Road maintenance quality detection system based on intelligent decision

Abstract

The invention relates to the technical field of road data processing, in particular to a road maintenance quality detection system based on intelligent decision, which comprises an image acquisition module, a model calibration module, a maintenance planning module, a monitoring analysis module, a vehicle monitoring module, an evaluation module and a judgment module. Adjusting the preset maintenance duration corresponding to the single region to a corresponding value, adjusting the maintenance thickness of each road section corresponding to the single region to a corresponding value, adjusting the preset maintenance area to a corresponding value, or adjusting the number of pavement image information used for training a built-in model in a model calibration module to a corresponding value; after road maintenance is completed, road conditions are classified according to continuously monitored road image information, subsequent maintenance parameters are adjusted according to a classification result, and the road maintenance efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of road data processing, and particularly to a road maintenance quality detection system based on intelligent decision-making. Background Art

[0002] With the rapid development and increasing perfection of road construction, the road mileage has been continuously increasing, and the road traffic volume has gradually increased in recent years. Along with the overloaded operation of roads, it directly leads to an increasing number of road diseases, which has a great impact on the road use performance. It not only directly affects the overall performance and lifespan of the highway, but also is directly related to the driving safety, comfort, and economy. Therefore, road maintenance is an important link to maintain the road use quality.

[0003] However, the contradiction between traditional road maintenance management means and the current large-scale maintenance requirements has become increasingly acute. For road inspection and repair work, it mainly still relies on inspectors to describe the disease locations in words, lacking a favorable road disease repair inspection model, and consuming a large amount of time and human resources. Due to the wide range and complexity of road repair projects, manual inspection has strong subjectivity, and different inspectors may produce inconsistent results, which will bring certain difficulties to the road repair decision-making work.

[0004] Chinese Patent Publication No.: CN117522175B, discloses a road maintenance decision-making method and system. The method includes determining the PCI of the target road; determining the preliminary decision-making level of the target road maintenance according to the PCI; controlling the road detection equipment to obtain the detection data of the target road; controlling the workstation to determine the road disease data according to the detection data; the workstation is a road disease identification and processing platform based on the road disease model; determining the in-depth decision-making level of the target road maintenance according to the road disease data; determining the target road maintenance decision-making mode according to the preliminary decision-making level of the target road maintenance and the in-depth decision-making level of the target road maintenance; It can be seen that the above technical solution has the following problems: It does not consider classifying the road conditions based on the continuously monitored road image information after completing the road maintenance, and does not consider adjusting the subsequent maintenance parameters according to the classification results, which affects the road maintenance efficiency. Summary of the Invention

[0005] Therefore, the present invention provides a road maintenance quality detection system based on intelligent decision-making to overcome the problems in the prior art that it does not consider classifying the road conditions based on the continuously monitored road image information after completing the road maintenance, does not consider adjusting the subsequent maintenance parameters according to the classification results, and affects the road maintenance efficiency.

[0006] To achieve the above object, the present invention provides a road maintenance quality detection system based on intelligent decision-making, including: An image acquisition module, which includes a plurality of image acquirers respectively arranged in each section of each road surface area for acquiring road surface image information; A model calibration module, which is connected to the image acquisition module and is used for periodically calibrating the damaged areas in the road surface image information, and determining the areas to be maintained as the damaged areas with the damaged area larger than the preset maintenance area; A maintenance plan module, which is connected to the model calibration module and is used for determining maintenance parameters based on the damaged areas calibrated by the model calibration module, including maintenance area, maintenance thickness and maintenance duration; A monitoring and analysis module, which is respectively connected to the image acquisition module, the model calibration module and the maintenance plan module, and is used for determining the flatness based on the road surface image information of the corresponding section re-acquired by the image acquisition module after a preset operation duration under the condition that the maintenance of each area to be maintained is completed; A vehicle monitoring module, which is used for counting the passing numbers of heavy-duty vehicles on each section in each cycle; An evaluation module, which is respectively connected to the monitoring and analysis module and the vehicle monitoring module, and is used for marking the maintenance labels of the corresponding sections based on the flatness of each section; A determination module, which is respectively connected to the evaluation module, the maintenance plan module and the model calibration module, and is used for determining whether the maintenance of each section in a single road surface area is qualified based on the number of each type of maintenance label of each section in a single road surface area, and when it is determined that the maintenance of each section in a single road surface area is abnormal, determining the maintenance duration, maintenance thickness or maintenance area of a single road surface area, or determining the number of road surface image information for training the built-in model in the model calibration module.

[0007] Further, the monitoring and analysis module is used for determining the flatness based on the road surface image information, for obtaining the gray difference between adjacent pixel points in the road surface image information, calculating the standard deviation of each gray difference, and obtaining the flatness for a single road surface image information; The evaluation module is used for marking the maintenance labels of the corresponding sections based on the flatness, including: If the flatness is less than or equal to the first preset flatness, the maintenance label of a single section is determined as a stable section; If the flatness is less than or equal to the second preset flatness and greater than the first preset flatness, the maintenance label of a single section is marked in combination with the heavy-duty vehicle flow parameter of a single section; If the flatness is greater than the second preset flatness, the maintenance label of a single section is marked based on the average rainfall in the road surface area corresponding to a single section within the preset operation duration.

[0008] Further, the evaluation module is used for marking the maintenance labels of a single section based on the heavy-duty vehicle flow parameter of a single section, including: Obtain the passing quantity of overloaded vehicles on a single road section within several cycles, calculate the variance of the passing quantity in each cycle, and obtain the overloaded vehicle flow parameter; If the overloaded vehicle flow parameter is less than or equal to the first preset flow parameter, determine the maintenance label of the single road section as a stable road section; If the overloaded vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, determine the maintenance label of the single road section as a fluctuating road section; If the overloaded vehicle flow parameter is greater than the second preset flow parameter, mark the maintenance label of the single road section based on the average rainfall in the road surface area corresponding to the single road section within the preset operation duration.

[0009] Further, the evaluation module is used to mark the maintenance label of the single road section based on the average rainfall in the road surface area corresponding to the single road section within the preset operation duration, including: If the average rainfall is less than or equal to the first preset rainfall, mark the maintenance label of the single road section based on the flatness difference parameter between the single road section and the adjacent road section; If the average rainfall is less than or equal to the second preset rainfall and greater than the first preset rainfall, determine the maintenance label of the single road section as a sensitive road section; If the average rainfall is greater than the second preset rainfall, determine the maintenance label of the single road section as a weak road section.

[0010] Further, the evaluation module is used to mark the maintenance label of the single road section based on the flatness difference parameter between the single road section and the adjacent road section, including: Calculate the difference between the flatness of the single road section and the flatness of the single road section connected to the single road section obtained, and obtain the flatness difference; Calculate the average value of the flatness differences corresponding to the two road sections connected to the single road section, and obtain the flatness difference parameter; If the flatness difference parameter is less than or equal to the preset difference parameter, determine the maintenance label of the single road section as a sensitive road section; If the flatness difference parameter is greater than the preset difference parameter, determine the maintenance label of the single road section as a vulnerable road section.

[0011] Further, the determination module is used to determine whether the maintenance of each road section in a single road surface area is qualified based on the quantity of each type of maintenance label of each road section in the statistically single road surface area, including: Statistically count the quantity of each type of maintenance label in the single road surface area, and determine the maintenance label with the largest quantity as the problem label; If the problem label is a stable road section, determine that the maintenance of each road section in the single road surface area is qualified, and control each module to continue running with the current operating parameters; If the problem label is a fluctuating section, it is determined that the maintenance of each section within a single road surface area is abnormal, and based on the average value of the heavy vehicle flow parameters of each fluctuating section within the single road surface area, the preset maintenance duration corresponding to the single road surface area is adjusted to the corresponding value; If the problem label is a sensitive section, it is determined that the maintenance of each section within a single road surface area is abnormal, and based on the average value of the flatness of each sensitive section within the single road surface area, the repair thickness of each section corresponding to the single road surface area is adjusted to the corresponding value; If the problem label is a vulnerable section, it is determined that the maintenance of each section within a single road surface area is abnormal, and based on the average value of the flatness difference parameters of each vulnerable section within the single road surface area, the preset maintenance area is adjusted to the corresponding value; If the problem label is a weak section, it is determined that the maintenance of each section within a single road surface area is abnormal, and based on the average rainfall within the single road surface area, the number of road surface image information used to train the built-in model in the model calibration module is adjusted to the corresponding value.

[0012] Further, the determination module is used to adjust the preset maintenance duration corresponding to a single road surface area to the corresponding value based on the average value of the heavy vehicle flow parameters of each fluctuating section within the single road surface area, where, The average value of the heavy vehicle flow parameters of each fluctuating section within the single road surface area is denoted as the average flow; The reduction amplitude of the preset maintenance duration is proportional to the average flow.

[0013] Further, the determination module is used to adjust the repair thickness of each section corresponding to a single road surface area to the corresponding value based on the average value of the flatness of each sensitive section within the single road surface area, where, The average value of the flatness of each sensitive section within the single road surface area is denoted as the average flatness value; The increase amplitude of the repair thickness of each section corresponding to the single road surface area is proportional to the average flatness value.

[0014] Further, the determination module is used to adjust the preset maintenance area to the corresponding value based on the average value of the flatness difference parameters of each vulnerable section within the single road surface area, where, The average value of the flatness difference parameters of each vulnerable section within the single road surface area is denoted as the average difference; The reduction amplitude of the preset maintenance area is proportional to the average difference.

[0015] Further, the determination module is used to adjust the number of road surface image information used to train the built-in model in the model calibration module to the corresponding value based on the average rainfall within the single road surface area, where, The increase in the quantity of road surface image information used to train the built-in model in the model calibration module is proportional to the average rainfall.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows. An image acquisition module, a model calibration module, a maintenance plan module, a monitoring and analysis module, a vehicle monitoring module, an evaluation module, and a determination module are provided to correspond the maintenance labels of each section based on the flatness of each section. Based on the quantity of each type of maintenance label of each section within a single road surface area obtained through statistics, it is determined whether the maintenance of each section within the single road surface area is qualified. When it is determined that the maintenance of each section within the single road surface area is abnormal, the maintenance duration, repair thickness, or maintenance area of the single road surface area is determined, or the quantity of road surface image information used to train the built-in model in the model calibration module is determined. After the road maintenance is completed, the road conditions are classified according to the continuously monitored road image information, and the subsequent maintenance parameters are adjusted according to the classification results, improving the maintenance efficiency of the road.

[0017] Furthermore, based on the flatness corresponding to the maintenance label of each section, by obtaining the gray level difference between adjacent pixel points in the road surface image information, the standard deviation of each gray level difference is calculated. The gray level difference reflects the undulation of the road surface. The greater the standard deviation, the worse the flatness of the road surface. The accuracy and real-time performance of flatness detection are improved. When the flatness is less than or equal to the first preset flatness, the standard deviation of the gray level difference between adjacent pixel points on the road surface is in a lower range. At this time, there is almost no obvious undulation or damage on the road surface, and the flatness is good. This section is marked as a stable section. When the flatness is less than or equal to the second preset flatness and greater than the first preset flatness, there is a certain degree of unevenness on the road surface, but it is not very serious. At this time, it is necessary to further evaluate in combination with the heavy vehicle flow parameters. The maintenance label of a single section is marked based on the heavy vehicle flow parameter of the single section. When the heavy vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, the passing quantity of heavy vehicles is relatively stable, and the impact on the road surface is relatively small. This section is marked as a stable section. When the heavy vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, there are certain fluctuations in the passing quantity of heavy vehicles, and the pressure on the road surface will also change. This section is marked as a fluctuating section. The road conditions are classified in a refined manner, providing an accurate basis for determining the situation of the subsequent corresponding road surface area, realizing the automatic identification and detection of road damage conditions, improving the detection efficiency, reducing the workload and subjectivity of manual inspections, improving the accuracy of detection results, and further improving the road maintenance efficiency.

[0018] Furthermore, based on the average rainfall in the road surface area corresponding to a single road section within the preset operation duration, the maintenance label of the single road section is marked. When the average rainfall is less than or equal to the first preset rainfall, the impact of rainfall on the road surface is relatively small. At this time, further evaluation is carried out based on the flatness difference parameter with the adjacent road section. When the average rainfall is greater than the first preset rainfall but less than or equal to the second preset rainfall, the road surface is relatively vulnerable to rain erosion, and this road section is marked as a sensitive road section; when the average rainfall is greater than the second preset rainfall, the road surface is greatly affected by rain, and the model calibration module fails to identify this road section in time, resulting in a situation of too high flatness, and the recognition accuracy of the road damage situation in the current environment is low. This road section is marked as a weak road section. Based on the flatness difference parameter with the adjacent road section, the maintenance label of the single road section is marked. When the flatness difference parameter is less than or equal to the preset difference parameter, the flatness difference between this road section and the adjacent road section is small. Due to the unevenness caused by overall environmental factors, this road section is marked as a sensitive road section; when the flatness difference parameter is greater than the preset difference parameter, in this case, the flatness of the single road section is quite different from that of the adjacent road section. When a vehicle passes by the adjacent damaged road, the vibration and pressure generated are transmitted to the current road section, resulting in this road section being more likely to be damaged. Therefore, this road section is marked as a vulnerable road section. According to multiple factors such as flatness, heavy vehicle flow parameter, rainfall parameter, and flatness difference parameter, the road sections are classified and marked, comprehensively evaluating the road conditions, realizing the refined classification of road conditions, providing an accurate basis for subsequent maintenance decisions, and improving the road maintenance efficiency.

[0019] Further, count the number of maintenance labels of each type within a single road surface area, and determine the maintenance label with the largest number as the problem label. According to different problem labels, analyze the reasons for abnormal maintenance, and accordingly adjust parameters such as the preset maintenance duration, repair thickness, preset maintenance area, and the number of image information for training the model. The dynamic adjustment of maintenance management is realized, and the maintenance parameters are optimized in a timely manner according to the actual road conditions and maintenance effects, improving the scientificity and effectiveness of road maintenance. If the problem label is a fluctuating section, at this time, due to the frequent driving of heavy-load vehicles in a single road surface area, the road surface is prone to damage. In this case, adjust the preset maintenance duration corresponding to the single road surface area to the corresponding value to adjust the maintenance cycle and detect road surface abnormalities in a timely manner. When the problem label is a sensitive section, at this time, since each section within a single area is prone to damage, adjust the repair thickness of each section corresponding to the single road surface area to the corresponding value to reduce the maintenance frequency; when the problem label is a vulnerable section, at this time, a large number of sections within a single road surface area are affected by the vibration and pressure generated when vehicles pass by adjacent damaged roads, resulting in the sections being more prone to damage. In this case, adjust the preset maintenance area. When the section is damaged, deal with it in a timely manner; when the problem label is a weak section, a single road surface area is easily damaged by rainwater, and the current model cannot well identify the damage characteristics of this road surface working environment. In this case, increase the training volume of the model to ensure that when there is slight damage, the damage to the road surface can be identified and maintained in a timely manner, and adjust the number of road surface image information used to train the built-in model in the model calibration module to the corresponding value. Enable the built-in model of the model calibration module to better adapt to the road surface conditions under different rainfall conditions and improve the model's ability to identify road surface damage characteristics. The maintenance efficiency of the road is improved. Description of the Drawings

[0020] Figure 1 It is a block diagram of the module of the road maintenance quality detection system based on intelligent decision-making according to an embodiment of the present invention; Figure 2 It is a logical decision-making diagram of the evaluation module based on the maintenance label corresponding to the section marked by the flatness; Figure 3 It is a logical decision-making diagram of the evaluation module based on the heavy-load vehicle flow parameters of a single section to mark the maintenance label of a single section; Figure 4 It is a logical decision-making diagram of the evaluation module based on the average rainfall of the road surface area corresponding to a single section within the preset operation duration to mark the maintenance label of a single section. Detailed Embodiments

[0021] In order to make the objectives and advantages of the present invention more clearly understood, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only for explaining the present invention and are not used to limit the present invention.

[0022] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and do not limit the protection scope of the present invention.

[0023] It should be noted that in the description of the present invention, the terms indicating directions or positional relationships such as "upper", "lower", "left", "right", "inner", "outer", etc. are based on the directions or positional relationships shown in the drawings. This is only for convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0024] In addition, it should also be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installation", "connection", and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0025] Please refer to Figure 1 、 Figure 2 、 Figure 3 and Figure 4 as shown, which are respectively the module block diagram of the road maintenance quality detection system based on intelligent decision-making according to the embodiments of the present invention, the logical decision diagram of the evaluation module for marking the maintenance label of the corresponding section based on the flatness mark, the logical decision diagram of the evaluation module for marking the maintenance label of a single section based on the heavy vehicle flow parameters of a single section, and the logical decision diagram of the evaluation module for marking the maintenance label of a single section based on the average rainfall of the road surface area corresponding to a single section within a preset operation duration; An intelligent decision-making-based road maintenance quality detection system according to an embodiment of the present invention includes: An image acquisition module, which includes a plurality of image acquisition devices respectively arranged in each section within each road surface area for acquiring road surface image information; A model calibration module, which is connected to the image acquisition module and is used to periodically calibrate the damaged areas in the road surface image information, and determine the damaged areas with a damaged area larger than the preset maintenance area as the areas to be maintained; A maintenance plan module, which is connected to the model calibration module and is used to determine maintenance parameters based on the damaged areas calibrated by the model calibration module, including maintenance area, maintenance thickness, and maintenance duration; A monitoring and analysis module, which is respectively connected to the image acquisition module, the model calibration module, and the maintenance plan module, and is used to determine the flatness based on the road surface image information of the corresponding road section re-acquired by the image acquisition module after a preset operation duration under the condition that the maintenance of each of the to-be-maintained areas is completed; A vehicle monitoring module, which is used to count the passing numbers of heavy-duty vehicles on each road section within each cycle; An evaluation module, which is respectively connected to the monitoring and analysis module and the vehicle monitoring module, and is used to mark the maintenance labels of the corresponding road sections based on the flatness of each road section; A determination module, which is respectively connected to the evaluation module, the maintenance plan module, and the model calibration module, and is used to determine whether the maintenance of each road section within a single road surface area is qualified based on the number of each type of maintenance label of each road section within a single road surface area, and when it is determined that the maintenance of each road section within a single road surface area is abnormal, adjust the preset maintenance duration corresponding to the single road surface area to a corresponding value, adjust the maintenance thickness of each road section corresponding to the single road surface area to a corresponding value, adjust the preset maintenance area to a corresponding value, or adjust the number of road surface image information used to train the built-in model in the model calibration module to a corresponding value.

[0026] Specifically, the specific structure of the model calibration module is not described. It can be any logical component. It can be understood that it is possible to input road surface image information into the model calibration module, and the model calibration module can frame the damaged areas in the input image information. This will not be elaborated here.

[0027] Specifically, each image acquirer in the image acquisition module is distributed on each road section within each road surface area. The specific structure of the image acquirer is not limited. It can be a high-definition camera installed at positions such as street lamp poles and traffic monitoring poles. It can be understood that it is only necessary to ensure that the road surface conditions of the entire road section can be covered. This will not be elaborated here.

[0028] Specifically, the model calibration module is a deep learning model. The model calibration module realizes the recognition of damaged areas such as cracks and potholes on the road surface through learning a large amount of road surface image data, and frames and marks these areas. The deep learning model can be a convolutional neural network (CNN).

[0029] Specifically, the model calibration module uses road surface image information as a training set to train the model. The damaged areas are marked in the images in the training set so that the model can learn the characteristics of these areas. The trained model is deployed into the model calibration module for processing the real-time acquired road surface images.

[0030] Specifically, the method for determining the maintenance parameters in the maintenance plan module is as follows: The specific method for determining the maintenance area is not limited. The damaged area selected by the model calibration module can be used to calculate the pixel area of this area in the image, and then combined with the shooting parameters of the image acquirer, the shooting height and angle, the pixel area is converted into the actual maintenance area. By using the known image resolution and shooting height, the actual ground area corresponding to each pixel is calculated, so as to obtain the actual area of the maintenance area, with the unit of square meters. This is the prior art and will not be elaborated here.

[0031] The method for determining the maintenance thickness is not limited. The maintenance thickness can be determined based on the maintenance area. In this embodiment, optionally, the maintenance thickness is selected within the range of [2 cm, 8 cm]. , where H is the maintenance thickness, S is the maintenance area, and D is the average gray level of the area to be maintained. It can be understood that the average gray level in this embodiment is an unsigned integer, α is the first thickness coefficient, β is the second thickness coefficient, α is taken as 0.001, and β is taken as 0.01. This is the prior art and will not be elaborated here.

[0032] The method for determining the maintenance duration is not limited. The maintenance duration is proportional to the maintenance area and the maintenance thickness. A maintenance duration calculation model can be established according to historical maintenance data. According to the determined maintenance area and maintenance thickness, substituting them into the calculation model to obtain the maintenance duration. In this embodiment, optionally, , where T is the maintenance duration, with the unit of hours, H0 is the unit thickness, and H0 is set to 1 cm, S0 is the unit area, and S0 is set to 1 cm 2 , this is the prior art and will not be elaborated here.

[0033] Specifically, the image acquisition module collects the road surface image information of each section at a set time interval and transmits it to the model calibration module. The model calibration module processes the input road surface image information, identifies the damaged area and performs frame selection and marking, and then transmits the marking information to the maintenance plan module. The maintenance plan module determines the maintenance area, maintenance thickness and maintenance duration according to the marking information and generates a maintenance plan. The maintenance personnel maintain the road surface according to the maintenance plan. After the maintenance is completed, the monitoring and analysis module triggers the image acquisition module to re-acquire the road surface image information of the corresponding section after a preset operation duration, analyzes the road surface flatness, and evaluates the maintenance quality.

[0034] Specifically, the specific method for the vehicle monitoring module to count the passing numbers of heavy-duty vehicles on each road section in each cycle is not limited. The vehicle monitoring module may include high-definition monitoring cameras installed on the road sections for capturing images and videos of vehicles, obtaining the outlines of the vehicles captured by the high-definition monitoring cameras, and marking the vehicles with an outline area greater than a preset outline area as heavy-duty vehicles, so as to count the passing numbers of heavy-duty vehicles on each road section in each cycle. This is the prior art and will not be elaborated herein.

[0035] Specifically, a single road surface area includes several road sections.

[0036] Specifically, an image acquisition module, a model calibration module, a maintenance plan module, a monitoring and analysis module, a vehicle monitoring module, an evaluation module, and a determination module are set up to mark the maintenance labels of the corresponding road sections based on the flatness of each road section; determine whether the maintenance of each road section in a single road surface area is qualified based on the counted numbers of various types of maintenance labels of each road section in the single road surface area, and when it is determined that the maintenance of each road section in the single road surface area is abnormal, adjust the preset maintenance duration corresponding to the single road surface area to a corresponding value, adjust the repair thickness of each road section corresponding to the single road surface area to a corresponding value, adjust the preset maintenance area to a corresponding value, or adjust the number of road surface image information used to train the built-in model in the model calibration module to a corresponding value. After the road maintenance is completed, classify the road conditions based on the continuously monitored road image information, and adjust the subsequent maintenance parameters according to the classification results, thereby improving the maintenance efficiency of the road.

[0037] Specifically, the monitoring and analysis module is used to determine the flatness based on the road surface image information, obtain the gray difference values between adjacent pixel points in the road surface image information, calculate the standard deviation of each gray difference value, and obtain the flatness of a single road surface image information; The evaluation module is used to mark the maintenance labels of the corresponding road sections based on the flatness, including: If the flatness is less than or equal to the first preset flatness, determine the maintenance label of a single road section as a stable road section; If the flatness is less than or equal to the second preset flatness and greater than the first preset flatness, mark the maintenance label of a single road section in combination with the heavy-duty vehicle flow parameter of the single road section; If the flatness is greater than the second preset flatness, mark the maintenance label of a single road section based on the average rainfall in the road surface area corresponding to the single road section within the preset operation duration.

[0038] Specifically, the first preset flatness is selected within the range of [0.17, 0.22], and the second preset flatness is selected within the range of [0.53, 0.64].

[0039] Specifically, the evaluation module is used to mark the maintenance label of a single road section based on the heavy vehicle flow parameters of the single road section, including: Obtain the passing quantity of heavy vehicles on a single road section within several cycles, calculate the variance of the passing quantity in each cycle, and obtain the heavy vehicle flow parameter; If the heavy vehicle flow parameter is less than or equal to the first preset flow parameter, determine the maintenance label of the single road section as a stable road section; If the heavy vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, determine the maintenance label of the single road section as a fluctuating road section; If the heavy vehicle flow parameter is greater than the second preset flow parameter, mark the maintenance label of the single road section based on the average rainfall in the road surface area corresponding to the single road section within the preset operation duration.

[0040] Specifically, the first preset flow parameter is selected within the range of [7, 10], and the second preset flow parameter is selected within the range of [18, 20].

[0041] Specifically, mark the maintenance label of the corresponding road section based on the flatness. By obtaining the gray difference between adjacent pixel points in the road surface image information, calculate the standard deviation of each gray difference. The gray difference reflects the undulation of the road surface. The larger the standard deviation, the worse the flatness of the road surface. This improves the accuracy and real-time performance of flatness detection. When the flatness is less than or equal to the first preset flatness, the standard deviation of the gray difference between adjacent pixel points on the road surface is in a lower range. At this time, the road surface has almost no obvious undulation or damage, and the flatness is good. Mark this road section as a stable road section. When the flatness is less than or equal to the second preset flatness and greater than the first preset flatness, there is a certain degree of unevenness on the road surface, but it is not very serious. At this time, it is necessary to further evaluate in combination with the heavy vehicle flow parameter. Mark the maintenance label of the single road section based on the heavy vehicle flow parameter of the single road section. When the heavy vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, the passing quantity of heavy vehicles is relatively stable, and the impact on the road surface is relatively small. Mark this road section as a stable road section. When the heavy vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, the passing quantity of heavy vehicles has a certain fluctuation, and the pressure on the road surface will also change. Mark this road section as a fluctuating road section. Conduct refined classification of the road conditions, provide an accurate basis for determining the situation of the subsequent corresponding road surface area, realize automatic identification and detection of road damage conditions, improve the detection efficiency, reduce the workload and subjectivity of manual inspections, improve the accuracy of detection results, and further improve the road maintenance efficiency.

[0042] Specifically, the evaluation module is used to mark the maintenance label of a single road section based on the average rainfall in the road surface area corresponding to the single road section within a preset operation duration, including: If the average rainfall is less than or equal to the first preset rainfall, mark the maintenance label of the single road section based on the flatness difference parameter between the single road section and the adjacent road section; If the average rainfall is less than or equal to the second preset rainfall and greater than the first preset rainfall, determine the maintenance label of the single road section as a sensitive road section; If the average rainfall is greater than the second preset rainfall, determine the maintenance label of the single road section as a weak road section.

[0043] Specifically, the first preset rainfall is selected within the range of [17mm, 20mm], and the second preset rainfall is selected within the range of [27mm, 30mm].

[0044] Specifically, the evaluation module is used to mark the maintenance label of a single road section based on the flatness difference parameter between the single road section and the adjacent road section, including: Calculate the difference between the flatness of the single road section and the flatness of the single road section connected to the single road section obtained, to obtain the flatness difference; Calculate the average value of the flatness differences corresponding to the two road sections connected to the single road section, to obtain the flatness difference parameter; If the flatness difference parameter is less than or equal to the preset difference parameter, determine the maintenance label of the single road section as a sensitive road section; If the flatness difference parameter is greater than the preset difference parameter, determine the maintenance label of the single road section as a vulnerable road section.

[0045] Specifically, the preset difference parameter is selected within the range of [0.15, 0.22].

[0046] Specifically, the maintenance label of a single road segment is marked based on the average rainfall in the road surface area corresponding to the single road segment within the preset operation duration. When the average rainfall is less than or equal to the first preset rainfall, the impact of rainfall on the road surface is relatively small. At this time, further evaluation is carried out based on the flatness difference parameter with the adjacent road segment. When the average rainfall is greater than the first preset rainfall but less than or equal to the second preset rainfall, the road surface is relatively vulnerable to rain erosion, and this road segment is marked as a sensitive road segment; when the average rainfall is greater than the second preset rainfall, the road surface is greatly affected by rain, and the model calibration module fails to identify this road segment in time, resulting in a situation of too high flatness, and the recognition accuracy of the road damage condition in the current environment is low, and this road segment is marked as a weak road segment. The maintenance label of a single road segment is marked based on the flatness difference parameter with the adjacent road segment. When the flatness difference parameter is less than or equal to the preset difference parameter, the flatness difference between this road segment and the adjacent road segment is small. Due to the unevenness caused by the overall environmental factors, this road segment is marked as a sensitive road segment; when the flatness difference parameter is greater than the preset difference parameter, in this case, the flatness of a single road segment is quite different from that of the adjacent road segment. When a vehicle passes by the adjacent damaged road, the vibration and pressure generated are involved in the current road segment, resulting in this road segment being more vulnerable to damage. Therefore, this road segment is marked as a vulnerable road segment. According to multiple factors such as flatness, heavy vehicle flow parameter, rainfall parameter, and flatness difference parameter, the road segments are classified and marked, comprehensively evaluating the road conditions, realizing the refined classification of the road conditions, providing an accurate basis for subsequent maintenance decisions, and improving the road maintenance efficiency.

[0047] Specifically, the determination module is used to determine whether the maintenance of each road segment in a single road surface area is qualified based on the quantity of each type of maintenance label of each road segment in the statistically single road surface area, including: Count the quantity of each type of maintenance label in a single road surface area, and determine the label with the largest quantity as the problem label; If the problem label is a stable road segment, it is determined that the maintenance of each road segment in a single road surface area is qualified, and each module is controlled to continue to operate with the current operating parameters; If the problem label is a fluctuating road segment, it is determined that the maintenance of each road segment in a single road surface area is abnormal, and the preset maintenance duration corresponding to the single road surface area is adjusted to the corresponding value based on the average value of the heavy vehicle flow parameters of each fluctuating road segment in the single road surface area; If the problem label is a sensitive road segment, it is determined that the maintenance of each road segment in a single road surface area is abnormal, and the repair thickness of each road segment corresponding to the single road surface area is adjusted to the corresponding value based on the average value of the flatness of each sensitive road segment in the single road surface area; If the problem label is a vulnerable section, determine the maintenance anomaly for each section within a single road surface area, and adjust the preset maintenance area to the corresponding value based on the average of the flatness difference parameters of each vulnerable section within the single road surface area; If the problem label is a weak section, determine the maintenance anomaly for each section within a single road surface area, and adjust the quantity of road surface image information used to train the built-in model in the model calibration module to the corresponding value based on the average rainfall within the single road surface area.

[0048] Specifically, use the newly determined training set to perform model fine-tuning on the model.

[0049] Specifically, under the condition that the problem label is a weak section, the road surface image information used to train the built-in model in the model calibration module is all selected from the road surface image information obtained within the road surface area where the average rainfall is greater than the second preset rainfall.

[0050] Specifically, the maintenance labels of each type for each section within a single road surface area include stable sections, fluctuating sections, sensitive sections, vulnerable sections, and weak sections.

[0051] Specifically, count the number of maintenance tags of each type within a single road surface area, and determine the problem tag as the maintenance tag with the largest quantity. Based on different problem tags, analyze the reasons for abnormal maintenance, and accordingly adjust parameters such as the preset maintenance duration, repair thickness, preset maintenance area, and the quantity of image information for training the model. The dynamic adjustment of maintenance management is realized, and the maintenance parameters are optimized in a timely manner according to the actual road conditions and maintenance effects, improving the scientificity and effectiveness of road maintenance. If the problem tag is a fluctuating section, at this time, due to the frequent driving of heavy-duty vehicles in a single road surface area, the road surface is prone to damage. In this case, adjust the preset maintenance duration corresponding to the single road surface area to the corresponding value to adjust the maintenance cycle and detect road surface abnormalities in a timely manner. When the problem tag is a sensitive section, at this time, since each section within a single area is prone to damage, adjust the repair thickness of each section corresponding to the single road surface area to the corresponding value to reduce the maintenance frequency; when the problem tag is a vulnerable section, at this time, a large number of sections within a single road surface area are damaged due to the vibration and pressure generated when vehicles pass by adjacent damaged roads, resulting in the current section being more prone to damage. In this case, adjust the preset maintenance area. When the section is damaged, deal with it in a timely manner; when the problem tag is a weak section, a single road surface area is easily damaged by rainwater, and the current model cannot well identify the damage characteristics of this road surface working environment. In this case, increase the training volume of the model to ensure that when there is slight damage, the damage to the road surface is identified and maintained in a timely manner, and adjust the quantity of road surface image information used to train the built-in model in the model calibration module to the corresponding value. Enable the built-in model of the model calibration module to better adapt to the road surface conditions under different rainfall conditions, improve the model's ability to identify road surface damage characteristics, and improve the maintenance efficiency of the road.

[0052] Specifically, the determination module is used to adjust the preset maintenance duration corresponding to a single road surface area to the corresponding value based on the average value of the heavy-duty vehicle flow parameters of each fluctuating section within the single road surface area, where, The average value of the heavy-duty vehicle flow parameters of each fluctuating section within a single road surface area is denoted as the average flow; The reduction amplitude of the preset maintenance duration is proportional to the average flow.

[0053] In this embodiment, optionally, Compare the average flow with a first preset average flow and a second preset average flow; If the average flow is less than or equal to the first preset average flow, then adjust the preset maintenance duration corresponding to the single road surface area to 0.92 times the initial preset maintenance duration; If the average flow is less than or equal to the second preset average flow and greater than the first preset average flow, then adjust the preset maintenance duration corresponding to the single road surface area to 0.83 times the initial preset maintenance duration; If the average flow rate is greater than the second preset average flow rate, adjust the preset maintenance duration corresponding to a single road surface area to 0.73 times the initial preset maintenance duration; The first preset average flow rate is taken as 12, and the second preset average flow rate is taken as 16.

[0054] Specifically, the determination module is used to adjust the repair thickness of each road section corresponding to a single road surface area to a corresponding value based on the average value of the flatness of each sensitive road section within the single road surface area, where Denote the average value of the flatness of each sensitive road section within the single road surface area as the flatness average value; The increase amplitude of the repair thickness of each road section corresponding to a single road surface area is proportional to the flatness average value.

[0055] In this embodiment, optionally, Compare the flatness average value with the first preset flatness average value and the second preset flatness average value; If the flatness average value is less than or equal to the first preset flatness average value, adjust the repair thickness of each road section corresponding to a single road surface area to 1.1 times the initial repair thickness; If the flatness average value is less than or equal to the second preset flatness average value and greater than the first preset flatness average value, adjust the repair thickness of each road section corresponding to a single road surface area to 1.2 times the initial repair thickness; If the flatness average value is greater than the second preset flatness average value, adjust the repair thickness of each road section corresponding to a single road surface area to 1.3 times the initial repair thickness; The first preset flatness average value is taken as 0.4, and the second preset flatness average value is taken as 0.8.

[0056] Specifically, the determination module is used to adjust the preset maintenance area to a corresponding value based on the average value of the flatness difference parameters of each vulnerable road section within the single road surface area, where Denote the average value of the flatness difference parameters of each vulnerable road section within the single road surface area as the average difference; The reduction amplitude of the preset maintenance area is proportional to the average difference.

[0057] In this embodiment, optionally, Compare the average difference with the first preset average difference and the second preset average difference; If the average difference is less than or equal to the first preset average difference, adjust the preset maintenance area to 0.94 times the initial preset maintenance area; If the average difference is less than or equal to the second preset average difference and greater than the first preset average difference, adjust the preset maintenance area to 0.87 times the initial preset maintenance area; If the average difference is greater than the second preset average difference, the preset maintenance area is adjusted to 0.77 times the initial preset maintenance area; The first preset average difference is taken as 0.24, and the second preset average difference is taken as 0.28.

[0058] Specifically, the determination module is used to adjust the quantity of road surface image information for training the built-in model in the model calibration module to a corresponding value based on the average rainfall in a single road surface area, where The increase amplitude of the quantity of road surface image information for training the built-in model in the model calibration module is proportional to the average rainfall.

[0059] In this embodiment, optionally, The average rainfall is compared with the first preset rainfall comparison threshold and the second preset rainfall comparison threshold; If the average rainfall is less than or equal to the first preset rainfall comparison threshold, the quantity of road surface image information for training the model is adjusted to 1.14 times the initial quantity; If the average rainfall is less than or equal to the second preset rainfall comparison threshold and greater than the first preset rainfall comparison threshold, the quantity of road surface image information for training the model is adjusted to 1.25 times the initial quantity; If the average rainfall is greater than the second preset rainfall comparison threshold, the quantity of road surface image information for training the model is adjusted to 1.32 times the initial quantity; The first preset rainfall comparison threshold is taken as 32 mm, and the second preset rainfall comparison threshold is taken as 37 mm.

[0060] Specifically, the maintenance cycle is dynamically adjusted according to the fluctuation of the passing of heavy-duty vehicles. When the fluctuation of the passing of heavy-duty vehicles is large, the maintenance frequency is increased to timely discover and handle road surface problems, improving the safety and service life of the road. It avoids the problems of untimely maintenance or over-maintenance that may be caused by a fixed maintenance cycle.

[0061] So far, the technical solution of the present invention has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the protection scope of the present invention.

[0062] The above are only the preferred embodiments of the present invention and are not used to limit the present invention; for those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A road maintenance quality detection system based on intelligent decision-making, characterized in that Including: An image acquisition module, which includes a plurality of image acquirers respectively arranged in each section of each road surface area for acquiring road surface image information; A model calibration module, which is connected to the image acquisition module and is used for periodically calibrating the damaged areas in the road surface image information, and determining the areas to be maintained as the damaged areas where the damaged area is larger than the preset maintenance area; A maintenance plan module, which is connected to the model calibration module and is used for determining maintenance parameters based on the damaged areas calibrated by the model calibration module, including maintenance area, maintenance thickness and maintenance duration; A monitoring and analysis module, which is respectively connected to the image acquisition module, the model calibration module and the maintenance plan module, and is used for determining the flatness based on the road surface image information of the corresponding section re-acquired by the image acquisition module after a preset operation duration under the condition that the maintenance of each area to be maintained is completed; A vehicle monitoring module, which is used for counting the passing numbers of heavy-duty vehicles on each section in each cycle; An evaluation module, which is respectively connected to the monitoring and analysis module and the vehicle monitoring module, and is used for marking the maintenance labels of the corresponding sections based on the flatness of each section; A determination module, which is respectively connected to the evaluation module, the maintenance plan module and the model calibration module, and is used for determining whether the maintenance of each section in a single road surface area is qualified based on the number of each type of maintenance label of each section in the single road surface area, and when it is determined that the maintenance of each section in the single road surface area is abnormal, determining the maintenance duration, maintenance thickness or maintenance area of the single road surface area, or determining the number of road surface image information used for training the built-in model in the model calibration module.

2. The road maintenance quality detection system based on intelligent decision-making according to claim 1, characterized in that The monitoring and analysis module is used for determining the flatness based on the road surface image information, for obtaining the gray difference between adjacent pixel points in the road surface image information, calculating the standard deviation of each gray difference, and obtaining the flatness of a single road surface image information; The evaluation module is used for marking the maintenance label of the corresponding section based on the flatness, including: If the flatness is less than or equal to the first preset flatness, the maintenance label of a single section is determined as a stable section; If the flatness is less than or equal to the second preset flatness and greater than the first preset flatness, the maintenance label of a single section is marked in combination with the heavy-duty vehicle flow parameter of the single section; If the flatness is greater than the second preset flatness, the maintenance label of a single section is marked based on the average rainfall in the road surface area corresponding to the single section within the preset operation duration.

3. The road maintenance quality detection system based on intelligent decision-making according to claim 2, characterized in that, The evaluation module is used for marking the maintenance label of a single section based on the heavy-duty vehicle flow parameter of the single section, including: Obtaining the passing numbers of heavy-duty vehicles on a single section in a plurality of cycles, calculating the variance of the passing numbers in each cycle, and obtaining the heavy-duty vehicle flow parameter; If the heavy-duty vehicle flow parameter is less than or equal to the first preset flow parameter, the maintenance label of a single section is determined as a stable section; If the heavy-duty vehicle flow parameter is less than or equal to the second preset flow parameter and greater than the first preset flow parameter, the maintenance label of a single section is determined as a fluctuating section; If the flow parameter of the overloaded vehicle is greater than the second preset flow parameter, mark the maintenance label of a single road section based on the average rainfall in the road surface area corresponding to the single road section within the preset operation duration.

4. The road maintenance quality detection system based on intelligent decision-making according to claim 3, wherein The evaluation module is used to mark the maintenance label of a single road section based on the average rainfall in the road surface area corresponding to the single road section within the preset operation duration, including: If the average rainfall is less than or equal to the first preset rainfall, mark the maintenance label of the single road section based on the flatness difference parameter between the single road section and the adjacent road section; If the average rainfall is less than or equal to the second preset rainfall and greater than the first preset rainfall, determine the maintenance label of the single road section as a sensitive road section; If the average rainfall is greater than the second preset rainfall, determine the maintenance label of the single road section as a weak road section.

5. The road maintenance quality detection system based on intelligent decision-making according to claim 4, characterized in that, The evaluation module is used to mark the maintenance label of a single road section based on the flatness difference parameter between the single road section and the adjacent road section, including: Calculate the difference between the flatness of the single road section and the flatness of the single road section connected to the single road section obtained, to obtain the flatness difference; Calculate the average value of the flatness difference corresponding to the two road sections connected to the single road section, to obtain the flatness difference parameter; If the flatness difference parameter is less than or equal to the preset difference parameter, determine the maintenance label of the single road section as a sensitive road section; If the flatness difference parameter is greater than the preset difference parameter, determine the maintenance label of the single road section as a vulnerable road section.

6. The road maintenance quality detection system based on intelligent decision-making according to claim 5, wherein, The determination module is used to determine whether the maintenance of each road section in a single road surface area is qualified based on the number of maintenance labels of each type of each road section in the single road surface area, including: Count the number of maintenance labels of each type in the single road surface area, and determine the maintenance label with the largest number as the problem label; If the problem label is a stable road section, determine that the maintenance of each road section in the single road surface area is qualified, and control each module to continue to run with the current operation parameters; If the problem label is a fluctuating road section, determine that the maintenance of each road section in the single road surface area is abnormal, and adjust the preset maintenance duration corresponding to the single road surface area to the corresponding value based on the average value of the flow parameters of the overloaded vehicles in each fluctuating road section in the single road surface area; If the problem label is a sensitive road section, determine that the maintenance of each road section in the single road surface area is abnormal, and adjust the repair thickness of each road section corresponding to the single road surface area to the corresponding value based on the average value of the flatness of each sensitive road section in the single road surface area; If the problem label is a vulnerable road section, determine that the maintenance of each road section in the single road surface area is abnormal, and adjust the preset maintenance area to the corresponding value based on the average value of the flatness difference parameters of each vulnerable road section in the single road surface area; If the problem label is a weak road section, determine that the maintenance of each road section in the single road surface area is abnormal, and adjust the number of road surface image information for training the built-in model in the model calibration module to the corresponding value based on the average rainfall in the single road surface area.

7. The road maintenance quality detection system based on intelligent decision-making according to claim 6, characterized in that, The determination module is used to adjust the preset maintenance duration corresponding to the single road surface area to the corresponding value based on the average value of the flow parameters of the overloaded vehicles in each fluctuating road section in the single road surface area, where Record the average value of the flow parameters of the overloaded vehicles in each fluctuating road section in the single road surface area as the average flow; The reduction amplitude of the preset curing duration is proportional to the average flow rate.

8. The road maintenance quality detection system based on intelligent decision-making according to claim 7, characterized in that The determination module is used to adjust the repair thickness of each road section corresponding to a single road surface area to a corresponding value based on the average flatness of each sensitive road section within the single road surface area, where the average of the flatness of each sensitive road section within the single road surface area is denoted as the average flatness; the increase amplitude of the repair thickness of each road section corresponding to the single road surface area is proportional to the average flatness.

9. The road maintenance quality detection system based on intelligent decision-making according to claim 8, characterized in that, The determination module is used to adjust the preset maintenance area to a corresponding value based on the average value of the flatness difference parameters of each vulnerable road section within the single road surface area, where the average value of the flatness difference parameters of each vulnerable road section within the single road surface area is denoted as the average difference; the reduction amplitude of the preset maintenance area is proportional to the average difference.

10. The road maintenance quality detection system based on intelligent decision-making according to claim 9, characterized in that, The determination module is used to adjust the quantity of road surface image information for training the built-in model in the model calibration module to a corresponding value based on the average rainfall within the single road surface area, where the increase amplitude of the quantity of road surface image information for training the built-in model in the model calibration module is proportional to the average rainfall.

Images