A prediction method for highway corrugated beam guardrail heightening based on machine learning

Abstract

The present invention belongs to the technical field of corrugated beam guardrails. In order to solve the problem of existing reliance on traditional field measurements and manual selection methods to select guardrail heightening schemes, a method for predicting the heightening of highway corrugated beam guardrails based on machine learning is provided. The method includes obtaining a data set, which is the influencing factors affecting guardrail heightening of a pilot section and the actual value of guardrail heightening of the pilot section, establishing a prediction model, inputting the influencing factors affecting guardrail heightening of the pilot section, and outputting the actual value of guardrail heightening of the pilot section, obtaining the influencing factors of the section to be constructed, inputting them into the prediction model, and optimizing the prediction model. The prediction model outputs the predicted value of guardrail heightening of the section to be constructed, establishing a random forest model, and the random forest model selects the optimal guardrail heightening scheme based on the output of the prediction model and the anti-collision level of the guardrail. This method reduces the cost of field investigations, realizes automated scheme selection, and improves work efficiency.

Description

Technical Field

[0001] The present invention belongs to the technical field of corrugated beam guardrails, and in particular relates to a method for predicting the heightening of highway corrugated beam guardrails based on machine learning. Background Art

[0002] With the continuous increase in traffic volume and the length of highway service time, the asphalt pavement of the highway will suffer certain damage due to reasons such as roadbed settlement. In engineering projects, the road surface is generally raised by adding an asphalt overlay layer, which results in the height of the center of the corrugated beam guardrail not meeting the current specifications, thereby causing buses or trucks with a higher center of gravity to overturn or deviate from their driving trajectory, which has a serious impact on road traffic safety.

[0003] The general practice in existing projects is to dismantle and rebuild the original guardrail, but the guardrail facilities are abandoned after dismantling, and the construction and renovation work is large, which will cause huge waste of resources. In addition, the column pull-out construction will damage the original roadbed. In order to avoid the above problems, the sleeve raising method, eccentric block raising method and accessory component raising method are currently adopted for raising the guardrail. In the sleeve raising method, there are problems such as the applicable guardrail section is single, it is necessary to drill holes on the original columns, the construction is cumbersome, and it cannot meet the heightening requirements of guardrails of different levels and heights; in the eccentric block raising method, there is a problem that the heightening requirements of the guardrail are limited by the size of the block; in the accessory component raising method, there is a problem of insufficient flexibility and unsuitability for the height increase of corrugated beams of different levels. In addition, the adjustment and modification of the guardrail height mainly rely on experience or simple calculation methods, and lack of systematic prediction and optimization solutions. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides a highway corrugated beam guardrail heightening prediction method based on machine learning that can meet diverse heightening needs and achieve optimization of protection effectiveness and cost-effectiveness.

[0005] The present invention provides a method for predicting the heightening of a highway corrugated beam guardrail based on machine learning, comprising the following steps:

[0006] S1. Obtain the influencing factors affecting guardrail heightening and the actual value of guardrail heightening on the pilot road section, and use the influencing factors affecting guardrail heightening and the actual value of guardrail heightening on the pilot road section as a data set;

[0007] S2. Establish a prediction model, where the input of the prediction model is the factors affecting the height increase of the guardrail on the pilot road section, and the output of the prediction model is the actual value of the height increase of the guardrail on the pilot road section;

[0008] S3. Obtaining an influencing factor of the road section to be constructed, inputting the influencing factor of the road section to be constructed into a prediction model, and optimizing the prediction model, so that the prediction model outputs a predicted value of the guardrail height increase of the road section to be constructed;

[0009] S4. Establish a random forest model based on the predicted value h of the guardrail height increase in the road section to be constructed. i And the anti-collision level of the guardrail, choose the optimal guardrail heightening plan.

[0010] Optionally, the guardrail heightening solution includes improving the guardrail bracket, improving the guardrail anti-blocking block and adding sleeves;

[0011] Improving the guardrail bracket includes increasing the height of the original bracket and increasing the distance between the first threaded hole connecting the original bracket to the corrugated beam of the guardrail and the second threaded hole connecting the original bracket to the column of the guardrail;

[0012] Improve the barrier block of the guardrail by adjusting the distance from the third threaded hole connecting the original barrier block to the corrugated beam of the guardrail to the bottom of the barrier block, and the distance from the fourth threaded hole connecting the original barrier block to the column of the guardrail to the bottom of the barrier block;

[0013] The additional sleeve is embedded in the guardrail column, and a fifth threaded hole and a sixth threaded hole are opened in the height direction of the sleeve. The fifth threaded hole is connected to the guardrail column by a bolt, and the sixth threaded hole is connected to the guardrail bracket or the guardrail anti-blocking block by a bolt.

[0014] Optional, for Class A double corrugated beam guardrails equipped with anti-blocking blocks, when h i When the range is 0-50mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when h i When the distance is between 50 and 100 mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, and reduce the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when h i When the distance is between 100 and 130 mm, reduce the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, increase the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block and add a sleeve; when h i When the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block is 130-180 mm, reduce the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block, and add a sleeve; when h i When the diameter is between 180 and 250 mm, add sleeves to the upright posts of the guardrail.

[0015] Optional, for Class A three-corrugated beam guardrails equipped with anti-blocking blocks, when h i When the range is 0-50mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when h iWhen the distance is between 50 and 100 mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, and reduce the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when h i When the distance is between 100 and 150 mm, reduce the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, increase the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block, and add a sleeve; when h i When the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block is 150-250mm, reduce or increase the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block, and add a sleeve; when h i When the distance is between 250 and 300 mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, reduce the distance from the fourth threaded hole to the bottom of the anti-blocking block, and add a sleeve; when h i =200mm, add a sleeve to the guardrail post.

[0016] Optionally, for Class B or Class C double corrugated beam guardrails equipped with brackets, when h i When the height is between 0 and 65 mm, improve the bracket of the guardrail; when h i When the height is between 65 and 130 mm, improve the bracket of the guardrail and add a sleeve; when h i When the diameter is between 130 and 250 mm, add sleeves to the upright posts of the guardrail.

[0017] Optionally, the influencing factors include road characteristics, environmental impacts, and guardrail characteristics.

[0018] Optionally, in step S1, operations of removing outliers and missing values, stabilizing the data, and normalizing the data are performed on the data set.

[0019] Optionally, in step S3, optimizing the prediction model includes: dividing the data set into a training set and a validation set, with the training set accounting for 70% of the data set and the validation set accounting for 30% of the data set, applying a neural network algorithm and using the training set to train the prediction model, and using the validation set to validate the prediction model.

[0020] The technical solution provided by the embodiment of the present invention has the following beneficial effects compared with the prior art:

[0021] An embodiment of the present invention provides a method for predicting the heightening of highway corrugated beam guardrails based on machine learning. A prediction model is established by obtaining the influencing factors of a pilot section and the actual value of the guardrail heightening. The influencing factors of the section to be constructed are then obtained and input into the prediction model. The prediction model outputs the predicted value of the guardrail heightening of the section to be constructed, and a random forest model is used to select the guardrail heightening scheme with the highest protective effectiveness and cost-effectiveness. This method reduces the expenses of field investigations, realizes automated scheme selection, reduces construction costs, and improves work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0024] Figure 1 This is a flowchart of the prediction of highway corrugated beam guardrail height increase based on machine learning according to an embodiment of the present invention;

[0025] Figure 2 This is a schematic structural diagram of an existing bracket according to an embodiment of the present invention;

[0026] Figure 3 This is a schematic structural diagram of an improved bracket according to an embodiment of the present invention;

[0027] Figure 4 This is a schematic structural diagram of the improved connection between the bracket and the column according to an embodiment of the present invention;

[0028] Figure 5 This is a schematic structural diagram of the original anti-blocking block according to an embodiment of the present invention;

[0029] Figure 6 This is a schematic structural diagram of an improved anti-blocking block according to an embodiment of the present invention;

[0030] Figure 7 This is a schematic structural diagram of a sleeve according to an embodiment of the present invention;

[0031] Figure 8 This is a schematic structural diagram of an A-class double corrugated beam guardrail with an anti-blocking block according to an embodiment of the present invention, with only an additional sleeve;

[0032] Figure 9 This is a structural schematic diagram of the A-level three-corrugated beam guardrail with anti-obstruction blocks according to an embodiment of the present invention, which uses improved anti-obstruction blocks and additional sleeves.

[0033] Among them, 1. column; 2. bracket; 201. first threaded hole; 202. second threaded hole; 3. anti-obstruction block; 301. third threaded hole; 302. fourth threaded hole; 4. sleeve; 401. fifth threaded hole; 402. sixth threaded hole. DETAILED DESCRIPTION

[0034] In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the scheme of the present invention will be further described below. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features therein can be combined with each other.

[0035] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present invention, rather than all the embodiments.

[0036] Reference Figures 1 to 9 As shown, this embodiment provides a method for predicting the height of a highway corrugated beam guardrail based on machine learning, comprising the following steps:

[0037] S1. Obtain the influencing factors and actual values ​​of guardrail height increase on the pilot road section, and use the influencing factors and actual values ​​of guardrail height increase on the pilot road section as a data set. Specifically, the influencing factors include road characteristics, environmental impacts, and guardrail characteristics. Road characteristics include highway category, longitudinal slope, and curve radius. Environmental impacts include traffic volume, proportion of heavy vehicles, precipitation, and geological conditions. Guardrail characteristics include corrugated beam style, expected vehicle type, and driving speed. To ensure fairness in data comparison, the mapminmax function in MATLAB is used to perform scale normalization on the obtained data to ensure comparability among the variables. That is, the collected data set is subjected to operations such as removing outliers and missing values, stabilization, and data normalization.

[0038] S2. Establish a prediction model, where the input of the prediction model is the factors affecting the height increase of the guardrail on the pilot road section, and the output of the prediction model is the actual value of the height increase of the guardrail on the pilot road section;

[0039] S3. Obtain the influencing factors of the road section to be constructed, input the influencing factors of the road section to be constructed into the prediction model, and optimize the prediction model. The prediction model outputs the predicted value of the guardrail height increase of the road section to be constructed. Specifically, the data set is divided into a training set and a validation set, with the training set accounting for 70% of the data set and the validation set accounting for 30% of the data set. The prediction model is trained using the training set by applying a neural network algorithm, and the prediction model is validated using the validation set. By calculating the root mean square error, mean absolute percentage error, coefficient of determination, and mean absolute error, the prediction accuracy, stability, and generalization performance of the prediction model are comprehensively evaluated, thereby optimizing an efficient prediction model.

[0040] S4. Establish a random forest model based on the predicted value h of the guardrail height increase in the road section to be constructed. i And the anti-collision level of the guardrail, choose the optimal guardrail heightening plan.

[0041] Further, refer to Figure 1 As shown, the guardrail heightening solution includes an improved guardrail bracket, an improved guardrail anti-blocking block and an additional sleeve 4;

[0042] Improvement of the guardrail bracket includes increasing the height of the original bracket, increasing the distance between the first threaded hole connecting the original bracket to the corrugated beam of the guardrail and the second threaded hole connecting the original bracket to the column 1 of the guardrail, for details, refer to Figure 2 and Figure 3 As shown, the improved bracket 2 has a higher height than the original bracket, and the vertical distance between the first threaded hole 201 and the second threaded hole 202 is increased. The improved bracket 2 is suitable for Class B or Class C corrugated beam guardrails. The length × width × height of the original bracket is 300 × 4.5 × 70 mm. The height of the improved bracket 2 is increased from 70 mm to 155 mm. The position of the second threaded hole 202 remains unchanged. The vertical distance between the first threaded hole 201 and the second threaded hole 202 is increased. When ensuring that the minimum distance from the first threaded hole 201 to the edge of the bracket 2 is 35 mm, the maximum height of the guardrail is 65 mm.

[0043] Improve the barrier block of the guardrail to adjust the distance from the third threaded hole connecting the original barrier block to the corrugated beam of the guardrail to the bottom of the barrier block, and the distance from the fourth threaded hole connecting the original barrier block to the column 1 of the guardrail to the bottom of the barrier block. For details, refer to Figure 5 and Figure 6 As shown, the improved anti-blocking block 3 is suitable for Class A corrugated beam guardrail, wherein, in the Class A double corrugated beam guardrail, the guardrail is raised by changing the distance from the third threaded hole 301 or the fourth threaded hole 302 to the bottom of the anti-blocking block. When the minimum distance from the third threaded hole 301 or the fourth threaded hole 302 to the edge of the anti-blocking block 3 in the vertical direction is ensured to be 50 mm, the maximum height of the guardrail is raised to 50 mm. In the Class A triple corrugated beam guardrail, the guardrail is raised by changing the distance from the third threaded hole 301 or the fourth threaded hole 302 to the bottom of the anti-blocking block 3. It is necessary to ensure that the distance between the upper and lower third threaded holes 301 on the same side or the upper and lower fourth threaded holes 302 on the same side is 200 mm, and to ensure that the minimum distance from the third threaded hole 301 or the fourth threaded hole 302 to the vertical edge of the anti-blocking block 3 is 50 mm, and to achieve a maximum height increase of 50 mm for the guardrail by changing the distance from the third threaded hole 301 or the fourth threaded hole 302 to the bottom of the anti-blocking block 3 separately, and to achieve a maximum height increase of 100 mm for the guardrail by changing the distance from the third threaded hole 301 and the fourth threaded hole 302 to the bottom of the anti-blocking block 3 together;

[0044] The additional sleeve 4 is embedded in the column 1 of the guardrail, and a fifth threaded hole 401 and a sixth threaded hole 402 are opened in the height direction of the sleeve 4. The fifth threaded hole 401 is connected to the column 1 of the guardrail by a bolt, and the sixth threaded hole 402 is connected to the bracket of the guardrail or the anti-blocking block of the guardrail by a bolt. For details, refer to Figure 7 and Figure 8 As shown, the sleeve 4 and the column 1 of the guardrail are made of the same material. The section where the sleeve 4 is connected to the column 1 is processed by cold pressing technology, and the diameter of the section where the sleeve 4 is connected to the column 1 is 2 mm smaller than the diameter of the column 1. The fifth threaded hole 401 is the same size as the original bolt hole on the column 1, and the connection between the sleeve 4 and the column 1 can be achieved without drilling the column 1 again.

[0045] This embodiment achieves the raising of the guardrail by improving the bracket 2, improving the anti-blocking block 3 or adding the sleeve 4. The guardrail can also be raised by combining two of the above methods. Among them, the bracket 2, the anti-blocking block 3 and the sleeve 4 are all prefabricated in a standardized factory. The sleeve 4 needs to be cold-pressed first and then galvanized for anti-corrosion treatment to ensure the service life of the bracket 2, the anti-blocking block 3 and the sleeve 4.

[0046] Furthermore, for the A-level double corrugated beam guardrail provided with the anti-blocking block 3, when h i When the distance from the third threaded hole 301 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is increased from 0 to 50 mm, specifically, the height of the anti-blocking block 3 remains unchanged, the position of the fourth threaded hole 302 remains unchanged, and the distance from the third threaded hole 301 to the bottom of the anti-blocking block 3 increases by 0 to 50 mm; when h i When the height is 50-100 mm, the distance from the third threaded hole 301 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is increased, and the distance from the fourth threaded hole 302 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is reduced. Specifically, the height of the anti-blocking block 3 remains unchanged, the distance from the third threaded hole 301 to the bottom of the anti-blocking block 3 is increased by 0-50 mm, and the distance from the fourth threaded hole 302 to the bottom of the anti-blocking block 3 is reduced by 50 mm; when h i When the distance is between 100 and 130 mm, the distance between the third threaded hole 301 of the anti-blocking block 3 and the bottom of the anti-blocking block 3 is reduced, the distance between the fourth threaded hole 302 of the anti-blocking block 3 and the bottom of the anti-blocking block 3 is increased, and the sleeve 4 is added. Specifically, the sleeve 4 is embedded in the column 1, and the height of the sleeve 4 exceeding the column 1 is 180 mm. The distance between the third threaded hole 301 of the anti-blocking block and the bottom of the anti-blocking block 3 is reduced by 0 to 50 mm, and the distance between the fourth threaded hole 302 and the bottom of the anti-blocking block 3 is increased by 50 mm. When h iWhen the distance between the third threaded hole 301 of the anti-blocking block 3 and the bottom of the anti-blocking block 3 is 130-180 mm, the distance between the third threaded hole 301 of the anti-blocking block 3 and the bottom of the anti-blocking block 3 is reduced by 0-50 mm, and the sleeve 4 is attached. Specifically, the sleeve 4 is embedded in the column 1, and the height of the sleeve 4 exceeding the column 1 is 180 mm. The distance between the third threaded hole 301 of the anti-blocking block and the bottom of the anti-blocking block 3 is reduced by 0-50 mm. When h i When the height is 180-250 mm, a sleeve 4 is added to the column 1 of the guardrail. Specifically, the original anti-blocking block is still used, and the sleeve 4 is embedded in the column 1. The total height of the sleeve 4 is 420 mm, and the height of the part of the sleeve 4 embedded in the column 1 is 220 mm. The distance between the fifth threaded hole 401 and the bottom edge of the sleeve 4 is 70 mm, and the distance between the sixth threaded hole 402 and the top edge of the sleeve 4 is 150 mm.

[0047] For the A-level three-corrugated beam guardrail equipped with anti-blocking block 3, when h i When the distance between the third threaded hole 301 of the anti-blocking block 3 and the bottom of the anti-blocking block 3 is 0-50 mm, the distance between the third threaded hole 301 and the bottom of the anti-blocking block 3 is increased by 0-50 mm, and the distance between the two third threaded holes 301 is ensured to be 200 mm; when h i When the distance is 50-100 mm, the distance from the third threaded hole 301 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is increased, and the distance from the fourth threaded hole 302 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is reduced. Specifically, the distance from the third threaded hole 301 to the bottom of the anti-blocking block 3 is increased by 0-50 mm, and the distance from the fourth threaded hole 302 to the bottom of the anti-blocking block 3 is reduced by 0-50 mm, so that the distance between the third threaded hole 301 and the fourth threaded hole 302 is increased by 50 mm to 100 mm; when h i When the distance is 100-150 mm, the distance from the third threaded hole 301 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is reduced, the distance from the fourth threaded hole 302 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is increased, and the sleeve 4 is added. Specifically, the sleeve 4 is embedded in the column 1, and the height of the sleeve 4 exceeding the column 1 is 200 mm. The distance from the third threaded hole 301 to the bottom of the anti-blocking block 3 is reduced by 0-50 mm, and the distance from the fourth threaded hole 302 to the bottom of the anti-blocking block 3 is increased by 0-50 mm. When h i When the distance between the third threaded hole 301 and the bottom of the anti-blocking block 3 is 150-250 mm, the distance between the third threaded hole 301 and the bottom of the anti-blocking block 3 is increased, and the sleeve 4 is added. Specifically, the sleeve 4 is embedded in the column 1, and the height of the sleeve 4 exceeding the column 1 is 200 mm. The position of the fourth threaded hole 302 remains unchanged, and the distance between the third threaded hole 301 and the bottom of the anti-blocking block 3 is reduced or increased by 0-50 mm. When h iWhen the distance is 250-300 mm, the distance from the third threaded hole 301 of the anti-blocking block 3 to the bottom of the anti-blocking block 3 is increased, the distance from the fourth threaded hole 302 to the bottom of the anti-blocking block 3 is reduced, and the sleeve 4 is added. Specifically, the sleeve 4 is embedded in the column 1, and the height of the sleeve 4 exceeding the column 1 is 200 mm. The distance from the third threaded hole 301 to the bottom of the anti-blocking block 3 is increased by 0-50 mm, and the distance from the fourth threaded hole 302 to the bottom of the anti-blocking block 3 is reduced by 0-50 mm. When h i =200mm, a sleeve 4 is added to the column 1 of the guardrail. Specifically, the sleeve 4 is embedded in the column 1, and the height of the sleeve 4 exceeding the column 1 is 200mm.

[0048] For Class B or Class C double corrugated beam guardrails equipped with brackets, when h i When the height of the bracket 2 of the guardrail is 0-65 mm, the height of the bracket 2 is improved. Specifically, the height of the bracket 2 is increased to 155 mm, the position of the second threaded hole 202 of the bracket 2 remains unchanged, and the vertical distance between the first threaded hole 201 and the second threaded hole 202 is increased to 0-65 mm. When the minimum distance from the first threaded hole 201 to the edge of the bracket 2 is ensured to be 35 mm, the maximum height of the guardrail is 65 mm. When h i When the height is 65-130 mm, the bracket 2 of the guardrail is improved and the sleeve 4 is added. Specifically, the height of the sleeve 4 is increased to 130 mm. The position of the second threaded hole 202 of the bracket 2 remains unchanged, and the vertical distance between the first threaded hole 201 and the second threaded hole 202 is reduced, so that the position of the first threaded hole 201 is lowered by 0-65 mm. When h i When the height is between 130 and 250 mm, a sleeve 4 is added to the column 1 of the guardrail. The original bracket is still used. The embedded height between the sleeve 4 and the column 1 is 170 mm, and the distance from the fifth threaded hole 401 to the top of the column 1 is 150 mm.

[0049] In a detailed case study of a Class III highway in a mountainous area, the highway has a total length of 20 km and is equipped with Class B double corrugated beam guardrails, covering a length of 12.3 km. During the initial design, the guardrail center-to-ground height was set at 600mm. After 14 years of operation, through a combination of on-site inspections and data collection, key information for the mileage range from K010+963 to K014+263 was accurately recorded. This includes an average daily traffic volume of 8,409 vehicles, of which 40% are heavy vehicles weighing over 12 tons. Furthermore, the longitudinal slope distribution of the line is flat for 3.8km, 1.1km of 1% slope, 2.3km of 2% slope, 0.7km of 3% slope, 3.6km of 4% slope, and 0.8km of 5% slope. The highway's geological composition includes a 0.5km section of loess fill, a 1km section of rock fill, and a 10.8km section of excavated earth. Corrugated beam installation adheres to a spacing of one column every two meters. This led to the construction of a parameter matrix containing 6,150 rows and seven columns. The first column records the mileage number, and a corresponding 6,150-row, one-column matrix of corrugated beam center height adjustment data was obtained.

[0050] During the data preprocessing stage, the collected data matrices of 6150×7 and 6150×1 dimensions were strictly processed for outlier elimination, noise reduction and standardization. The data were systematically divided into a 70% training set and a 30% validation set. Multiple rounds of optimization training were carried out using the MATLAB built-in BP neural network algorithm to ensure that the accuracy of the prediction model could withstand rigorous verification.

[0051] Subsequently, data collection was expanded to include a 7.7km section of the highway not covered, covering similar traffic and geographical parameters. By analyzing this data, the appropriate height of the guardrail was predicted.

[0052] Based on this predicted value, a matrix of 3850 rows and 1 column was constructed as input. The random forest algorithm was used to optimize and determine the most reasonable guardrail heightening plan with minimizing the comprehensive engineering cost as the objective function, providing a scientific decision-making basis for the upgrade of highway safety protection measures.

[0053] The following are specific embodiments:

[0054] Example 1:

[0055] For a Class B double-corrugated beam guardrail on a certain road section, column 1 has a size of φ114×4.5mm and a wall thickness of 4.5mm. The BP neural network method predicts that the center of the corrugated beam is 55cm from the ground and needs to be raised by 5cm. The random forest method determines that only bracket 2 should be replaced to increase the height.

[0056] The original bracket height is 70mm, and the improved bracket 2 height is increased to 155mm. The second threaded hole 202 is 55mm away from the lower edge of the bracket 2, and the first threaded hole 201 is 35mm away from the upper edge of the bracket 2. The dimensions of the first threaded hole 201 and the second threaded hole 202 are 22×50mm and 22×28mm respectively, which are consistent with the bolt hole dimensions of the original bracket. Figure 3 The modified bracket 2 is shown as a standby.

[0057] Reference Figure 4 As shown, the bracket 2 is transported to the site and the emergency lane is used as a construction access road. First, the corrugated beam is lifted with a forklift, and then the bolt connection between the corrugated beam and the original bracket is removed, and the bolt connection between the column 1 and the original bracket is removed. The bolts between the improved bracket 2 and the corrugated beam and column 1 can be installed.

[0058] Example 2:

[0059] In a modified Class A double corrugated beam guardrail (2006 specification), the dimensions of column 1 are φ140×4.5mm, the wall thickness is 4.5mm, the height of the original barrier block is 200mm, and the bolt holes on the two opposite sides of the original barrier block are arranged oppositely, with two bolt holes on each side. The distance between the two bolt holes and the top and bottom of the original barrier block are both 100mm. Based on the BP neural network method, it is predicted that the height of the corrugated beam guardrail from the ground is 400mm and needs to be increased by 200mm. After classification using the random forest method, the height increase scheme of only adding sleeve 4 is selected.

[0060] The sleeve 4 is made of the same steel as the original column 1, with a total height of 420 mm. The section where the sleeve 4 is connected to the column 1 is made by cold pressing, and the outer diameter of this part is 129 mm. The embedded height between the column 1 and the sleeve 4 is 220 mm. The dimensions of the fifth threaded hole 401 and the sixth threaded hole 402 on the sleeve 4 are 24×18 mm, wherein the distance between the fifth threaded hole 401 and the bottom edge of the sleeve 4 is 70 mm, and the distance between the sixth threaded hole 402 and the top edge of the sleeve 4 is 150 mm.

[0061] Reference Figure 8 As shown, the sleeve 4 is transported to the site, and the bolt connections between the corrugated beam, the column 1 and the original anti-blocking block are first removed. Secondly, the sleeve 4 is embedded in the column 1, and the fifth threaded hole 401 of the sleeve 4 is aligned with the bolt hole of the column 1, and connected with bolts, and the original anti-blocking block is connected with the corrugated beam and the sixth threaded hole 402 of the attached sleeve 4 by bolts.

[0062] Example 3:

[0063] For a Class A three-corrugated beam guardrail (2017 version of the specification) on a certain road section, the original barrier block has a height of 400mm. Two bolt holes are opened in the vertical direction on the side connected to the column 1, and two bolt holes are opened in the vertical direction on the side connected to the corrugated beam. The two bolt holes on the same side are 100mm away from the top and bottom edges of the original barrier block respectively, and the distance between the two bolt holes on the same side is 100mm. The BP neural network method is used to predict that the height of the center of the corrugated beam from the ground is 450mm, and it needs to be raised by 250mm. The random forest model classification is used to select two heightening schemes: adding a sleeve 4 and replacing the improved barrier block 3. The additional sleeve 4 can increase the guardrail by 200mm. The position of the fourth threaded hole 302 of the barrier block 3 remains unchanged, and the distance from the third threaded hole 301 to the bottom of the barrier block 3 is increased by 50mm, so that the guardrail is raised by 250mm as a whole.

[0064] The sleeve 4 is made of the same steel as the column 1. The total height of the sleeve 4 is 420 mm. The section where the sleeve 4 is connected to the column 1 is made by cold pressing, and the outer diameter of this part is 129 mm. The embedded height between the column 1 and the sleeve 4 is maintained at 220 mm. The dimensions of the fifth threaded hole 401 and the sixth threaded hole 402 on the sleeve 4 are 24×18 mm, wherein the distance between the fifth threaded hole 401 and the bottom edge of the sleeve 4 is 70 mm, and the distance between the sixth threaded hole 402 and the top edge of the sleeve 4 is 150 mm.

[0065] Reference Figure 9 As shown, during the specific installation, first lift the corrugated beam with a forklift, then remove the bolt connection between the corrugated beam, the column 1 and the original anti-blocking block, then embed the sleeve 4 into the column 1, align the fifth threaded hole 401 of the sleeve 4 with the bolt hole of the column 1, and connect them with bolts, and connect the original anti-blocking block with the corrugated beam and the sixth threaded hole 402 of the attached sleeve 4 through bolts.

[0066] 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 entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further restrictions, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0067] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments described herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the height of highway corrugated beam guardrails based on machine learning, characterized in that: The following steps are involved: S1. Obtain the influencing factors affecting guardrail heightening and the actual value of guardrail heightening on the pilot road section, and use the influencing factors affecting guardrail heightening and the actual value of guardrail heightening on the pilot road section as a data set; S2. Establish a prediction model, where the input of the prediction model is the factors affecting the height increase of the guardrail on the pilot road section, and the output of the prediction model is the actual value of the height increase of the guardrail on the pilot road section; S3. Obtaining an influencing factor of the road section to be constructed, inputting the influencing factor of the road section to be constructed into a prediction model, and optimizing the prediction model, so that the prediction model outputs a predicted value of the guardrail height increase of the road section to be constructed; S4. Build a random forest model based on the predicted value of the guardrail height increase in the road section to be constructed. And the anti-collision level of the guardrail, choose the best guardrail heightening solution; The guardrail heightening scheme includes an improved guardrail bracket, an improved guardrail anti-blocking block and an additional sleeve (4); The improved guardrail bracket includes increasing the height of the original bracket and increasing the distance between the first threaded hole connecting the original bracket to the corrugated beam of the guardrail and the second threaded hole connecting the original bracket to the upright post (1) of the guardrail; Improve the barrier block of the guardrail by adjusting the distance from the third threaded hole connecting the original barrier block with the corrugated beam of the guardrail to the bottom of the barrier block, and the distance from the fourth threaded hole connecting the original barrier block with the upright post (1) of the guardrail to the bottom of the barrier block; The additional sleeve (4) is embedded in the guardrail column (1), and a fifth threaded hole (401) and a sixth threaded hole (402) are provided in the height direction of the sleeve (4); the fifth threaded hole (401) is connected to the guardrail column (1) by a bolt, and the sixth threaded hole (402) is connected to the guardrail bracket or the guardrail anti-blocking block by a bolt; For Class B or Class C double corrugated beam guardrails equipped with brackets, When the diameter is 0~65mm, improve the bracket of the guardrail; when When the diameter is 65~130mm, improve the bracket of the guardrail and add a sleeve (4); when When the diameter is 130-250 mm, a sleeve (4) is added to the upright post (1) of the guardrail.

2. The method for predicting the height of highway corrugated beam guardrail based on machine learning according to claim 1 is characterized in that: For Class A double corrugated beam guardrails equipped with anti-blocking blocks, When the range is 0~50mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when When the distance is between 50 and 100 mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, and reduce the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when When the diameter is between 100 and 130 mm, the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block is reduced, the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block is increased, and a sleeve (4) is added; when When the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block is 130-180 mm, the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block is reduced, and a sleeve (4) is added; when When the distance is 180-250 mm, a sleeve (4) is added to the upright post (1) of the guardrail.

3. The method for predicting the height of highway corrugated beam guardrail based on machine learning according to claim 1 is characterized in that: For Class A three-corrugated beam guardrails equipped with anti-blocking blocks, When the range is 0~50mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when When the distance is between 50 and 100 mm, increase the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block, and reduce the distance from the fourth threaded hole of the anti-blocking block to the bottom of the anti-blocking block; when When the distance is between 100 and 150 mm, the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block is reduced, the distance between the fourth threaded hole of the anti-blocking block and the bottom of the anti-blocking block is increased, and a sleeve (4) is added; when When the distance between the third threaded hole of the anti-blocking block and the bottom of the anti-blocking block is 150-250 mm, the distance is reduced or increased, and a sleeve (4) is added; when When the distance is between 250 and 300 mm, the distance from the third threaded hole of the anti-blocking block to the bottom of the anti-blocking block is increased, the distance from the fourth threaded hole to the bottom of the anti-blocking block is reduced, and a sleeve (4) is added; when When the guardrail is installed, a sleeve (4) is attached to the upright post (1) of the guardrail.

4. The method for predicting the height of highway corrugated beam guardrail based on machine learning according to claim 1 is characterized in that: Influencing factors include road characteristics, environmental impacts, and guardrail characteristics.

5. The method for predicting the height of highway corrugated beam guardrail based on machine learning according to claim 1 is characterized in that: In step S1, the data set is subjected to operations of removing outliers and missing values, stabilizing the data, and normalizing the data.

6. The method for predicting the heightening of highway corrugated beam guardrails based on machine learning according to claim 1, characterized in that: In step S3, optimizing the prediction model includes: dividing the data set into a training set and a validation set, with the training set accounting for 70% of the data set and the validation set accounting for 30% of the data set; applying a neural network algorithm and using the training set to train the prediction model; and using the validation set to validate the prediction model.

Images