrasonic probe. The laser scanning unit uses a 905 nm wavelength laser beam to scan laterally along the ice surface, generating point cloud data containing height information, and reconstructs the three-dimensional shape of the ice surface using a triangulation algorithm. The image acquisition unit is equipped with a 5-megapixel CMOS sensor, capturing ice surface texture images at a rate of 15 frames per second, using an edge detection algorithm to identify texture boundaries, and calculating the number of effective textures per unit area. The ultrasonic probe emits pulse waves at a frequency of 40 kHz, receives the reflected signals from the ice surface, and calculates the texture depth value using the time-of-flight difference. All sensor nodes synchronously transmit data to a central server via the LoRa wireless protocol.

[0101] Through the above technical solutions, this application realizes full-dimensional dynamic monitoring of the micro-texture features of ice surface. Laser scanning technology accurately captures the spatial distribution features of surface roughness, image recognition technology tracks the temporal variation of texture density in real time, and ultrasonic detection technology effectively obtains three-dimensional morphological data of texture depth. The spatiotemporal synchronous acquisition of the three sets of sensor data provides high-precision input parameters for the anti-skid performance evaluation model, solving the problem of evaluation error accumulation caused by the lack of single-dimensional data in traditional methods.

[0102] In some of the solutions described above in this application, the road surface sensor network needs to collect micro-texture data of the ice surface. However, in practical applications, traditional sensors may not be able to accurately capture the dynamic changes in the three-dimensional morphology, texture distribution density, and depth of the ice surface, resulting in insufficient accuracy and completeness of data acquisition and affecting the reliability of subsequent anti-skid performance evaluation. This application further proposes a road surface sensor network comprising multiple distributed sensor nodes. Each sensor node includes a surface roughness sensor, a texture distribution density sensor, and a texture depth sensor. The surface roughness sensor uses laser scanning technology to generate three-dimensional topographic data of the ice surface. The texture distribution density sensor calculates the texture distribution density based on image recognition technology. The texture depth sensor uses an ultrasonic probe to measure the texture depth. The system comprises several sensors: a surface roughness sensor projects a laser beam onto the ice surface using laser scanning technology, receives the reflected light signal to generate 3D point cloud data, and then constructs the 3D topography of the ice surface; a texture distribution density sensor acquires images of the ice surface using a high-resolution camera, identifies texture boundaries using an edge detection algorithm, and calculates the texture distribution density per unit area based on pixel density; and a texture depth sensor emits ultrasonic pulses onto the ice surface using an ultrasonic probe and calculates the texture depth based on the reflection time difference. Multiple sensor nodes are distributed to cover different areas of the ice surface, ensuring comprehensive data acquisition. Specifically, laser scanning technology captures the microscopic undulations of the ice surface using high-precision optical elements, generating three-dimensional morphological data that quantifies surface roughness parameters. Image recognition technology extracts texture features through grayscale thresholding and combines morphological processing to eliminate noise interference, ensuring the accuracy of texture distribution density calculation. Ultrasonic probes penetrate the ice surface with high-frequency sound waves, accurately measuring texture depth based on the time delay and intensity changes of the echo signal. Distributed sensor nodes transmit surface roughness, texture distribution density, and depth data to a central server in real time via a synchronous data acquisition protocol. This data, combined with temperature, humidity, and wind speed data collected by environmental monitoring terminals, supports the construction and updating of a dynamic model of anti-skid performance. Through the collaborative work of multiple sensors, the dynamic changes in the microscopic texture of the ice surface are fully captured, providing high-precision input for anti-skid performance evaluation, thereby improving the real-time performance and reliability of the evaluation results.

[0103] As a preferred embodiment, the solution of this application is implemented as follows: The central server comprises five functional modules that operate collaboratively. The data integrity detection module performs integrity scanning on the received ice surface roughness, texture depth, and ambient temperature data using a preset data field verification algorithm, and automatically generates a detection report containing markers of missing data locations. The data preprocessing module uses a wavelet transform algorithm to denoise the three-dimensional morphology data of the ice surface, converts humidity and wind speed data of different dimensions into dimensionless values ​​using a range normalization method, and uniformly converts the texture distribution data in image format into a matrix data structure. The feature extraction module extracts three key feature vectors from the preprocessed time-series data based on principal component analysis: the surface roughness variation coefficient, the texture distribution dispersion index, and the temperature-humidity coupling change rate. The model building module trains the anti-skid performance prediction model using a support vector machine algorithm, and simultaneously builds a validation model based on a random forest algorithm. The two models share the feature vector input layer but have independent hidden layer structures. The error analysis module calculates the deviation between the predicted value and the measured anti-slip coefficient using residual analysis. When the average residual of five consecutive time nodes exceeds the threshold, it triggers a model parameter optimization instruction and generates texture maintenance suggestions.

[0104] Through the above technical solution, this application achieves fully automated processing of ice surface anti-skid performance data, effectively solving the error accumulation problem caused by the fragmented data processing in traditional methods. The collaborative operation of each functional module ensures the accuracy of feature extraction and the timeliness of model updates. Specifically, the data integrity detection module avoids invalid data input through a dynamic verification mechanism, the feature extraction module uses dimensionality reduction technology to improve the identification of key features, the dual-model architecture of the model building module enhances the reliability of prediction results, and the residual monitoring mechanism of the error analysis module enables the self-optimization function of the evaluation system. This technical solution improves the response speed of ice surface anti-skid performance evaluation to the minute level, and the prediction accuracy reaches engineering application standards, providing precise technical support for road safety maintenance.

[0105] In some of the solutions mentioned above in this application, the information transmission method between the user terminal device and the central server has problems of insufficient real-time performance and poor display adaptability, which makes it impossible to present the suggestions for optimizing the anti-slip performance of the ice surface to road managers or drivers in a timely and intuitive manner, affecting the timeliness of traffic safety decisions.

[0106] This application further proposes that the user terminal device includes a smartphone, tablet computer or dedicated display device, which establishes a communication connection with the central server via the Internet to receive and display suggestions for optimizing the anti-slip performance of the ice surface.

[0107] The user terminal devices employ various hardware forms to adapt to different application scenarios. Smartphones and tablets receive data via mobile networks and support touch interaction; dedicated display devices are equipped with fixed installation interfaces and high-resolution screens, suitable for large-screen monitoring in traffic control centers. Communication connections are based on the Internet protocol to achieve bidirectional data transmission. The central server encapsulates optimization suggestions into structured data packets and pushes them to the terminal devices via HTTP or MQTT protocols. The display function is implemented through the terminal device's built-in graphics rendering engine, presenting optimization suggestions in the form of visual charts overlaid with text descriptions.

[0108] Specifically, the optimization suggestions generated by the central server are encrypted and transmitted to user terminal devices via the internet. Smartphones use a dedicated application to parse the data and display a trend chart of anti-skid performance changes and a list of maintenance suggestions on the interactive interface; tablets use a split-screen mode to simultaneously display real-time data comparisons from multiple monitoring points. Dedicated display devices connect to a large-size display screen via a video output interface to map the anti-skid performance error distribution of different areas of the ice surface in the form of a dynamic heat map. For example, when the texture depth of a certain section is detected to be below the safety threshold, a red warning box automatically pops up on the terminal device interface, and a voice prompt is generated. The display module automatically adjusts the chart ratio and font size according to the device screen size to ensure readability of information on different terminals. The data refresh frequency is synchronized with the model update cycle of the central server, achieving real-time matching between the anti-skid performance evaluation results and the actual on-site conditions.

[0109] As a preferred embodiment, the solution of this application is implemented as follows: The user terminal device adopts an industrial-grade touch screen with wireless communication capabilities. This device has a built-in dedicated data parsing module and graphics rendering engine, and establishes a two-way encrypted communication channel with the central server via HTTPS protocol. When the central server generates anti-skid performance optimization suggestions, the data packet is transmitted to the user terminal device via the Internet. The data parsing module decompresses the timestamp information, risk level parameters, and geographic coordinates in the optimization suggestions. The graphics rendering engine generates a visualization interface based on the decompressed data, where the risk level is displayed as a color-gradient heatmap overlaid on an electronic map. For example, on the monitoring screen of a highway maintenance center, operators can view the spatiotemporal evolution trend of the anti-skid performance of ice surfaces on different road sections in real time. Simultaneously, the device supports multi-touch operation to retrieve historical data for comparative analysis.

[0110] Through the above technical solution, this application solves the technical problems of traditional evaluation systems having a single information transmission channel and insufficient real-time display capabilities, and realizes efficient analysis and visualization of ice surface anti-skid performance optimization suggestions from multiple types of terminal devices. Specifically, the collaborative work of dedicated display devices and mobile terminals ensures the reliability of data reception in different application scenarios. The encrypted communication mechanism effectively avoids the risk of data tampering during remote transmission, and the graphical interface design further improves the understandability of complex anti-skid performance data and the efficiency of decision support.

[0111] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0112] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the anti-skid performance of an ice surface anti-skid texture structure, characterized by, Includes the following steps: S1. Acquire micro-texture data of the ice surface and environmental condition data. The micro-texture data of the ice surface includes surface roughness, texture distribution density, and texture depth. The environmental condition data includes ambient temperature, humidity, wind speed, and vehicle traffic frequency. S2. Classify the micro-texture data of the ice surface and environmental condition data to obtain classification results. Based on the classification results, establish a data partitioning model and deploy an anti-skid performance evaluation mirror system on a distributed computing platform. In the data partitioning model, preprocess and extract features from the corresponding micro-texture data of the ice surface and environmental condition data to extract surface roughness variation features, texture distribution density variation features, and dynamic environmental condition variation features. S3. Retrieve historical data of the ice surface, combine surface roughness variation characteristics, texture distribution density variation characteristics, and environmental condition dynamic variation characteristics to construct a first dynamic variation model of the anti-slip performance of the ice surface, and update the model in real time. In the anti-slip performance evaluation mirror system, use the micro-texture data of the ice surface and environmental condition data to train the support vector machine model to construct a second dynamic variation model of the anti-slip performance of the ice surface. In the anti-slip performance evaluation mirror system, update the second dynamic variation model of the anti-slip performance of the ice surface in real time. Compare the prediction results of the first dynamic variation model of the anti-slip performance of the ice surface with those of the second dynamic variation model of the anti-slip performance of the ice surface. If the difference in the prediction results is within the preset range, the first dynamic variation model of the anti-slip performance of the ice surface takes effect. S4. Substitute the real-time collected micro-texture data of the ice surface and environmental condition data into the dynamic change model of the anti-slip performance of the first ice surface to obtain the dynamic change result of the anti-slip performance of the ice surface. Collect the real-time anti-slip performance data of the ice surface, compare the real-time anti-slip performance data with the dynamic change result, and generate the anti-slip performance error. S5, analyze the anti-slip performance error, extract the time node information corresponding to the error, establish the correlation between the time node information and the anti-slip performance evaluation index of the ice surface, and generate optimization suggestions for the anti-slip performance of the ice surface.

2. The method of evaluating the skid-resistance performance of an ice-surface skid- resistant texture construction according to claim 1, wherein Step S1 also includes the following steps: The system detects the integrity of microscopic texture data and environmental condition data on the ice surface. The detection content includes surface roughness, texture distribution density, texture depth, ambient temperature, humidity, wind speed, and vehicle traffic frequency. A raw data detection report is generated, which contains information on qualified and missing data. The raw data detection report is compared with historical data of the ice surface. If the data types are consistent, the microscopic texture data and environmental condition data of the ice surface are determined to match the historical data. If the data types are inconsistent, the discrepancies are identified and the data types are marked, generating discrepancy data category information. New data detection rules are established based on the information of the difference data categories. The data corresponding to the new data detection rules and the data corresponding to the original data detection report are substituted into the preset digital twin simulation model to generate dynamic evolution images of the micro-texture of the ice surface. The key point coordinate information in the dynamic evolution image of the micro-texture of the ice surface is obtained. The accuracy of the key point coordinate information is verified by using the known coordinates of the basic points on the ice surface. If the key point coordinate information is verified, the micro-texture data of the ice surface and the environmental condition data are marked as valid data.

3. The method of evaluating the skid-resistance performance of an ice-surface skid- resistant texture construction according to claim 1, wherein Step S2 also includes the following steps: Verify whether the data meets the predetermined classification criteria, which include data category, environmental condition type and timestamp label. Perform preprocessing operations on the classified data, including data denoising, data normalization and data format conversion. Receive the anti-slip performance evaluation request, determine the type of features to be extracted based on the evaluation request, and use time series feature analysis to extract the selected features from the preprocessed data. The effectiveness of the extracted features is verified using historical data or known anti-skid performance evaluation results. Based on the verification results, the feature extraction strategy is adjusted, including modifying the feature type and optimizing the feature extraction algorithm.

4. The method of evaluating the slip resistance of an ice surface texture according to claim 1, wherein Step S3 also includes the following steps: The real-time collected micro-texture data of the ice surface and environmental condition data are simultaneously input into the first dynamic change model of the anti-slip performance of the ice surface and the second dynamic change model of the anti-slip performance of the ice surface. The first dynamic change model of the anti-skid performance of the ice surface and the second dynamic change model of the anti-skid performance of the ice surface are predicted based on the input data, respectively, and the corresponding dynamic change results of the anti-skid performance are generated. The dynamic change results of the anti-skid performance include the first dynamic change results of the anti-skid performance and the second dynamic change results of the anti-skid performance. Compare the dynamic changes of the first and second anti-skid performance, and calculate the difference between them. Define a difference index to quantify the difference between the dynamic change results of the first anti-slip performance and the dynamic change results of the second anti-slip performance. Calculate the prediction error of the real-time dynamic change model of the first ice surface anti-slip performance and the dynamic change model of the second ice surface anti-slip performance, and compare the prediction error with the preset error range. If the prediction error is within the preset error range, the dynamic change model of the anti-slip performance of the first ice surface is deemed to be effective. If the prediction error exceeds the preset range, the dynamic change model of the anti-slip performance of the first ice surface and the dynamic change model of the anti-slip performance of the second ice surface will be optimized, and the above steps will be repeated for verification.

5. The method of evaluating the slip resistance of an ice surface texture according to claim 1, wherein Step S4 also includes the following steps: A data processing grouping mechanism is established to establish a grouping correspondence between the real-time anti-slip performance data and the dynamic change results of the anti-slip performance of the ice surface and the data processing grouping mechanism. Data grouping tasks are generated based on the grouping correspondence, and the data grouping tasks are executed to process the real-time anti-slip performance data and the dynamic change results of the anti-slip performance of the ice surface in parallel.

6. An anti-skid performance evaluation system for an ice surface anti-skid texture structure, which applies the anti-skid performance evaluation method for an ice surface anti-skid texture structure according to any one of claims 1 to 5, characterized by, It includes a central server, an environmental monitoring terminal, a road surface sensor network, and user terminal equipment. The central server establishes communication connections with the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment respectively. The central server integrates and analyzes the data received from the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment, generates optimization suggestions for the anti-skid performance of the ice surface, and transmits them to the environmental monitoring terminal, the road surface sensor network, and the user terminal equipment.

7. The system for evaluating the skid-resisting performance of an ice-surface skid-resisting texture structure according to claim 6, wherein The environmental monitoring terminal includes a temperature sensor, a humidity sensor, a wind speed sensor, and a vehicle traffic frequency detector. The temperature sensor uses thermistor technology to measure the ambient temperature, the humidity sensor measures humidity based on the capacitive humidity sensing principle, the wind speed sensor uses ultrasonic speed measurement technology to measure wind speed, and the vehicle traffic frequency detector uses geomagnetic induction technology to count the vehicle traffic frequency.

8. The anti-slip performance evaluation system for the anti-slip texture structure on the ice surface as described in claim 6, characterized in that, The road surface sensor network consists of multiple distributed sensor nodes. Each sensor node includes a surface roughness sensor, a texture distribution density sensor, and a texture depth sensor. The surface roughness sensor uses laser scanning technology to generate three-dimensional topographic data of the ice surface. The texture distribution density sensor calculates the texture distribution density based on image recognition technology. The texture depth sensor uses an ultrasonic probe to measure the texture depth.

9. The anti-slip performance evaluation system for the anti-slip texture structure on the ice surface as described in claim 6, characterized in that, The central server includes a data integrity detection module, a data preprocessing module, a feature extraction module, a model building module, and an error analysis module. The data integrity detection module is used to detect data integrity and generate a raw data detection report. The data preprocessing module is used to perform noise reduction, normalization, and format conversion on the data. The feature extraction module is used to extract surface roughness variation features, texture distribution density variation features, and dynamic environmental condition variation features. The model building module is used to construct a dynamic variation model of the anti-slip performance of the first ice body surface and a dynamic variation model of the anti-slip performance of the second ice body surface. The error analysis module is used to analyze the anti-slip performance error and generate optimization suggestions.

10. The anti-slip performance evaluation system for the anti-slip texture structure on the ice surface as described in claim 6, characterized in that, User terminal devices include smartphones, tablets, or dedicated display devices. These devices establish a communication connection with a central server via the Internet to receive and display suggestions for optimizing the anti-slip performance of the ice surface.

Images