Maintenance decision-making method and system for semi-rigid base asphalt pavement of expressway

Abstract

The invention relates to the technical field of pavement maintenance, in particular to a highway semi-rigid base asphalt pavement maintenance decision-making method and system. The method comprises the following steps: collecting different types of monitoring data of a target highway; performing dimension reduction processing on the monitoring data, and predicting a target expressway maintenance demand based on the dimension-reduced data and a prediction model to obtain a prediction result of the target expressway maintenance demand; and an optimal maintenance strategy is formulated and executed by integrating the prediction results of the current available resources and the target expressway maintenance demand. The maintenance effect of the semi-rigid base asphalt of the expressway pavement is improved, and the quality of the expressway is guaranteed.

Description

Technical Field

[0001] The invention relates to the technical field of pavement maintenance, and in particular to a method and system for making decisions on semi-rigid base asphalt pavement maintenance for expressways. Background Art

[0002] With the rapid development of the national economy and the continuous increase in traffic flow, highway infrastructure is facing severe tests. The traditional pavement maintenance method mainly relies on manual inspection and experience judgment. This method is not only time-consuming and labor-intensive, but also difficult to make accurate and timely maintenance decisions due to the lack of systematic data support. In addition, due to the lack of an effective early warning mechanism, problems can often be discovered and dealt with only after they become serious, which not only increases maintenance costs, but may also lead to traffic safety accidents. Although there are some pavement maintenance management systems on the market, most of the systems have a single function and cannot achieve comprehensive monitoring and intelligent analysis of the internal conditions of the pavement structure, nor can they dynamically adjust maintenance strategies according to actual conditions. Therefore, there is an urgent need for a more efficient and intelligent highway maintenance decision-making method and technical means to improve the efficiency and quality of maintenance work.

[0003] The existing highway maintenance semi-rigid base asphalt pavement maintenance decision-making technology has the following main problems: first, the data collection means are limited, and it is difficult to obtain comprehensive and continuous road condition information; second, the data processing capacity is insufficient, and it is impossible to effectively extract useful information from massive data. These two problems cause maintenance work to often lag behind actual needs and fail to respond to changes in road conditions in a timely manner, which in turn affects maintenance quality and efficiency. Therefore, there is an urgent need for a road maintenance solution that can integrate multiple sensing technologies and utilize big data analysis and intelligent decision-making algorithms. Summary of the invention

[0004] In order to solve the problem that the existing method has poor decision-making effect in maintaining semi-rigid base asphalt pavement of highway, which affects the maintenance quality, the purpose of the present invention is to provide a maintenance decision-making method and system for semi-rigid base asphalt pavement of highway, and the technical scheme adopted is as follows:

[0005] In a first aspect, the present invention provides a method for making a maintenance decision for a semi-rigid base asphalt pavement of a highway, the method comprising the following steps:

[0006] Collect different types of monitoring data of the target highway;

[0007] Performing dimensionality reduction processing on the monitoring data, and predicting the maintenance demand of the target expressway based on the dimensionality reduced data and the prediction model to obtain a prediction result of the maintenance demand of the target expressway;

[0008] The optimal maintenance strategy is formulated and implemented by integrating the current available resources and the forecast results of the maintenance demand of the target highway.

[0009] Preferably, after collecting different types of monitoring data of the target highway, the method further includes: preprocessing the monitoring data, wherein the preprocessing includes data cleaning, missing value filling, outlier detection and correction, data normalization and data standardization.

[0010] Preferably, the step of predicting the maintenance demand of the target expressway based on the dimensionally reduced data and the prediction model to obtain a prediction result of the maintenance demand of the target expressway includes:

[0011] The prediction model is trained using sample monitoring data of the target highway to obtain a trained prediction model; wherein, during the training process, samples and features are randomly selected to construct multiple decision trees; for the regression task, the final output is the average value of the prediction values ​​of all decision trees; the prediction model is constructed using a random forest algorithm;

[0012] The dimensionally reduced data is input into a trained prediction model to obtain a prediction result of the maintenance demand of the target expressway.

[0013] Preferably, the step of formulating and executing an optimal maintenance strategy based on the forecast results of the currently available resources and the maintenance demand of the target highway includes:

[0014] Determine the location, type and time of maintenance for the target expressway based on the forecast results of the maintenance demand of the target expressway;

[0015] The optimal maintenance strategy is formulated and executed based on the location, type, time and currently available resources of the target highway to be maintained, wherein the currently available resources include funds, manpower and equipment.

[0016] Preferably, after executing the optimal maintenance strategy, the method further includes: if problems occur during the execution, adjusting the maintenance strategy.

[0017] Preferably, a PCA algorithm is used to perform dimensionality reduction processing on the monitoring data.

[0018] Preferably, the collection of different types of monitoring data of the target highway includes:

[0019] Sensors are installed inside the target highway structure, and the sensors are used to collect different types of monitoring data of the highway; wherein the sensors include temperature sensors, humidity sensors, strain gauges, earth pressure gauges, multi-layer settlement displacement gauges, crack detection sensors and pothole detection sensors, and the interior of the structure includes the road surface, the inside of the surface layer, the inside of the water-stable base layer and the top of the roadbed.

[0020] In a second aspect, the present invention provides a highway semi-rigid base asphalt pavement maintenance decision system, the system comprising:

[0021] Data collection module, used to collect different types of monitoring data of the target highway;

[0022] A data processing module is used to perform dimensionality reduction processing on the monitoring data, and predict the maintenance demand of the target expressway based on the data after dimensionality reduction and the prediction model to obtain a prediction result of the maintenance demand of the target expressway;

[0023] The resource scheduling module is used to formulate and execute the optimal maintenance strategy based on the current available resources and the forecast results of the maintenance demand of the target highway.

[0024] The present invention has at least the following beneficial effects:

[0025] The present invention first collects different types of monitoring data of the target expressway, then performs dimensionality reduction processing on the collected monitoring data, and uses a prediction model to predict the maintenance needs of the target expressway, and further integrates the current available resources and the predicted results of the maintenance needs of the target expressway to formulate and execute the optimal maintenance strategy. This process enables the changes in road conditions to be captured in a timely manner, so that responses can be made at the early stage of problems, avoiding cost increases and safety hazards caused by delayed maintenance, improving the maintenance effect of semi-rigid base asphalt on the highway pavement, and ensuring the quality of the highway. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0027] Figure 1 The present invention provides a flowchart of a method for making decisions on maintenance of semi-rigid base asphalt pavement of a highway according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, a method and system for maintaining a semi-rigid base asphalt pavement of a highway proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments.

[0029] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0030] The following is a detailed description of a method and system for maintaining a semi-rigid base asphalt pavement of a highway provided by the present invention in conjunction with the accompanying drawings.

[0031] An embodiment of a maintenance decision method for semi-rigid base asphalt pavement of a highway:

[0032] This embodiment proposes a maintenance decision method for semi-rigid base asphalt pavement of highway. Figure 1 As shown, a highway semi-rigid base asphalt pavement maintenance decision method of this embodiment includes the following steps:

[0033] Step S1, collecting different types of monitoring data of the target highway.

[0034] First, different types of sensors are installed inside the structure of the target highway. The sensors collect different types of monitoring data of the target highway in real time through built-in sensing elements. It should be noted that the target highway in this embodiment is a highway to be tested with a semi-rigid base asphalt pavement. In this embodiment, the installed sensors include temperature sensors, humidity sensors, strain gauges, earth pressure gauges, multi-layer settlement displacement gauges, crack detection sensors, and pothole detection sensors. The interior of the structure includes the road surface, the inside of the surface layer, the inside of the water-stable base layer, and the top of the roadbed. These locations usually bear a large traffic load or are easily affected by environmental factors. Specifically, the pavement temperature sensor is used to measure the temperature change of the pavement to evaluate the thermal stress of the pavement and prevent pavement damage caused by excessive temperature changes. The humidity sensor detects the pavement humidity of the target highway to prevent the risk of pavement sliding caused by excessive humidity or damage caused by water infiltration into the roadbed. The strain gauge is arranged at the bottom of each layer of the asphalt surface layer and the semi-rigid base layer to detect the horizontal direction of each structural layer to prevent excessive strain in the structural layer and cause interlayer bonding failure. The settlement displacement meter detects the vertical displacement of the pavement structure as a whole to prevent excessive vertical deformation. The soil pressure gauge is arranged in each structural layer to observe the vertical stress of the layer and understand the stress state of the structural layer. The crack detection sensor monitors whether there are cracks in the pavement, promptly detects damage to the pavement structure, and prevents the cracks from spreading and causing larger-scale damage. The pothole detection sensor detects whether there are potholes in the road surface to ensure driving safety and reduce accidents caused by vehicles avoiding potholes. Embedded mounting bracket: According to the type and size of the sensor, fix the mounting bracket in the selected position in advance to ensure that the sensor is installed firmly and is not easily affected by external forces; fix the sensor on the mounting bracket to ensure that the direction of the sensor is correct and the required data can be accurately collected; pay attention to waterproof and dustproof treatment during installation. After the sensor is installed, debug and calibrate the sensor to ensure that the collected data is accurate. As other implementation methods, the type of sensor and the installation location of the sensor can be selected according to the specific situation. This embodiment sets the collection frequency of monitoring data to once a minute, that is, all types of monitoring data are collected once every minute. In specific applications, the implementer can set the collection frequency of monitoring data according to the specific situation.

[0035] At this point, different types of monitoring data of the target highway have been obtained, and the acquired monitoring data has been transmitted to the central server through either wireless or wired methods. Wired transmission: directly connected to the central server through optical fiber or cable, suitable for short-distance transmission or scenarios where communication lines have been laid; wireless transmission: using IoT technologies such as LoRa and NB-IoT to achieve remote data transmission, suitable for long-distance transmission or scenarios where it is inconvenient to lay cables; wireless transmission solutions need to consider factors such as signal strength and data encryption to ensure the security and reliability of data transmission.

[0036] The monitoring data collected by the sensor needs to be converted into a unified format for the central server to receive and parse. It can be converted into data formats such as JSON and XML. The MQTT protocol has the characteristics of lightweight and low bandwidth, and is more suitable for IoT application scenarios. Therefore, the HTTP or MQTT protocol in the TCP / IP protocol family is used to transmit data to ensure the orderly transmission and security of data.

[0037] Step S2, performing dimensionality reduction processing on the monitoring data, and predicting the maintenance demand of the target expressway based on the dimensionality reduced data and the prediction model to obtain a prediction result of the maintenance demand of the target expressway.

[0038] The data set involved in this embodiment contains a large number of observation samples, each of which contains multiple features. First, the data needs to be preprocessed to ensure the effectiveness of subsequent PCA operations; all types of monitoring data collected from the target highway are preprocessed, and the preprocessing includes data cleaning, missing value filling, outlier detection and correction, data normalization and standardization processing to eliminate the influence of dimension. Data preprocessing is a prior art and will not be described in detail here.

[0039] Next, the principal component analysis algorithm (PCA) will be used to reduce the dimension of the pre-processed monitoring data. Specifically, the covariance matrix is ​​calculated using the pre-processed monitoring data. The covariance matrix is ​​used to reflect the linear relationship between features. The covariance matrix is ​​decomposed by eigenvalue to obtain eigenvalues ​​and corresponding eigenvectors. The eigenvalues ​​are arranged in descending order, and the eigenvectors corresponding to the first k largest eigenvalues ​​are selected as the principal components. The basis for selecting k is the cumulative contribution rate. The goal is to select the smallest k so that the cumulative contribution rate reaches a certain threshold. The threshold can be 80% or 90%, which is set by the implementer according to the specific situation. The specific calculation formula for the cumulative contribution rate is:

[0040]

[0041] Among them, Cumulative Variance represents the cumulative contribution rate, λ i represents the i-th eigenvalue, k represents the number of eigenvalues, that is, the principal components corresponding to the first k largest eigenvalues ​​selected, m represents the total number of features, that is, the total number of features in the original data set, and λ j represents the jth eigenvalue.

[0042] Project the original data into the space formed by the first k eigenvectors to obtain the reduced-dimensional data, namely:

[0043] Y=X norm V k

[0044] Among them, Y represents the data after dimensionality reduction, X norm Represents the original data, V k represents the matrix consisting of the first k eigenvectors.

[0045] Random Forests is an integrated learning method that constructs multiple decision trees for classification or regression and determines the final result by voting or averaging. Compared with a single decision tree, Random Forests can reduce the risk of overfitting and improve the accuracy and stability of the model.

[0046] Next, the reduced-dimensional data is used to construct an isolation forest to predict the maintenance needs of the target highway. Specifically, the prediction model is trained using sample monitoring data of the road to obtain a trained prediction model. During the training process, samples and features are randomly selected to construct multiple decision trees. For regression tasks, the final output is the average of the prediction values ​​of all decision trees. The prediction model is constructed using a random forest algorithm.

[0047] Among them, model training and evaluation specifically include:

[0048] 1. Standardized data: Use the training set data to calculate the mean and standard deviation, and standardize the data;

[0049] 2. PCA transformation: Apply the PCA algorithm to obtain the data after dimensionality reduction;

[0050] 3. Feature importance: evaluate the importance of principal components corresponding to different eigenvalues;

[0051] 4. Visualization: Draw a scatter plot of the data before and after dimensionality reduction to observe the changes in data distribution;

[0052] 5. Dimensionality reduction effect evaluation: Calculate the correlation of data before and after dimensionality reduction to test the effectiveness of PCA;

[0053] 6. Downstream task evaluation: Further train the classifier and evaluate its performance on the test set.

[0054] (1) Data preparation

[0055] The sample data set contains N samples, each sample has M features, and the sample data set is represented as an N×M matrix. The sample data set is divided into a training set Dtrain and a test set Dtest, for example, in a ratio of 70% / 30%. If the features in the data set have different dimensions or types (such as numerical and categorical), the data needs to be standardized or encoded.

[0056] (2) Constructing a random forest

[0057] The construction process of random forest mainly includes the following steps:

[0058] Step 1: Generate B sub-datasets from the original data set Dtrain by bootstrap sampling. Each sub-dataset is used to train a decision tree. The size of each sub-dataset is usually equal to the size of the original data set. However, due to random sampling, some samples may be selected multiple times, while some samples will not be selected.

[0059] Step 2: For each sub-dataset b∈[1, B], in the process of building a decision tree, a part of the features are randomly selected at each node as candidate features for splitting, and then the best splitting features and splitting points are selected to minimize the classification error or regression error;

[0060] The specific steps are as follows:

[0061] 1. Feature selection: At each node, randomly select p features from all features as candidate features for splitting, where p <M;

[0062] 2. Optimal split point selection: For the selected features, calculate the purity (such as information gain or Gini index) of each possible split point; select the split point that increases the purity the most;

[0063] 2.1. The Gini index is defined as Among them, G is the Gini index, p c is the probability of category c appearing in a node. The lower the Gini index, the purer the node.

[0064] 2.2. Information gain is defined as IG(T,A)=Entropy(T)-Entropy(T|A), where Entropy(T) is the entropy of the node, Entropy(T|A) is the entropy after splitting according to feature A, and IG(T,A) represents information gain;

[0065] 3. Split node: Use the best split point to divide the data set into two subsets, and continue splitting recursively until the stopping condition is met, such as reaching a predetermined maximum depth or the number of samples in the node is less than a certain threshold.

[0066] Step 3: The growth of the decision tree can be based on one of the following stopping conditions:

[0067] reaching a predetermined maximum depth;

[0068] The number of samples in a node is less than a certain threshold;

[0069] For classification tasks, all samples belong to the same category;

[0070] No further splitting is possible, i.e. all features cannot further reduce impurity.

[0071] (3) Prediction results

[0072] For classification tasks, the final prediction result is the voting result of the prediction results of most decision trees; for regression tasks, the final prediction result is the average of the prediction values ​​of all decision trees.

[0073] (4) The experimental process includes:

[0074] Step 1: Dataset division

[0075] Divide the sample data set into a training set and a test set, for example, in a ratio of 70% / 30%, use the training set data to train the random forest model, and evaluate the performance of the model on the test set;

[0076] Step 2: Model training

[0077] Sample selection: Generate multiple sub-datasets from the training set through bootstrap sampling;

[0078] Construct decision tree: For each sub-dataset, build a decision tree using random feature selection and optimal split point selection;

[0079] Stop condition: Set the maximum depth of the decision tree or the minimum number of samples in a node as the stopping condition;

[0080] Step 3: Model Evaluation

[0081] Classification task evaluation: Use indicators such as accuracy, recall, and F1 score to evaluate model performance;

[0082] Regression task evaluation: Use indicators such as mean square error (MSE) and mean absolute error (MAE) to evaluate model performance;

[0083] Step 4: Feature Importance Evaluation

[0084] Randomize features: For each feature, randomize the value of that feature and re-evaluate model performance;

[0085] Calculate performance drop: Calculate the drop in model performance after randomizing features as a measure of feature importance.

[0086] So far, this embodiment has obtained the prediction results of the maintenance demand of the target highway using the prediction model.

[0087] Step S3, formulate and execute the optimal maintenance strategy based on the current available resources and the forecast results of the maintenance demand of the target highway.

[0088] In the above steps, the forecast results of the maintenance demand of the target highway are determined. Next, the location, type and time of the target highway to be maintained are determined based on the forecast results of the maintenance demand of the target highway. Combined with the location, type, time of the target highway to be maintained and the currently available maintenance resources, the optimal maintenance strategy is formulated and implemented, where the currently available resources include funds, manpower and equipment.

[0089] Among them, the resource scheduling process of the resource scheduling module includes:

[0090] Resource Allocation:

[0091] Based on maintenance decisions, determine the priority of maintenance tasks, and allocate personnel, equipment and materials accordingly. Use scheduling algorithms (such as greedy algorithms and genetic algorithms) to allocate resources to ensure maximum resource utilization;

[0092] Execution monitoring:

[0093] Use GPS tracking devices and mobile applications to monitor the execution of maintenance plans, ensure that resources are in place on time, and provide real-time feedback on the execution status. If there are any deviations, make immediate adjustments;

[0094] Adjustment strategy:

[0095] If problems are encountered during the execution process, such as sudden weather changes or equipment failures, the maintenance strategy should be adjusted in a timely manner. After the adjustment, resources should be reallocated to ensure the smooth completion of the maintenance task.

[0096] Specifically, the interface design of the user interaction module uses Web technology (HTML / CSS / JavaScript) to develop user-friendly web applications; uses Django or Flask framework to build back-end services to process front-end requests and interact with the database; and designs intuitive user interfaces to display road conditions, maintenance decisions, and resource scheduling.

[0097] The user interaction module can realize data visualization functions, such as map marking, table display, line chart, etc., so that users can clearly understand the current situation; provide a parameter adjustment interface, allowing users to enter specific parameters to adjust maintenance strategies; set up permission management to ensure that only authorized users can access sensitive information.

[0098] The core operating principle of the highway semi-rigid base asphalt pavement maintenance decision-making method provided in this embodiment is to integrate multi-source sensor data and support intelligent maintenance decisions through advanced data processing and analysis technology. First, various sensors deployed at key locations on the highway are used to obtain real-time highway status information. These sensors can detect parameters including but not limited to road surface temperature, humidity level, crack conditions, and the presence or absence of potholes, and transmit the collected data to the central server via wired or wireless means. During the data transmission process, Internet of Things technologies such as linear frequency modulation spread spectrum modulation technology (LongRange, LoRa) and narrowband Internet of Things technology based on cellular networks (Narrow Band Internet of Things, NB-IoT) are used to ensure the secure transmission of data. At the same time, factors such as signal strength and data encryption are taken into account to ensure the reliability and security of the data. Once the data arrives at the central server, it enters the preprocessing stage. The work in this stage includes steps such as data cleaning, missing value filling, outlier detection and correction, and data normalization to ensure data quality and lay a solid foundation for subsequent data analysis. In addition, the principal component analysis (PCA) algorithm is also applied to reduce the dimension of the data, which not only improves the data processing efficiency, but also retains the key information of the data.

[0099] The preprocessed data is then fed into the data analysis module, which uses the random forest machine learning algorithm to analyze the data and predict future road maintenance needs. It builds multiple decision trees by randomly selecting samples and features, and uses majority voting or averaging to arrive at the final prediction result. This process enables the system to respond to changes in road conditions at an early stage, effectively avoiding the additional costs and safety hazards caused by untimely maintenance.

[0100] According to the prediction results, the decision generation module combines the existing maintenance resources, such as funds, manpower, equipment, etc., to formulate the optimal maintenance strategy. The resource scheduling module schedules the corresponding maintenance resources according to the generated decision, and tracks the progress of the maintenance work in real time through the execution monitoring module. If any problems are encountered during the execution process, adjustments can be made quickly to ensure the smooth progress of the maintenance work. Finally, the user interaction module provides an intuitive and easy-to-understand graphical interface, allowing users to view road conditions, maintenance decisions and resource scheduling information, and allows users to adjust maintenance strategies by inputting specific parameters through the interface, realizing the flexibility and efficiency of maintenance operations. The operating principle of the entire system revolves around real-time data collection, intelligent analysis, and scientific decision-making, aiming to improve the efficiency and effectiveness of road maintenance work through intelligent means.

[0101] In summary, the maintenance decision-making method for semi-rigid asphalt pavement on highways collects real-time data by deploying advanced sensing equipment and ensures the quality of data through data preprocessing steps. Subsequently, machine learning techniques such as random forest algorithms are used to conduct in-depth analysis of the processed data to predict the maintenance needs of the road. This process enables the system to capture changes in road conditions in a timely manner, so that it can respond at the early stage of problems, avoiding increased costs and safety hazards caused by delayed maintenance.

[0102] The maintenance decision method of semi-rigid base asphalt pavement of highway is based on the analysis results. The decision generation module will formulate the optimal maintenance strategy according to the results of the prediction model and the available resources. At the same time, the system also has the ability to quickly adjust the maintenance plan when an emergency is predicted. The resource scheduling module is responsible for scheduling maintenance resources according to the decision results to ensure the reasonable allocation of resources. The graphical interface provided by the user interaction module allows users to intuitively understand the road conditions, maintenance decisions and resource scheduling, and can adjust the strategy through the interface. Such a design makes maintenance work more flexible and efficient, helps to maximize benefits and minimize costs, and solves the following problems of existing road maintenance technology: first, the data collection means are limited, and it is difficult to obtain comprehensive and continuous road condition information; second, the data processing capacity is insufficient, and it is impossible to effectively extract useful information from massive data. These two problems cause maintenance work to often lag behind actual needs and cannot respond to changes in road conditions in a timely manner, thereby affecting maintenance quality and efficiency. The maintenance method of semi-rigid base asphalt pavement of highway using big data analysis and intelligent decision-making algorithm provided in this embodiment realizes the problems of real-time monitoring, intelligent analysis and scientific decision-making of road conditions.

[0103] An embodiment of a highway semi-rigid base asphalt pavement maintenance decision system:

[0104] An embodiment of the present invention provides a highway semi-rigid base asphalt pavement maintenance decision system, which may include a data acquisition module, a data processing module and a resource scheduling module.

[0105] Among them, the data acquisition module is used to collect different types of monitoring data of the target highway;

[0106] A data processing module is used to perform dimensionality reduction processing on the monitoring data, and predict the maintenance demand of the target expressway based on the data after dimensionality reduction and the prediction model to obtain a prediction result of the maintenance demand of the target expressway;

[0107] The resource scheduling module is used to formulate and execute the optimal maintenance strategy based on the current available resources and the forecast results of the maintenance demand of the target highway.

[0108] It should be understood that the various modules of a highway semi-rigid base asphalt pavement maintenance decision system can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art can understand that the above-mentioned method and system can be implemented using computer executable instructions and / or included in a processor control code, for example, such codes are provided on a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The system and its modules of this specification can not only be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software (for example, firmware).

[0109] For more details about the above modules, please refer to other places in this manual and will not be repeated here.

[0110] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A maintenance decision method for semi-rigid base asphalt pavement of highway, characterized in that: The method comprises the following steps: Collect different types of monitoring data of the target highway; Performing dimensionality reduction processing on the monitoring data, and predicting the maintenance demand of the target expressway based on the dimensionality reduced data and the prediction model to obtain a prediction result of the maintenance demand of the target expressway; The optimal maintenance strategy is formulated and implemented based on the current available resources and the forecast results of the maintenance demand of the target highway.

2. A highway semi-rigid base asphalt pavement maintenance decision method according to claim 1, characterized in that: After collecting different types of monitoring data of the target highway, it also includes: preprocessing the monitoring data, and the preprocessing includes data cleaning, missing value filling, outlier detection and correction, data normalization and data standardization.

3. The maintenance decision method for semi-rigid base asphalt pavement of a highway according to claim 1 is characterized in that: The target highway maintenance demand is predicted based on the dimension-reduced data and the prediction model to obtain the prediction result of the target highway maintenance demand, including: The prediction model is trained using sample monitoring data of the target highway to obtain a trained prediction model; wherein, during the training process, samples and features are randomly selected to construct multiple decision trees; for the regression task, the final output is the average value of the prediction values ​​of all decision trees; the prediction model is constructed using a random forest algorithm; The dimensionally reduced data is input into a trained prediction model to obtain a prediction result of the maintenance demand of the target expressway.

4. A highway semi-rigid base asphalt pavement maintenance decision method according to claim 1, characterized in that: The optimal maintenance strategy is formulated and executed based on the forecast results of the currently available resources and the maintenance demand of the target highway, including: Determine the location, type and time of maintenance for the target expressway based on the forecast results of the maintenance demand of the target expressway; The optimal maintenance strategy is formulated and executed based on the location, type, time and currently available resources of the target highway to be maintained, wherein the currently available resources include funds, manpower and equipment.

5. The maintenance decision method for semi-rigid base asphalt pavement of a highway according to claim 1 is characterized in that: After executing the optimal maintenance strategy, it also includes: if problems occur during the execution process, the maintenance strategy is adjusted.

6. A highway semi-rigid base asphalt pavement maintenance decision method according to claim 1, characterized in that: The PCA algorithm is used to perform dimensionality reduction processing on the monitoring data.

7. A highway semi-rigid base asphalt pavement maintenance decision method according to claim 1, characterized in that: The different types of monitoring data collected on the target highway include: Sensors are installed inside the target highway structure, and the sensors are used to collect different types of monitoring data of the highway; wherein the sensors include temperature sensors, humidity sensors, strain gauges, earth pressure gauges, multi-layer settlement displacement gauges, crack detection sensors and pothole detection sensors, and the interior of the structure includes the road surface, the inside of the surface layer, the inside of the water-stable base layer and the top of the roadbed.

8. A highway semi-rigid base asphalt pavement maintenance decision system, characterized in that: The system includes: Data collection module, used to collect different types of monitoring data of the target highway; A data processing module is used to perform dimensionality reduction processing on the monitoring data, and predict the maintenance demand of the target expressway based on the dimensionality reduced data and the prediction model to obtain a prediction result of the maintenance demand of the target expressway; The resource scheduling module is used to formulate and execute the optimal maintenance strategy based on the current available resources and the forecast results of the maintenance demand of the target highway.

Images