A road surface defect monitoring and prediction method and system

Abstract

The present invention relates to the field of urban traffic management technology, and in particular to a road surface defect monitoring and prediction method and system, the method comprising the following steps: constructing a road surface digital twin model of a target road surface; constructing a historical road surface monitoring data set, and obtaining accurate road surface monitoring index data and high-definition road surface image data in real time; according to the road surface digital twin model and the road surface monitoring data set, using a dual-source weight fusion method to obtain the index weights of each road surface monitoring index; updating the road surface digital twin model based on the accurate road surface monitoring index data, index weights and high-definition road surface image data, to achieve real-time monitoring and prediction of target road surface defects, and formulate a road surface maintenance strategy. The present invention is based on the road surface digital twin model, takes into account the influence of various road surface monitoring indicators on road surface defects, and achieves real-time, accurate and efficient monitoring and prediction of road surface defects, and then timely formulates a road surface maintenance strategy to ensure the road's traffic capacity and safety.

Description

Technical Field

[0001] The present invention relates to the technical field of urban traffic management, and in particular to a road surface defect monitoring and prediction method and system. Background Art

[0002] In the rapidly developing modern transportation system, roads are important infrastructure that connect cities and villages and promote economic exchanges. The condition of roads is directly related to the smoothness and safety of traffic. However, with the surge in the number of vehicles and the extension of the service life of roads, the problem of road surface defects has become increasingly prominent and has become one of the key factors restricting traffic efficiency and safety. Road surface defects such as cracks and potholes not only affect driving comfort, but may also cause traffic accidents, posing a serious threat to people's lives and property safety.

[0003] Traditional methods of monitoring road surface defects mainly rely on manual inspections and regular testing. Although manual inspections are intuitive, they are limited by factors such as labor costs, inspection frequency, and subjective judgment, making it difficult to achieve efficient and accurate monitoring of large areas of roads. Regular inspections, on the other hand, have time lags and often fail to detect and deal with road surface defects in a timely manner, resulting in the accumulation of problems and exacerbating damage to road structures and driving safety. In addition, with the rise of intelligent transportation systems, the need for real-time monitoring and early warning of road conditions is becoming increasingly urgent. Traditional monitoring methods have obvious deficiencies in data processing, information integration, and intelligent analysis, and are difficult to meet the needs of modern traffic management.

[0004] As an advanced technology that has emerged in recent years, digital twin technology provides a new idea for road defect monitoring and prediction. This technology can integrate a variety of sensors and monitoring equipment to achieve real-time and comprehensive monitoring of road conditions. It can also build a digital mirror of the physical world to achieve a comprehensive, real-time, and dynamic simulation of physical entities and their operating environment, discover potential road defects in advance, provide scientific decision-making support for road maintenance, effectively reduce maintenance costs and traffic accident risks, and promote the intelligent development of the transportation field. At present, there are some methods for monitoring road defects using digital twin technology, but these methods simply use monitoring data to directly update the digital twin model, and the accuracy of road defect monitoring and prediction needs to be improved. Summary of the invention

[0005] In view of the defects in the prior art, the present invention provides a road surface defect monitoring and prediction method and system.

[0006] In order to achieve the above-mentioned purpose, in the first aspect, the present invention provides a method for monitoring and predicting road surface defects, the method comprising the following steps: constructing a road surface digital twin model of the target road surface; collecting historical road surface monitoring data to construct a historical road surface monitoring data set, and obtaining accurate road surface monitoring index data and road surface high-definition image data in real time; according to the road surface digital twin model and the road surface monitoring data set, using a dual-source weight fusion method to obtain the index weights of each road surface monitoring index; updating the road surface digital twin model based on the accurate road surface monitoring index data, the index weights and the road surface high-definition image data, to achieve real-time monitoring of target road surface defects; predicting the accurate road surface monitoring index data, and then predicting the target road surface defects according to the index weights and the road surface digital twin model, and formulating a road surface maintenance strategy. Based on the road surface digital twin model, the present invention takes into account the influence of various road surface monitoring indicators on road surface defects, realizes real-time, accurate and efficient monitoring and prediction of road surface defects, and then timely formulates road surface maintenance strategies to ensure the traffic capacity and safety of roads.

[0007] Optionally, the historical road surface monitoring data includes historical road surface monitoring index data and historical road surface image data;

[0008] The collecting of historical road surface monitoring data to construct a historical road surface monitoring data set and obtaining accurate road surface monitoring index data and high-definition road surface image data in real time includes the following steps:

[0009] Collecting historical road surface monitoring data, and preprocessing the historical road surface monitoring data, thereby constructing a historical road surface monitoring data set;

[0010] The road surface monitoring index data and the road surface image data are acquired in real time, and preprocessed to obtain the accurate road surface monitoring index data and the high-definition road surface image data.

[0011] Furthermore, the indicator weights are obtained based on the historical pavement monitoring data set, and the importance of each pavement monitoring indicator is taken into account when updating the pavement information on the pavement digital twin model. In addition, the precise data of pavement monitoring indicators and the high-definition image data of the pavement can accurately reflect the real-time condition of the pavement, making the pavement defect monitoring and prediction results based on the pavement digital twin model more objective and accurate.

[0012] Optionally, the real-time acquisition of road surface monitoring index data and road surface image data, and preprocessing to obtain the accurate road surface monitoring index data and the high-definition road surface image data comprises the following steps:

[0013] Performing data cleaning on the road surface monitoring indicator data to obtain accurate data of the road surface monitoring indicator;

[0014] The road surface image data is smoothed and denoised, and the contrast is enhanced to obtain the road surface high-definition image data.

[0015] Furthermore, by preprocessing the road surface monitoring index data and road surface image data so that they can accurately reflect the real-time condition of the road surface, the updated road surface digital twin model can also accurately reflect the real-time condition of the road surface, thereby realizing accurate monitoring of the road surface.

[0016] Optionally, the step of obtaining the indicator weight of each road surface monitoring indicator by using a dual-source weight fusion method according to the road surface digital twin model and the road surface monitoring data set comprises the following steps:

[0017] Performing image recognition on the historical road surface image data in the road surface monitoring data set to obtain road surface defects, wherein the road surface defects include cracks and potholes;

[0018] Calculate the pavement defect evaluation factor value corresponding to each group of pavement monitoring data in the pavement monitoring data set according to the image recognition result;

[0019] Based on the pavement monitoring data set and the pavement defect evaluation factor value, using a permutation importance method to obtain a first weight of the pavement monitoring index;

[0020] Obtaining a second weight of the pavement monitoring index using an item-by-item analysis method according to the pavement digital twin model, the pavement monitoring data set, and the pavement defect evaluation factor value;

[0021] The indicator weight is calculated according to the first weight and the second weight.

[0022] Furthermore, the first weight is obtained based on the machine learning method, and the second weight is obtained based on the statistical test method. Combining the two to obtain the indicator weight can avoid the problem of one-sidedness of the weight obtained by using a single method and improve the accuracy of the road defect monitoring prediction results.

[0023] Optionally, the obtaining the second weight of the pavement monitoring index by using an item-by-item analysis method according to the pavement digital twin model, the pavement monitoring data set and the pavement defect evaluation factor value comprises the following steps:

[0024] Randomly selecting a group of road surface monitoring data from the road surface monitoring data set as the array to be adjusted, and the remaining groups of road surface monitoring data as the adjustment arrays;

[0025] For a certain road surface monitoring indicator, the value of the road surface monitoring indicator in the adjustment array is used to replace the value of the road surface monitoring indicator in the array to be adjusted in sequence to obtain multiple groups of adjusted arrays, and the road surface monitoring indicator is marked as a road surface monitoring change indicator;

[0026] The adjusted array is used to update the pavement digital twin model, pavement defects are identified and multiple pavement defect evaluation factor comparison values ​​are calculated, and then the second weight of the pavement monitoring change index is calculated in combination with the pavement defect evaluation factor values.

[0027] Optionally, the second weight satisfies the following relationship:

[0028]

[0029] in, is the second weight, N is the number of the adjustment arrays, is the comparison value of the ith road surface defect evaluation factor, is the road surface defect evaluation factor value, is the value of the road surface monitoring change index in the i-th adjustment array, is the value of the road surface monitoring change index in the array to be adjusted.

[0030] Optionally, updating the pavement digital twin model based on the pavement monitoring index precise data, the index weights and the pavement high-definition image data to achieve real-time monitoring of target pavement defects comprises the following steps:

[0031] Calculate weighted road surface monitoring index data according to the road surface monitoring index accurate data and the index weight;

[0032] The weighted road surface monitoring index data and the road surface high-definition image data are used to update the road surface digital twin model to obtain a real-time twin road surface, thereby realizing real-time monitoring of the target road surface defects.

[0033] Furthermore, by combining the precise data of pavement monitoring indicators and indicator weights to calculate weighted pavement monitoring indicator data, and then fully considering the impact of different pavement monitoring indicators on pavement conditions, and introducing pavement high-definition image data to realize multi-source data fusion, the pavement digital twin model can more comprehensively reflect the actual conditions of the pavement, thereby improving the real-time and comprehensiveness of monitoring.

[0034] Optionally, predicting the precise data of the pavement monitoring index, predicting the target pavement defect according to the index weight and the pavement digital twin model, and formulating a pavement maintenance strategy comprises the following steps:

[0035] Predict the precise data of the road surface monitoring indicators to obtain the future indicator prediction data of each road surface monitoring indicator;

[0036] Calculate weighted road surface monitoring index prediction data according to the index prediction data and the index weight;

[0037] Using the weighted pavement monitoring index prediction data to update the real-time twin pavement to obtain the future twin pavement, thereby realizing the prediction of the target pavement defects;

[0038] Formulate a current pavement maintenance strategy based on the weighted pavement monitoring indicator data;

[0039] A future pavement maintenance strategy is formulated based on the real-time twin pavement, the weighted pavement monitoring index prediction data and the future twin pavement, and the current pavement maintenance strategy is integrated to obtain the pavement maintenance strategy.

[0040] Furthermore, by predicting the future of target road surface defects, road maintenance work can be planned in advance to avoid worsening of road surface problems. When formulating road maintenance strategies, integrating current road maintenance strategies with future road maintenance strategies can comprehensively consider the current and future states of the road surface, achieve the combination of long-term planning and short-term response, and formulate more scientific and reasonable maintenance strategies, which will help optimize the allocation of maintenance resources, reduce unnecessary waste, and improve the efficiency and benefits of maintenance work.

[0041] Optionally, formulating a future pavement maintenance strategy based on the real-time twin pavement, the weighted pavement monitoring index prediction data and the future twin pavement, and integrating the current pavement maintenance strategy to obtain the pavement maintenance strategy includes the following steps:

[0042] Arrange the weighted road surface monitoring index prediction data in descending order;

[0043] Deleting the smallest data in the weighted road surface monitoring index prediction data in sequence, and using the remaining weighted road surface monitoring index prediction data after each deletion to update the real-time twin road surface to obtain a comparison twin road surface, and calculating the similarity between the comparison twin road surface and the future twin road surface;

[0044] When the similarity is less than the similarity threshold for the first time, the operation of deleting the weighted road surface monitoring index prediction data is stopped, and the weighted road surface monitoring index prediction data remaining after the last deletion operation is used as the final judgment data;

[0045] A future pavement maintenance strategy is formulated according to the final judgment data, and the current pavement maintenance strategy is integrated to obtain the pavement maintenance strategy.

[0046] Furthermore, by sorting the weighted pavement monitoring index prediction data by size and deleting the smallest data one by one, and using the remaining weighted pavement monitoring index prediction data to update the real-time twin pavement, and then calculating the similarity with the future twin pavement, it is possible to more accurately identify the pavement monitoring index combination that has the greatest impact on the future state of the pavement, thereby integrating the current pavement maintenance strategy to obtain a more targeted pavement maintenance strategy, thereby minimizing maintenance costs while ensuring maintenance results.

[0047] In the second aspect, the present invention also provides a pavement defect monitoring and prediction system, which uses a pavement defect monitoring and prediction method provided by the present invention, and the system includes: a digital twin model construction module, which is used to construct a pavement digital twin model of the target pavement; a data acquisition and preprocessing module, which is used to collect historical pavement monitoring data to construct a historical pavement monitoring data set, and obtain real-time accurate pavement monitoring indicator data and pavement high-definition image data; a pavement defect monitoring and prediction module, which is used to obtain the indicators of each pavement monitoring indicator based on the pavement digital twin model and the pavement monitoring data set using a dual-source weight fusion method. The road surface monitoring index is used to collect and analyze the road surface defects, and the road surface monitoring index is used to collect and analyze the road surface defects. ... BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A schematic diagram of a process flow of a road surface defect monitoring and prediction method according to an embodiment of the present invention;

[0050] Figure 2A schematic diagram of the framework of a road surface defect monitoring and prediction system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments described herein are only for illustration and are not intended to limit the present invention. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that these specific details do not need to be adopted to implement the present invention. In other examples, in order to avoid confusing the present invention, known circuits, software or methods are not specifically described.

[0052] Throughout the specification, references to "one embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment of the present invention. Therefore, the phrases "in one embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily all refer to the same embodiment or example. In addition, particular features, structures, or characteristics may be combined in one or more embodiments or examples in any suitable combination and / or subcombination. In addition, it should be understood by those of ordinary skill in the art that the figures provided herein are for illustrative purposes and that the figures are not necessarily drawn to scale.

[0053] It should be noted in advance that, in an optional embodiment, except for independent explanations, the same symbols or letters appearing in all formulas have the same meanings and values.

[0054] In an alternative embodiment, see Figure 1 The present invention provides a road surface defect monitoring and prediction method, the method comprising the following steps:

[0055] S1. Build a digital twin model of the target road surface.

[0056] Specifically, in this embodiment, SketchUp is used to construct a road surface digital twin model using road geometry information, road texture information, and environmental information in the road design CAD drawings.

[0057] The pavement digital twin model can reflect the changes in the physical world's pavement in real time, can reflect the current pavement conditions in real time, and can simulate and predict pavement defects based on historical data and real-time data, and early detection and prevention of pavement cracks and potholes. At the same time, the pavement digital twin model can present complex pavement conditions in an intuitive and easy-to-understand way. Users can make various assumptions and tests through the model to evaluate the effects of different maintenance plans and make more scientific decisions. In addition, by building a pavement digital twin model, pavement management can be made more data-based and intelligent, improving the efficiency and effectiveness of pavement management.

[0058] Furthermore, in other optional embodiments, other software may be used to construct a road surface digital twin model, such as 3DS MAX, Revit, and Unity3D.

[0059] S2. Collect historical road surface monitoring data to construct a historical road surface monitoring data set, and obtain accurate road surface monitoring indicator data and high-definition road surface image data in real time.

[0060] Wherein, step S2 specifically includes the following steps:

[0061] S21. Collect historical road surface monitoring data, and pre-process the historical road surface monitoring data to construct a historical road surface monitoring data set.

[0062] Specifically, in this embodiment, the historical road surface monitoring data includes historical road surface monitoring index data and historical road surface image data. The road surface monitoring index includes the stress, strain, displacement, flatness, compaction, rutting depth, friction coefficient, temperature, humidity, average monthly rainfall, average daily traffic flow, annual average daily truck traffic volume and cumulative standard axle load action times of the target road surface. The historical road surface image data is the historical road surface image, and each historical road surface image has the same number of pixels. The specific method of obtaining the road surface monitoring index data can refer to the existing technology.

[0063] Further, the historical road surface monitoring data is preprocessed, and the specific method of preprocessing can refer to the subsequent step S22. Preprocessing the historical road surface monitoring data can make the obtained data accurately reflect the road surface conditions, provide a reliable data basis for the subsequent calculation of indicator weights, and thus improve the accuracy of road surface defect monitoring and prediction.

[0064] S22, acquiring road surface monitoring index data and road surface image data in real time, and performing preprocessing to obtain accurate road surface monitoring index data and high-definition road surface image data.

[0065] Wherein, step S22 specifically includes the following steps:

[0066] S221. Clean the road surface monitoring index data to obtain accurate road surface monitoring index data.

[0067] Specifically, in this embodiment, data cleaning specifically includes missing value processing, outlier processing, duplicate value processing and data consistency processing. Through data cleaning, the obtained road surface monitoring index accurate data can accurately reflect the real-time status of the target road surface, ensuring the real-time and accuracy of the subsequent use of the road surface digital twin model to monitor the target road surface.

[0068] S222: Smoothing and noise reduction processing are performed on the road surface image data, and contrast is enhanced to obtain high-definition road surface image data.

[0069] Specifically, in this embodiment, a mean shift filtering algorithm is used to smooth and reduce noise on the road surface image data, and a CLAHE algorithm is used to enhance the contrast of the road surface image data. This can make the road surface image clearer and smoother, and the details of cracks and potholes more prominent, and the edges of cracks and potholes clearer, so that the obtained road surface high-definition image data can accurately reflect the real-time condition of the target road surface, ensuring the real-time and accuracy of the subsequent monitoring of the target road surface using the road surface digital twin model, and at the same time can also improve the accuracy of subsequent identification of road surface defects.

[0070] S3. According to the pavement digital twin model and the pavement monitoring data set, a dual-source weight fusion method is used to obtain the indicator weights of each pavement monitoring indicator.

[0071] Wherein, step S3 specifically includes the following steps:

[0072] S31. Perform image recognition on the historical road surface image data in the road surface monitoring data set to obtain road surface defects, where the road surface defects include cracks and potholes.

[0073] Specifically, in this embodiment, the YOLOv7 algorithm is used to identify cracks and potholes on the historical road surface image, and the number of pixels of the historical road surface image occupied by the road surface defects is recorded.

[0074] S32. Calculate the pavement defect evaluation factor value corresponding to each group of pavement monitoring data in the pavement monitoring data set according to the image recognition result.

[0075] Specifically, in this embodiment, the road surface defect evaluation factor value is the ratio of the number of pixels of the road surface defect to the number of pixels of the road surface image. The road surface defect evaluation factor value can intuitively reflect the severity of the road surface defect.

[0076] S33. Based on the pavement monitoring data set and the pavement defect evaluation factor value, use the permutation importance method to obtain the first weight of the pavement monitoring index.

[0077] S34. According to the pavement digital twin model, the pavement monitoring data set and the pavement defect evaluation factor value, use an item-by-item analysis method to obtain the second weight of the pavement monitoring index.

[0078] Wherein, step S34 specifically includes the following steps:

[0079] S341. Randomly select a group of road surface monitoring data from the road surface monitoring data set as the array to be adjusted, and the remaining groups of road surface monitoring data as adjustment arrays.

[0080] Specifically, in this embodiment, the array to be adjusted is recorded as , is the value of the jth road surface monitoring index in the array to be adjusted, and n is the number of road surface monitoring indicators; the adjustment array is recorded as , is the value of the jth road surface monitoring indicator in the kth adjustment array, k=1,2,3,...,N, and N is the number of adjustment arrays.

[0081] S342. For a certain road surface monitoring indicator, the value of the road surface monitoring indicator in the adjustment array is used to replace the value of the road surface monitoring indicator in the array to be adjusted in turn to obtain multiple groups of adjusted arrays, and the road surface monitoring indicator is marked as a road surface monitoring change indicator.

[0082] Specifically, in this embodiment, it is assumed that The corresponding road surface monitoring index is used as the road surface monitoring change index, and then used in turn , , ,... replace , and obtain N adjusted arrays, namely , , ,... Since the number of road surface monitoring indicators is n, n×N adjusted arrays can be obtained in sequence.

[0083] S343. Use the adjusted array to update the pavement digital twin model, identify pavement defects and calculate multiple pavement defect evaluation factor comparison values, and then calculate the second weight of the pavement monitoring change index in combination with the pavement defect evaluation factor values.

[0084] Specifically, in this embodiment, the road surface digital twin model is updated using the adjusted array to obtain the twin road surface corresponding to the adjusted array, and then the road surface image in the twin road surface is extracted and the road surface defects are identified using the YOLOv7 algorithm.

[0085] Furthermore, the number of pixels of the road surface image in the extracted twin road surface should be the same as the number of pixels of the historical road surface image.

[0086] More specifically, the second weight satisfies the following relationship:

[0087]

[0088] in, is the second weight, is the comparison value of the ith road surface defect evaluation factor, is the road surface defect evaluation factor value, is the value of the road surface monitoring change index in the i-th adjustment array, It is the value of the road surface monitoring change index in the array to be adjusted.

[0089] S35. Calculate the indicator weight according to the first weight and the second weight.

[0090] Specifically, in this embodiment, the indicator weight is the average of the first weight and the second weight. The first weight is obtained based on the machine learning method, and the second weight is obtained based on the statistical test method. The combination of the two to obtain the indicator weight can avoid the problem of one-sidedness of the weight obtained by a single method, and improve the accuracy of the road defect monitoring prediction results.

[0091] S4. Update the pavement digital twin model based on the precise data of the pavement monitoring indicators, the indicator weights and the pavement high-definition image data to achieve real-time monitoring of target pavement defects.

[0092] Wherein, step S4 specifically includes the following steps:

[0093] S41. Calculate weighted road surface monitoring index data according to the road surface monitoring index precise data and the index weight.

[0094] Specifically, in this embodiment, the weighted road surface monitoring index data is the product of the road surface monitoring index precision data and the corresponding index weight.

[0095] S42. Use the weighted road surface monitoring index data and the road surface high-definition image data to update the road surface digital twin model to obtain a real-time twin road surface, thereby realizing real-time monitoring of the target road surface defects.

[0096] Specifically, in this embodiment, by calculating weighted road surface monitoring index data, and then fully considering the influence of different road surface monitoring indicators on the road surface condition, and introducing high-definition road surface image data to realize multi-source fusion of quantitative data and image data, the road surface digital twin model can more comprehensively reflect the actual situation of the road surface, thereby improving the real-time and comprehensiveness of monitoring.

[0097] S5. Predict the precise data of the pavement monitoring indicators, and then predict the target pavement defects according to the indicator weights and the pavement digital twin model, and formulate a pavement maintenance strategy.

[0098] Wherein, step S5 specifically includes the following steps:

[0099] S51. Predict the precise data of the road surface monitoring indicators to obtain the future indicator prediction data of each road surface monitoring indicator.

[0100] Specifically, in this embodiment, a prediction model for each road surface monitoring index is constructed using a BP neural network based on the road surface monitoring data set, and the established prediction model is used to obtain index prediction data.

[0101] Furthermore, in other optional embodiments, the indicator prediction data for road surface monitoring indicators such as temperature, humidity and average monthly rainfall can also be obtained based on the weather forecast of the location of the target road surface.

[0102] S52. Calculate weighted road surface monitoring index prediction data according to the index prediction data and the index weight.

[0103] Specifically, in this embodiment, the method for calculating the weighted road surface monitoring index prediction data is the same as the method for calculating the weighted road surface monitoring index data.

[0104] S53. Use the weighted road surface monitoring index prediction data to update the real-time twin road surface to obtain the future twin road surface, thereby realizing the prediction of the target road surface defects.

[0105] Specifically, in this embodiment, the weighted road surface monitoring index prediction data is used to update the real-time twin road surface to obtain the future twin road surface, and then the future road surface image in the future twin road surface is extracted and the YOLOv7 algorithm is used to identify road surface defects to achieve the prediction of the target road surface defects.

[0106] Furthermore, the number of pixels in the future road surface image should be the same as the number of pixels in the historical road surface image.

[0107] S54: Formulate a current pavement maintenance strategy based on the weighted pavement monitoring index data.

[0108] Specifically, in this embodiment, the weighted pavement monitoring index data corresponding to each pavement monitoring index are divided into three level ranges from the lowest to the highest according to expert opinions, namely, below normal range, normal, and above normal range, and then different response plans are retrieved and formulated for different level ranges of different weighted pavement monitoring index data to form a pavement maintenance plan library, and then the corresponding response plans are queried in the pavement maintenance plan library based on the weighted pavement monitoring index data to form the current pavement maintenance strategy.

[0109] S55. Formulate a future pavement maintenance strategy based on the real-time twin pavement, the weighted pavement monitoring index prediction data and the future twin pavement, and integrate the current pavement maintenance strategy to obtain the pavement maintenance strategy.

[0110] Wherein, step S55 specifically includes the following steps:

[0111] S551. Arrange the weighted road surface monitoring index prediction data in descending order.

[0112] Specifically, in this embodiment, the weighted road surface monitoring index prediction data are arranged in descending order as a weighted road surface monitoring index prediction data sequence, and are recorded as , It is the j-th weighted road surface monitoring index prediction data in the weighted road surface monitoring index prediction data sequence.

[0113] S552. Delete the smallest data in the weighted road surface monitoring index prediction data in sequence, and use the remaining weighted road surface monitoring index prediction data to update the real-time twin road surface after each deletion to obtain a comparison twin road surface, and calculate the similarity between the comparison twin road surface and the future twin road surface.

[0114] Specifically, in this embodiment, taking the first deletion of the smallest data in the weighted road surface monitoring index prediction data as an example, the deletion ,get , then use The real-time twin road surface is updated to obtain a comparison twin road surface. Next, a comparison road surface image of the comparison twin road surface is extracted, and then the similarity between the comparison road surface image and the future road surface image extracted in step S53 is calculated, and the result of the calculation is used as the similarity between the comparison twin road surface and the future twin road surface. By sorting the weighted road surface monitoring index prediction data by size and deleting the smallest data one by one, and using the remaining weighted road surface monitoring index prediction data to update the real-time twin road surface, and then calculating the similarity with the future twin road surface, it is possible to more accurately identify the combination of road surface monitoring indicators that has the greatest impact on the future state of the road surface, which is convenient for the subsequent integration of the current road surface maintenance strategy to obtain a more targeted road surface maintenance strategy, while ensuring the maintenance effect while minimizing the maintenance cost.

[0115] Furthermore, the number of pixels of the comparison road surface image should be the same as the number of pixels of the historical road surface image. In addition, this embodiment uses a structural similarity index to indicate the similarity between the comparison road surface image and the future road surface image.

[0116] S553: ​​When the similarity is less than the similarity threshold for the first time, the operation of deleting the weighted road surface monitoring index prediction data is stopped, and the weighted road surface monitoring index prediction data remaining after the last deletion operation is used as the final judgment data.

[0117] Specifically, in this embodiment, the similarity threshold is set to 0.92.

[0118] S554. Formulate a future pavement maintenance strategy according to the final judgment data, and integrate the current pavement maintenance strategy to obtain the pavement maintenance strategy.

[0119] Specifically, in this embodiment, according to the final judgment data obtained, the future pavement maintenance strategy is formulated using the method for formulating the current pavement maintenance strategy, and then the current pavement maintenance strategy is integrated with the future pavement maintenance strategy to obtain the pavement maintenance strategy.

[0120] More specifically, the specific process of integrating the current pavement maintenance strategy with the future pavement maintenance strategy is as follows: for a certain pavement monitoring indicator, determine whether the level range of the corresponding weighted pavement monitoring indicator data is not lower than the level range of the corresponding weighted pavement monitoring indicator prediction data. If so, the response plan for the pavement monitoring indicator in the current pavement maintenance strategy will not be changed. If not, the response plan for the pavement monitoring indicator in the future pavement maintenance strategy will be used to replace the response plan for the pavement monitoring indicator in the current pavement maintenance strategy. After completing the judgment of all pavement monitoring indicators, the modified current pavement maintenance strategy, that is, the final pavement maintenance strategy, can be obtained.

[0121] When formulating pavement maintenance strategies, integrating current pavement maintenance strategies with future pavement maintenance strategies can comprehensively consider the current and future status of the pavement, achieve a combination of long-term planning and short-term response, and formulate more scientific and reasonable maintenance strategies, which will help optimize the allocation of maintenance resources, reduce unnecessary waste, and improve the efficiency and benefits of maintenance work.

[0122] It should be noted that, in some cases, the actions described in the specification can be performed in a different order and still achieve the desired results. In this embodiment, the order of steps given is only to make the embodiment appear clearer and easier to explain, rather than to limit it.

[0123] In an alternative embodiment, see Figure 2 The present invention provides a pavement defect monitoring and prediction system. The system uses a pavement defect monitoring and prediction method provided by the present invention. The system includes a digital twin model construction module A1, a data acquisition and preprocessing module A2, a pavement defect monitoring and prediction module A3, a data storage module A4 and an information visualization module A5.

[0124] The digital twin model construction module A1 is used to construct a road surface digital twin model of the target road surface.

[0125] Specifically, in this embodiment, the digital twin model construction module A1 is electrically connected to the pavement defect monitoring and prediction module A3 and the information visualization module A5, respectively. The digital twin model construction module A1 is based on SketchUp software and uses the road geometry information, road texture information and environmental information in the road design CAD drawings to construct a pavement digital twin model.

[0126] The data acquisition and preprocessing module A2 is used to collect historical road surface monitoring data to construct a historical road surface monitoring data set, and to obtain accurate road surface monitoring index data and high-definition road surface image data in real time.

[0127] Specifically, in this embodiment, the data acquisition and preprocessing module A2 specifically executes the content described in step S2.

[0128] More specifically, the data acquisition and preprocessing module A2 includes a digital display screen and various sensors for collecting road monitoring index data, which transmit information to the data acquisition and preprocessing module A2 through the Internet of Things. Data that cannot be obtained using sensors can be input into the data acquisition and preprocessing module A2 through the digital display screen.

[0129] The pavement defect monitoring and prediction module A3 is used to obtain the indicator weights of each pavement monitoring indicator based on the pavement digital twin model and the pavement monitoring data set using a dual-source weight fusion method; update the pavement digital twin model based on the precise data of the pavement monitoring indicators, the indicator weights and the pavement high-definition image data to achieve real-time monitoring of target pavement defects; predict the precise data of the pavement monitoring indicators, and then predict the target pavement defects based on the indicator weights and the pavement digital twin model, and formulate a pavement maintenance strategy.

[0130] Specifically, in this embodiment, the pavement defect monitoring and prediction module A3 is electrically connected to the digital twin model construction module A1, the data acquisition and preprocessing module A2, the data storage module A4 and the information visualization module A5 respectively. The pavement defect monitoring and prediction module A3 specifically executes the contents described in steps S3-S5, which will not be repeated here.

[0131] The data storage module A4 is used to store the data collected and generated by the data collection and preprocessing module A2, and can also store the data generated by the road surface defect monitoring and prediction module A3.

[0132] Specifically, in this embodiment, the data storage module A4 is electrically connected to the data acquisition and preprocessing module A2, the road surface defect monitoring and prediction module A3 and the information visualization module A5 respectively.

[0133] Furthermore, the data storage module A4 also stores a road maintenance solution library.

[0134] The information visualization module A5 is used to display real-time road surface monitoring data and road surface digital twin information, and can output the road surface maintenance strategy.

[0135] In summary, the method provided by the present invention presents the complex road conditions in an intuitive and easy-to-understand manner by constructing a road surface digital twin model, thereby realizing the monitoring and prediction of road surface defects. The method enables the road surface monitoring index data and road surface image data to be preprocessed so that they can accurately reflect the real-time condition of the road surface, thereby enabling the updated road surface digital twin model to accurately reflect the real-time condition of the road surface, thereby realizing accurate monitoring of the road surface. When updating the road surface digital twin model, the method calculates the weighted road surface monitoring index data by combining the precise data of the road surface monitoring index and the index weight, thereby fully considering the degree of influence of different road surface monitoring indicators on the road surface state, so that the road surface digital twin model can more comprehensively reflect the actual situation of the road surface, thereby improving the real-time, comprehensiveness and accuracy of the monitoring. Among them, the index weight is obtained based on the fusion of the first weight and the second weight, thereby avoiding the problem that the weight obtained by a single method is one-sided, and further improving the accuracy of the road surface defect monitoring prediction results. After monitoring and predicting the road surface defects, the method can also formulate corresponding road surface maintenance strategies, while ensuring the maintenance effect and reducing the maintenance cost as much as possible, which is conducive to extending the service life of the road surface and improving the safety and traffic capacity of the road. In addition, the system provided by the present invention is compatible with the method provided by the present invention, and can stably execute the method provided by the present invention, thereby improving the practical application capability of the present invention.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A road surface defect monitoring and prediction method, characterized in that: The steps include: Build a pavement digital twin model of the target pavement; Collect historical road surface monitoring data to build a historical road surface monitoring data set, and obtain accurate road surface monitoring indicator data and high-definition road surface image data in real time; According to the pavement digital twin model and the pavement monitoring data set, a dual-source weight fusion method is used to obtain the indicator weight of each pavement monitoring indicator; Based on the precise data of the road surface monitoring index, the index weights and the high-definition image data of the road surface, the road surface digital twin model is updated to achieve real-time monitoring of target road surface defects; Predict the precise data of the pavement monitoring indicators, and then predict the target pavement defects according to the indicator weights and the pavement digital twin model, and formulate a pavement maintenance strategy; The historical road surface monitoring data includes historical road surface monitoring index data and historical road surface image data; The collecting of historical road surface monitoring data to construct a historical road surface monitoring data set and obtaining accurate road surface monitoring index data and high-definition road surface image data in real time includes the following steps: Collecting historical road surface monitoring data, and preprocessing the historical road surface monitoring data, thereby constructing a historical road surface monitoring data set; Acquire road surface monitoring index data and road surface image data in real time, and perform preprocessing to obtain accurate road surface monitoring index data and high-definition road surface image data; The step of obtaining the index weights of various road surface monitoring indexes by using a dual-source weight fusion method according to the road surface digital twin model and the road surface monitoring data set comprises the following steps: Performing image recognition on the historical road surface image data in the road surface monitoring data set to obtain road surface defects, wherein the road surface defects include cracks and potholes; Calculate the pavement defect evaluation factor value corresponding to each group of pavement monitoring data in the pavement monitoring data set according to the image recognition result; Based on the pavement monitoring data set and the pavement defect evaluation factor value, using a permutation importance method to obtain a first weight of the pavement monitoring index; Obtaining a second weight of the pavement monitoring index using an item-by-item analysis method according to the pavement digital twin model, the pavement monitoring data set, and the pavement defect evaluation factor value; Calculate the indicator weight according to the first weight and the second weight; The step of obtaining the second weight of the pavement monitoring index by using an item-by-item analysis method according to the pavement digital twin model, the pavement monitoring data set, and the pavement defect evaluation factor value comprises the following steps: Randomly selecting a group of road surface monitoring data from the road surface monitoring data set as the array to be adjusted, and the remaining groups of road surface monitoring data as the adjustment arrays; For a certain road surface monitoring indicator, the value of the road surface monitoring indicator in the adjustment array is used to replace the value of the road surface monitoring indicator in the array to be adjusted in sequence to obtain multiple groups of adjusted arrays, and the road surface monitoring indicator is marked as a road surface monitoring change indicator; Using the adjusted array to update the pavement digital twin model, identifying pavement defects and calculating a plurality of pavement defect evaluation factor comparison values, and then calculating a second weight of the pavement monitoring change index in combination with the pavement defect evaluation factor values; The second weight satisfies the following relationship: ， in, is the second weight, N is the number of the adjustment arrays, is the comparison value of the ith road surface defect evaluation factor, is the road surface defect evaluation factor value, is the value of the road surface monitoring change index in the i-th adjustment array, is the value of the road surface monitoring change index in the array to be adjusted.

2. A road surface defect monitoring and prediction method according to claim 1, characterized in that: The real-time acquisition of road surface monitoring index data and road surface image data, and preprocessing to obtain the accurate road surface monitoring index data and the high-definition road surface image data includes the following steps: Performing data cleaning on the road surface monitoring indicator data to obtain accurate data of the road surface monitoring indicator; The road surface image data is smoothed and denoised, and the contrast is enhanced to obtain the road surface high-definition image data.

3. A road surface defect monitoring and prediction method according to claim 1, characterized in that: The updating of the road surface digital twin model based on the road surface monitoring index precise data, the index weight and the road surface high-definition image data to achieve real-time monitoring of target road surface defects includes the following steps: Calculate weighted road surface monitoring index data according to the road surface monitoring index accurate data and the index weight; The weighted road surface monitoring index data and the road surface high-definition image data are used to update the road surface digital twin model to obtain a real-time twin road surface, thereby realizing real-time monitoring of the target road surface defects.

4. A road surface defect monitoring and prediction method according to claim 3, characterized in that: The predicting of the accurate data of the pavement monitoring index, and then predicting the target pavement defects according to the index weights and the pavement digital twin model, and formulating a pavement maintenance strategy includes the following steps: Predict the precise data of the road surface monitoring indicators to obtain the future indicator prediction data of each road surface monitoring indicator; Calculate weighted road surface monitoring index prediction data according to the index prediction data and the index weight; Using the weighted pavement monitoring index prediction data to update the real-time twin pavement to obtain the future twin pavement, thereby realizing the prediction of the target pavement defects; Formulate a current pavement maintenance strategy based on the weighted pavement monitoring indicator data; A future pavement maintenance strategy is formulated based on the real-time twin pavement, the weighted pavement monitoring index prediction data and the future twin pavement, and the current pavement maintenance strategy is integrated to obtain the pavement maintenance strategy.

5. A road surface defect monitoring and prediction method according to claim 4, characterized in that: The step of formulating a future pavement maintenance strategy based on the real-time twin pavement, the weighted pavement monitoring index prediction data and the future twin pavement, and integrating the current pavement maintenance strategy to obtain the pavement maintenance strategy includes the following steps: Arrange the weighted road surface monitoring index prediction data in descending order; Deleting the smallest data in the weighted road surface monitoring index prediction data in sequence, and using the remaining weighted road surface monitoring index prediction data after each deletion to update the real-time twin road surface to obtain a comparison twin road surface, and calculating the similarity between the comparison twin road surface and the future twin road surface; When the similarity is less than the similarity threshold for the first time, the operation of deleting the weighted road surface monitoring index prediction data is stopped, and the weighted road surface monitoring index prediction data remaining after the last deletion operation is used as the final judgment data; A future pavement maintenance strategy is formulated according to the final judgment data, and the current pavement maintenance strategy is integrated to obtain the pavement maintenance strategy.

6. A road surface defect monitoring and prediction system, the system using a road surface defect monitoring and prediction method according to any one of claims 1 to 5, characterized in that: The system comprises: A digital twin model construction module, wherein the digital twin model construction module is used to construct a road surface digital twin model of a target road surface; A data acquisition and preprocessing module, which is used to collect historical road surface monitoring data to construct a historical road surface monitoring data set, and obtain accurate road surface monitoring indicator data and high-definition road surface image data in real time; A pavement defect monitoring and prediction module, the pavement defect monitoring and prediction module is used to obtain the indicator weights of various pavement monitoring indicators using a dual-source weight fusion method based on the pavement digital twin model and the pavement monitoring data set; update the pavement digital twin model based on the pavement monitoring indicator precise data, the indicator weights and the pavement high-definition image data to achieve real-time monitoring of target pavement defects; predict the pavement monitoring indicator precise data, and then predict the target pavement defects based on the indicator weights and the pavement digital twin model, and formulate a pavement maintenance strategy; The step of obtaining the index weights of various road surface monitoring indexes by using a dual-source weight fusion method according to the road surface digital twin model and the road surface monitoring data set comprises the following steps: Performing image recognition on the historical road surface image data in the road surface monitoring data set to obtain road surface defects, wherein the road surface defects include cracks and potholes; Calculate the pavement defect evaluation factor value corresponding to each group of pavement monitoring data in the pavement monitoring data set according to the image recognition result; Based on the pavement monitoring data set and the pavement defect evaluation factor value, using a permutation importance method to obtain a first weight of the pavement monitoring index; Obtaining a second weight of the pavement monitoring index using an item-by-item analysis method according to the pavement digital twin model, the pavement monitoring data set, and the pavement defect evaluation factor value; Calculate the indicator weight according to the first weight and the second weight; The step of obtaining the second weight of the pavement monitoring index by using an item-by-item analysis method according to the pavement digital twin model, the pavement monitoring data set, and the pavement defect evaluation factor value comprises the following steps: Randomly selecting a group of road surface monitoring data from the road surface monitoring data set as the array to be adjusted, and the remaining groups of road surface monitoring data as the adjustment arrays; For a certain road surface monitoring indicator, the value of the road surface monitoring indicator in the adjustment array is used to replace the value of the road surface monitoring indicator in the array to be adjusted in sequence to obtain multiple groups of adjusted arrays, and the road surface monitoring indicator is marked as a road surface monitoring change indicator; Using the adjusted array to update the pavement digital twin model, identifying pavement defects and calculating a plurality of pavement defect evaluation factor comparison values, and then calculating a second weight of the pavement monitoring change index in combination with the pavement defect evaluation factor values; The second weight satisfies the following relationship: ， in, is the second weight, N is the number of the adjustment arrays, is the comparison value of the ith road surface defect evaluation factor, is the road surface defect evaluation factor value, is the value of the road surface monitoring change index in the i-th adjustment array, is the value of the road surface monitoring change index in the array to be adjusted; A data storage module, the data storage module is used to store the data collected and generated by the data collection and preprocessing module, and can also store the data generated by the road surface defect monitoring and prediction module; An information visualization module, which is used to display real-time pavement monitoring data and pavement digital twin information, and can output the pavement maintenance strategy.

Images