Search

CN-122020597-A - Highway road surface condition detection system and method based on artificial intelligence

CN122020597ACN 122020597 ACN122020597 ACN 122020597ACN-122020597-A

Abstract

The invention discloses an artificial intelligence-based expressway road surface condition detection system and a method thereof, relating to the technical field of intelligent traffic, wherein the system comprises a data acquisition module for processing and storing multidimensional road surface data; the invention relates to a method for optimizing the road surface of a highway, which comprises the steps of generating a disease trend report based on a cleaned data training model, making a maintenance scheme by a maintenance decision module, executing a feedback module to implement maintenance and collect effect data, forming a continuously optimized maintenance decision chain by a closed loop optimization module according to feedback adjustment algorithm parameters, accurately predicting the disease evolution trend of each road section by integrating road surface multidimensional data and dynamic training algorithm, scientifically dividing maintenance priority, matching a differentiated maintenance scheme, combining a full-flow closed loop control mechanism to continuously optimize algorithm performance, realizing reasonable distribution of maintenance resources, improving timeliness and suitability of disease treatment, prolonging the service life of the road surface, reducing maintenance cost and guaranteeing safe and stable operation of the highway.

Inventors

  • LIU MINGHUA
  • ZHU BING
  • ZHANG ZHEN

Assignees

  • 西南交通大学

Dates

Publication Date
20260512
Application Date
20260224

Claims (10)

  1. 1. Highway road surface condition detecting system based on artificial intelligence, characterized in that, this system includes: the data acquisition module is used for acquiring initial disease density, environmental factors, environmental factor duration time and traffic load data of each road section of the expressway, cleaning, completing and standardizing the acquired original data, and storing the acquired original data in an associated mode according to the ID and the time stamp of the road section; The evolution prediction module adopts a disease potential energy evolution algorithm based on the standardized data processed by the data acquisition module, carries out algorithm training according to road sections, wherein each road section training data set comprises historical data of not less than 3 years, adopts a sliding window mechanism to update the training data, generates a disease evolution trend report of each road section for 1-12 months in the future, and outputs a disease evolution dynamic potential energy result; The maintenance decision module is used for dividing maintenance value gain priority intervals by adopting a maintenance gain decision algorithm based on the trend report output by the evolution prediction module, and generating a specific maintenance scheme comprising maintenance time, process and operation range by corresponding different maintenance response timelines in different intervals; The execution feedback module organizes on-site maintenance operation according to the maintenance scheme formulated by the maintenance decision module, and acquires disease residue rate, road surface flatness and passing recovery time data according to a fixed period after maintenance is completed; And the closed-loop optimization module is used for receiving the feedback data transmitted by the execution feedback module, comparing and analyzing deviation, adjusting parameters of the disease potential energy evolution algorithm and the maintenance gain decision algorithm, and transmitting the adjusted parameters back to the evolution prediction module and the maintenance decision module after the data confirmation is verified by fixed time length.
  2. 2. The expressway road surface condition detection system based on artificial intelligence according to claim 1, wherein the data acquisition module is characterized in that initial disease density is disease number and disease area ratio in unit area of each road section of an expressway, vehicle-mounted laser radar and a high-definition inspection camera are adopted to acquire acquired environmental factors including environmental temperature, relative humidity, rainfall intensity, snowfall, sunshine duration and freeze thawing times, the duration of the environmental factors is duration of continuous occurrence of each environmental factor, a temperature and humidity sensor, a rain gauge, a sunshine sensor and a freeze thawing detector which are distributed along the expressway at fixed intervals are adopted to acquire the environmental factors in real time, start time and end time of each environmental factor are synchronously recorded, the duration of the environmental factors is determined, traffic load data includes the number of average heavy vehicles, the duty ratio of heavy vehicles, the axle load of average vehicles and the traffic frequency of vehicles, and the duration of the traffic load data is acquired by adopting expressway entrance and road section middle bayonet equipment.
  3. 3. The system for detecting the condition of the highway pavement based on artificial intelligence according to claim 1, wherein in the evolution prediction module, a mathematical expression of the disease potential energy evolution algorithm is as follows: Wherein, the Evolution of dynamic potential energy for diseases; the initial disease density is processed by the data acquisition module; Is an environmental coupling coefficient; Accumulating an impact value for the environment; representing the total number of categories of environmental factors; is the first The influence weight coefficient of the seed environmental factor on disease evolution; for a duration corresponding to the environmental factor; the traffic load amplification factor is used; Nonlinear effect of average equivalent axle load; the average daily equivalent axle load processed by the data acquisition module; Is the nonlinear coefficient of the load.
  4. 4. The expressway road surface condition detection system based on artificial intelligence according to claim 1 is characterized in that in the evolution prediction module, algorithm training is carried out according to road sections, wherein continuous and homogeneous training road sections are divided according to expressway topography, road surface structure types and design load levels, a training data set of each road section comprises historical data of not less than 3 years, specifically standardized historical data of initial disease density, environmental factors, duration and traffic load of the corresponding road section, and contemporaneous disease evolution dynamic potential energy historical observation values, for each divided road section, environmental coupling coefficients, traffic load amplification coefficients and load nonlinear coefficients of a disease potential energy algorithm are independently initialized, historical standardized data of the road section are input into an algorithm, coefficient values are iteratively adjusted, a sliding window mechanism with fixed duration is adopted, historical data of the earliest 1 month in a window is moved out, new month data are added, an updated training data set is formed, and the algorithm of the road section is retrained.
  5. 5. The expressway road surface condition detection system based on artificial intelligence according to claim 1, wherein the evolution prediction module is characterized in that the disease evolution trend report comprises the following main contents of disease evolution dynamic potential energy monthly change conditions of each road section calculated based on a disease potential energy evolution algorithm, contribution duty ratio of environment accumulation influence and traffic load acting in disease evolution of each period, and the historical evolution data of the corresponding road section for the past 3 years are combined, nonlinear change characteristics of the disease evolution of the current period are compared and analyzed, and trend content of each road section in the report comprises detailed historical data comparison information related to evolution characteristic description.
  6. 6. The system for detecting the condition of the highway pavement based on artificial intelligence according to claim 1, wherein in the maintenance decision module, a mathematical expression of a maintenance gain decision algorithm is as follows: Wherein, the Gain value for maintenance value; Is a disease risk conversion coefficient; evolution of dynamic potential energy for diseases; The time interval between the time point of maintenance decision and the time point of planning to implement maintenance; the benefit attenuation rate is obtained; The cost is repaired as a benchmark; Is the cost upscaling factor.
  7. 7. The expressway road surface condition detection system based on artificial intelligence according to claim 1, wherein the maintenance decision module is characterized in that the maintenance value gain priority interval is divided into a three-level priority interval by taking a maintenance value gain value MVG output by a maintenance gain decision algorithm as a judgment index and combining with a reference maintenance cost C, wherein the maintenance value gain value is judged to be a high-priority maintenance road section when the maintenance value gain value meets MVG not less than 0.8C, the maintenance value gain value is judged to be a medium-priority maintenance road section when the maintenance value gain value meets MVG not more than 0.3C and not more than 0.8C, and the maintenance value gain value is judged to be a low-priority maintenance road section when the MVG is not more than 0.3.
  8. 8. The expressway road surface condition detection system based on artificial intelligence according to claim 1, wherein the maintenance decision module is characterized in that a maintenance scheme is generated based on a maintenance value gain priority interval, and the main content comprises the steps of adopting a night half-width sealing operation mode, particularly a hot-mix asphalt rapid repair process, wherein the operation range covers a disease area and a peripheral 1-2m transition zone, setting temporary warning marks and diversion facilities synchronously, adopting a weekend local sealing operation mode, particularly a Saturday, a weekday 9:00-17:00 sealing 1-2 lanes, adopting cold patch asphalt or a micro-meter process, adopting a working range to cover a disease point accurately, cleaning a scene after operation, and recovering the working in time, adopting a maintenance scheme to start regular inspection in 30 days, carrying out 1 disease development situation tracking each month, adopting a maintenance scheme to cover a disease crack, adopting an emergency sealing scheme, and a full-scale sealing process, and providing a full-scale sealing performance standard for each maintenance scheme for a sealing layer in a medium priority road section within 72 hours after the triggering of the medium priority threshold.
  9. 9. The highway pavement condition detection system based on artificial intelligence according to claim 1, wherein the fixed duration verification data comprise continuously collecting disease development data, maintenance execution effect data, environment influence data and maintenance benefit data in at least one complete disease growth period after algorithm parameter adjustment is completed, wherein the disease development data comprise real-time state data of diseases, development rate data of diseases, coverage data of diseases and severity data of diseases, the maintenance execution effect data comprise execution state data of maintenance measures and implementation effect data of maintenance measures, the environment influence data comprise environment temperature and humidity data, environment precipitation data and environment illumination data, and the maintenance benefit data comprise maintenance cost data and maintenance efficiency data.
  10. 10. An artificial intelligence based highway pavement condition detection method for preparing the artificial intelligence based highway pavement condition detection system according to any one of claims 1 to 9, characterized by comprising the specific steps of: s100, data acquisition and preprocessing, namely acquiring initial disease density, environmental factors, environmental factor duration time and traffic load data of each road section, cleaning, complementing and standardizing the original data, and storing the original data in an associated mode according to the road section ID and the time stamp; S200, disease evolution prediction, namely training each road section by adopting a disease potential energy evolution algorithm based on the processed standardized data, wherein a training data set contains at least 3 calendar history data, updating the training data by a sliding window mechanism, generating a disease evolution trend report of each road section for 1-12 months in the future, and outputting a disease evolution dynamic potential energy result; S300, generating a maintenance scheme, namely dividing a maintenance price value gain priority interval by adopting a maintenance gain decision algorithm based on a trend report, and generating a concrete scheme containing maintenance time, process and operation range corresponding to different maintenance response timelines; s400, performing on-site maintenance, namely performing on-site operation according to a maintenance scheme, and collecting disease residual rate, road surface flatness and passing recovery time data according to a fixed period after the maintenance is finished; And S500, closed-loop optimization adjustment, namely receiving feedback data of the step S400, comparing analysis deviation, adjusting parameters of a disease potential energy evolution algorithm and a maintenance gain decision algorithm, confirming the adjusted parameters through fixed time length verification data, and returning the adjusted algorithm to the steps S200 and S300 for subsequent disease evolution prediction and maintenance decision making.

Description

Highway road surface condition detection system and method based on artificial intelligence Technical Field The invention relates to the technical field of intelligent traffic, in particular to an artificial intelligence-based expressway road surface condition detection system and an artificial intelligence-based expressway road surface condition detection method. Background The expressway is used as a core hub of a transportation system, bears key tasks of transregional passenger and cargo transportation, road surface conditions directly influence traffic safety, transportation efficiency and road network operation stability, along with economic and social development, traffic flow is continuously increased, traffic proportion of heavy vehicles is continuously increased, load pressure borne by the road surface is increasingly increased, complex changes of natural environment including comprehensive actions of various factors such as temperature fluctuation, precipitation and sunshine are accelerated, generation and evolution of road surface diseases are accelerated, a management mode which is dependent on manual inspection or single detection means is difficult to adapt to operation and maintenance requirements of a modern road network, rapid iteration of technologies such as manual intelligence and big data analysis provides technical support for integrating multidimensional detection data and realizing accurate prediction of diseases and scientific maintenance decision, and a set of intelligent system which can penetrate data acquisition, analysis prediction, maintenance execution and optimization adjustment is urgently needed in industry, so that the expressway road surface management is promoted to be transferred from passive repair to active prevention and experience driving to data driving, and challenges of the increasingly complex road network maintenance are met. The traditional highway pavement condition detection and maintenance technology has a plurality of limitations, the current requirement of fine management is difficult to meet, in a data acquisition link, a traditional mode is often limited to single type data acquisition, data dimension is incomplete, a pretreatment flow of a system is lacked, so that data quality is uneven, reliable support cannot be provided for subsequent analysis, in the aspect of disease prediction, the traditional method is lacked in differential consideration of road section characteristics, a unified prediction mode is adopted, historical data and dynamic change factors are not fully combined, the deviation between a prediction result and an actual disease development situation is larger, the disease evolution trend is difficult to predict in advance, in a maintenance decision link, the traditional mode depends on a manual experience formulation scheme, scientific priority division basis is lacked, maintenance opportunity, process selection and operation range setting are lacked pertinence, the problem that resource waste is caused easily to occur due to excessive maintenance or the fact that the maintenance is not timely aggravated is easily caused, in addition, the traditional technology is lacked an effective feedback optimization mechanism, the maintenance execution effect cannot be timely in reverse-feeding detection and decision link, the technical scheme is difficult to continuously adapt to the dynamic change of the pavement condition, and the overall management efficiency is low. Disclosure of Invention The invention aims to make up the defects of the prior art, provides an expressway road surface condition detection system and a method thereof based on artificial intelligence, the system comprises five modules of data acquisition, evolution prediction, maintenance decision, execution feedback and closed loop optimization, and the evolution prediction module is utilized to generate a disease evolution trend report by acquiring road surface diseases, environmental factors and traffic load data, the maintenance decision module divides maintenance priority and formulates a specific scheme according to the maintenance priority, the execution feedback module is responsible for field operation and effect evaluation, and the closed loop optimization module adjusts algorithm parameters according to feedback. In order to solve the technical problems, the invention provides the following technical scheme that on one hand, the system for detecting the pavement condition of the expressway based on artificial intelligence comprises: the data acquisition module is used for acquiring initial disease density, environmental factors, environmental factor duration time and traffic load data of each road section of the expressway, cleaning, completing and standardizing the acquired original data, and storing the acquired original data in an associated mode according to the ID and the time stamp of the road section; The evolution prediction module adopts a disease potential