Detection Methods for Pavement and Bridge Technical Conditions

Abstract

The present invention discloses a method for detecting the technical conditions of road surfaces and bridges, belonging to the field of road detection technology, which solves the problems of insufficient real-time performance, high detection cost, high implementation risk and long cycle in the existing detection of the technical conditions of road surfaces and bridges. The method includes the following steps: obtaining the positioning information and vehicle speed data of a large number of autonomous vehicles; calculating the average value and variance of the vehicle speed by an on-vehicle computing unit and screening the vehicles that meet the requirements; obtaining acceleration data, camera image data and lidar scan data respectively by an on-vehicle acceleration sensor, a camera and a lidar and storing them in an on-vehicle storage unit; analyzing the acceleration data, camera image data and lidar scan data by the on-vehicle computing unit and uploading them to a central server through an on-vehicle communication unit; and performing fusion analysis and storage on the analysis results of the acceleration data, camera image data and lidar scan data respectively by the central server.

Description

Technical Field

[0001] The present invention belongs to the technical field of road detection, in particular to a method for detecting the technical conditions of road surfaces and bridges. Background Art

[0002] According to the provisions of relevant national laws and regulations, it is necessary to regularly detect the technical conditions of road surfaces and bridges of each grade of highway to judge the health status of the highway, carry out maintenance work, and ensure the safe operation of the highway. The detection of highways is mainly divided into two parts: road surface detection and bridge detection. The dynamic characteristic detection of bridges is an important part of the technical condition detection of bridges;

[0003] At present, the road surface detection solution is a combination of daily inspections and regular inspections: daily inspections are to arrange inspection technicians to take inspection vehicles regularly (such as every day), and visually inspect to find obvious disease conditions; regular inspections are to use a road surface detection vehicle equipped with lasers, cameras, and sensors to conduct a detailed quantitative inspection of the road surface regularly (such as every year);

[0004] At present, there are two solutions for the dynamic characteristic detection of bridges: one is the manual detection method. First, close the emergency lane on the bridge deck. Technicians place multiple acceleration sensors at specific positions on the bridge deck, connect all sensors to the upper computer, and open the data acquisition software of the upper computer to collect and save data, and then analyze the dynamic characteristics of the bridge through the acceleration data; the other is the real-time monitoring method, which permanently installs sensors on the bridge and realizes real-time data collection and analysis by constructing an Internet of Things system;

[0005] In the existing road surface detection solutions, daily inspections require the investment of professional technicians and special vehicles, and the implementation costs of labor and equipment are relatively high. At the same time, daily inspections can only be mainly qualitative inspections, and the inspection refinement degree is insufficient. Regular inspections require the use of expensive road surface detection vehicles, the detection cost is relatively high, and the time interval between two inspections is relatively long, and the timeliness is insufficient;

[0006] In the existing bridge dynamic characteristic detection methods, the manual detection method requires closing traffic for a long time and purchasing several acceleration sensors, with high detection costs, high implementation risks and no real-time performance. The real-time monitoring method requires purchasing and installing sensors, communication equipment, power supply equipment, acquisition equipment, etc., with a long implementation period and high implementation costs. Summary of the Invention

[0007] Object of the Invention: To provide a method for detecting the technical conditions of road surfaces and bridges to solve the above problems existing in the prior art.

[0008] Technical Solution: A method for detecting the technical conditions of road surfaces and bridges, the method comprising the following steps:

[0009] 1) Obtain positioning information and vehicle speed data of a large number of autonomous vehicles;

[0010] 2) The vehicle-mounted computing unit calculates the average speed and variance of the vehicle speed, and filters the vehicles that meet the requirements according to the average value and variance;

[0011] 3) The vehicle-mounted acceleration sensor, camera, and lidar respectively obtain acceleration data, camera image data, and lidar scan data, and store them in the vehicle-mounted storage unit;

[0012] 4) The vehicle-mounted computing unit analyzes the acceleration data, camera image data, and lidar scan data, and uploads them to the central server through the vehicle-mounted communication unit;

[0013] 5) The central server respectively performs fusion analysis on the analysis results of the acceleration data, camera image data, and lidar scan data;

[0014] 6) The data storage unit of the central server stores the analysis results.

[0015] Preferably, the positioning information of many autonomous vehicles includes: the vehicle positioning information under the analysis of bridge dynamic characteristics and the vehicle positioning information under the analysis of road surface diseases;

[0016] Among them, the vehicle positioning information under the analysis of bridge dynamic characteristics includes: numbering a vehicle passing through a bridge within a certain period as i, i = 1, 2, 3... n, and obtaining the positioning information of vehicle i;

[0017] The vehicle positioning information under the analysis of road surface diseases includes: dividing the road surface into blocks according to lanes and lengths, denoted as itk, where t is the sequential number in the length direction, t = 1, 2, 3... n, k is the lane number, k = 1, 2, 3, 4, and i is the number of a vehicle passing through a certain road surface block within a certain period, i = 1, 2, 3... n.

[0018] Preferably, obtaining acceleration data by the vehicle-mounted acceleration sensor includes:

[0019] Judging the start and end times of the vehicle passing through the bridge through the positioning information, respectively intercepting the acceleration data of the vehicle on the bridge deck and on the road surface, and filtering the acceleration data.

[0020] Preferably, analyzing the acceleration data by the vehicle-mounted computing unit includes:

[0021] Using Fourier transform to convert the road surface acceleration data from the time domain to the frequency domain, and extracting the first three-order frequencies denoted as f1, f2, f3;

[0022] A band-stop filter is used to filter the bridge deck acceleration data, and the stopband frequency ranges are: [f1×0.95, f1×1.05], [f2×0.95, f2×1.05], [f3×0.95, f3×1.05];

[0023] The filtered data is transformed from the time domain to the frequency domain by Fourier transform, and the first three-order frequencies are extracted and denoted as

[0024] Preferably, the in-vehicle computing unit analyzes the camera image data, including:

[0025] Obtain the in-vehicle camera video stream data and identify road surface diseases based on image recognition technology.

[0026] Preferably, identifying road surface diseases based on image recognition technology includes:

[0027] Construct a road surface disease picture data set;

[0028] Train an image recognition algorithm based on the data set;

[0029] Apply the image recognition algorithm to the in-vehicle camera video.

[0030] Preferably, the in-vehicle computing unit analyzes the lidar scan data, including:

[0031] Establish a road surface block elevation point cloud through the in-vehicle lidar scan data;

[0032] Estimate the International Roughness Index (IRI) of the block according to the longitudinal elevation difference, judge whether there are ruts according to the lateral elevation difference, and judge whether there are local protrusions or depressions according to the local elevation difference.

[0033] Preferably, the central server performs a fusion analysis on the acceleration data analysis results, including:

[0034] Eliminate the larger 5% and smaller 5% of the data, and calculate the remaining vehicles The mean value of h1, h2, and h3 is obtained as the characteristic frequency of a certain section of the bridge for a certain period of time.

[0035] Preferably, the central server performs a fusion analysis on the camera image data analysis results, including:

[0036] When 85% of the vehicle data analysis results support the same conclusion, then adopt this conclusion, and the conclusion includes whether there are diseases and the type and size of the diseases.

[0037] Preferably, the central server performs a fusion analysis on the lidar scan data analysis results, including:

[0038] Quantitative calculation index: Exclude the largest 5% and the smallest 5% of the data, and calculate the mean of the remaining vehicles as the representative value of the index for this block;

[0039] Qualitative analysis index: When the analysis results of 85% of the vehicle data support the same conclusion, then adopt this conclusion.

[0040] In summary, the beneficial effects of the present invention are as follows:

[0041] 1. The real-time performance is greatly improved; for the detection of pavement technical conditions, the original frequency of quantitative detection was 1 time / year, and the present invention can quantitatively track the occurrence and development trend of pavement diseases in real time; for the detection frequency of bridge dynamic characteristics, it is about 5 years / time, and the present invention can detect the bridge dynamic characteristics in real time;

[0042] 2. The cost is greatly reduced; the existing traditional technologies require dedicated manpower, equipment, and time to complete the detection work of pavement and bridge technical conditions. Taking expressways as an example, currently, about 13,000 yuan / km needs to be invested in the detection of pavement technical conditions every year, while the present invention only needs to invest about 5,000 yuan / km every year. Currently, the average investment in the detection of bridge dynamic characteristics is about 30,000 yuan / bridge / time, and the present invention only needs to invest about 1,000 yuan / bridge / time;

[0043] 3. The overall social safety risk is reduced; on the one hand, it reduces pavement traffic control, reduces the number and scope of pavement operations, and the number of operation accidents will be significantly reduced; on the other hand, due to the improved real-time performance of the detection of road and bridge technical conditions, it helps to detect road and bridge safety hazards in a timely manner and avoid the occurrence of serious accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flowchart of the method for detecting the technical conditions of pavements and bridges provided by the present invention;

[0045] Figure 2 is a network architecture diagram of the method for detecting the technical conditions of pavements and bridges provided by the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0046] In the following description, numerous specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details. In other instances, some well-known technical features are not described to avoid confusion with the present invention.

[0047] To enable those skilled in the art to better understand the solution of the present invention, the technical solution of the present invention will be further described below in conjunction with specific embodiments.

[0048] Embodiment 1

[0049] AsFigure 1 and Figure 2 , the bridge technical condition detection method provided in this embodiment, the method includes the following steps:

[0050] 1) Obtain the positioning information and vehicle speed data of a large number of autonomous vehicles;

[0051] Among them, obtaining the positioning information of a large number of autonomous vehicles includes numbering a vehicle passing through the bridge within a certain period of time as i, i = 1, 2, 3... n, and obtaining the positioning information of vehicle i;

[0052] The speed data can be obtained through the speed sensor on the autonomous vehicle.

[0053] 2) Calculate the average value and variance of the vehicle speed by the in-vehicle computing unit, and screen the vehicles that meet the requirements according to the average value and variance;

[0054] 3) Obtain the acceleration data by the in-vehicle acceleration sensor and store it in the in-vehicle storage unit;

[0055] Among them, obtaining the acceleration data by the in-vehicle acceleration sensor includes judging the start and end times of the vehicle passing through the bridge through the positioning information, respectively intercepting the acceleration data of the vehicle on the bridge deck and on the road surface, and filtering the acceleration data.

[0056] 4) Analyze the acceleration data by the in-vehicle computing unit and upload it to the central server through the in-vehicle communication unit;

[0057] Among them, analyzing the acceleration data by the in-vehicle computing unit includes: using Fourier transform to convert the road surface acceleration data from the time domain to the frequency domain, and extracting the first three-order frequencies and recording them as f1, f2, f3; using a band-stop filter to filter the bridge deck acceleration data, and the stopband frequency ranges are respectively: [f1×0.95, f1×1.05], [f2×0.95, f2×1.05], [f3×0.95, f3×1.05]; using Fourier transform to convert the filtered data from the time domain to the frequency domain, and extracting the first three-order frequencies and recording them as

[0058] Exemplarily, if the sampling frequency is F s , the number of sampling points is N, and the frequency resolution is Then the algorithm for automatically extracting the first three orders is:

[0059] Select some data for manual analysis to determine the estimated value f'1 of the first-stage frequency;

[0060] Automatically analyze the data, perform high-pass filtering, and set the cut-off frequency to f'1 / 2;

[0061] Perform fast Fourier transform to convert the data from the time domain to the frequency domain, and obtain the frequency domain curve y = H(ωi )

[0062] Among them, ω i is the frequency value of the i-th point on the discrete frequency domain curve, and H(ω i ) is the amplitude of the i-th point.

[0063] Plot the trend graph of the frequency domain curve and find all trend points:

[0064] First, take the first point (ω1, H(ω1)) of the frequency domain curve as the starting point of the trend line, denoted as (x1, q1);

[0065] The method for calculating other points on the trend line is as follows:

[0066] The formula for calculating the moving window size is:

[0067]

[0068] where int is the rounding function;

[0069] Calculate the slope of the line connecting all amplitude points within the range of the frequency domain curve (ω1, ω w ) and the ω1 amplitude point, that is, within a moving window:

[0070]

[0071] Take the (ω i , H(ω i ) corresponding to the maximum value of all k i as the second point of the trend line, denoted as (x2, q2);

[0072] Take (x2, q2) as the starting point to find the trend point (x3, q3) of the next moving window range, and so on, to determine other trend points (x4, q4) …… (x m , q m ).

[0073] Find the characteristic frequency:

[0074] Calculate the slope of (x i , q i ) with adjacent trend points (x i-1 , q i-1 ) and (x i+1 , q i+1 ):

[0075]

[0076] If k i-1 > 0 and k i < 0, then calculate xi is the characteristic frequency, and the first three characteristic frequencies are taken as the first three natural frequencies.

[0077] 5) The central server performs fusion analysis on the analysis results of the acceleration data;

[0078] Eliminate the larger 5% and smaller 5% of the data, and calculate the remaining vehicles The mean values of h1, h2, and h3 are obtained as the characteristic frequencies of a certain period of time of the bridge.

[0079] 6) The data storage unit of the central server stores the analysis results.

[0080] The bridge technical condition detection method provided by this embodiment: Let many self-driving vehicles drive through the bridge, obtain and intercept the acceleration data of the vehicles on the bridge deck and on the road surface through acceleration sensors, analyze the acceleration data, calculate the mean values of the first three natural frequencies, and use them as the characteristic frequencies of a certain period of time of the bridge to detect the dynamic characteristics of the bridge, which can detect the dynamic characteristics of the bridge in real time, discover potential bridge safety hazards in time, avoid the occurrence of serious accidents, and there is no need to invest special manpower, equipment and time in the bridge technical condition detection work, the cost is greatly reduced, and at the same time, road traffic control can be reduced, the number and scope of road operations can be reduced, and the number of operation accidents will be significantly reduced.

[0081] Embodiment 2

[0082] As Figure 1 and Figure 2 , the road surface technical condition detection method provided by this embodiment includes the following steps:

[0083] 1) Obtain the positioning information and vehicle speed data of many self-driving vehicles;

[0084] Among them, obtaining the positioning information of many self-driving vehicles includes dividing the road surface into blocks according to lanes and lengths, denoted as itk, where t is the sequence number in the length direction, t = 1, 2, 3... n, k is the lane number, k = 1, 2, 3, 4, and i is the number of a vehicle passing through a certain road surface block during a certain period of time, i = 1, 2, 3... n;

[0085] The vehicle speed data can be obtained through the speed sensors on the self-driving vehicles.

[0086] 2) The on-vehicle computing unit calculates the mean value and variance of the vehicle speed, and filters the vehicles that meet the requirements according to the mean value and variance;

[0087] 3) The on-vehicle camera obtains the camera image data and stores it in the on-vehicle storage unit;

[0088] 4) The on-vehicle computing unit analyzes the camera image data and uploads it to the central server through the on-vehicle communication unit;

[0089] Among them, the in-vehicle computing unit analyzes the camera image data, including obtaining the in-vehicle camera video stream data and identifying road surface diseases based on image recognition technology.

[0090] Identifying road surface diseases based on image recognition technology includes:

[0091] Constructing a road surface disease picture data set;

[0092] Training an image recognition algorithm based on the data set;

[0093] Applying the image recognition algorithm to the in-vehicle camera video.

[0094] 5) The central server performs fusion analysis on the analysis results of the camera image data;

[0095] When the analysis results of 85% of the vehicle data support the same conclusion, then adopt this conclusion, and the conclusion includes whether there are diseases and the disease type and size.

[0096] 6) The data storage unit of the central server stores the analysis results.

[0097] The road surface technical condition detection method provided by this embodiment: The road surface is divided into blocks according to lanes and lengths, allowing many autonomous driving vehicles to drive in the road surface blocks, collecting image data through in-vehicle cameras, and identifying road surface diseases based on image recognition technology, which can quantitatively track the occurrence and development trend of road surface diseases in real time, timely discover potential safety hazards, avoid the occurrence of serious accidents, and there is no need to invest special manpower, equipment and time in the road surface technical condition detection work, the cost is greatly reduced, at the same time, road traffic control can be reduced, the number of road surface operations and the scope can be reduced, and the number of operation accidents will be significantly reduced.

[0098] Embodiment 3

[0099] Such as Figure 1 and Figure 2 , the road surface technical condition detection method provided by this embodiment, the method includes the following steps:

[0100] 1) Obtaining the positioning information and vehicle speed data of many autonomous driving vehicles;

[0101] Among them, obtaining the positioning information of many autonomous driving vehicles includes dividing the road surface into blocks according to lanes and lengths, denoted as itk, where t is the sequence number in the length direction, t = 1, 2, 3... n, k is the lane number, k = 1, 2, 3, 4, and i is the number of a vehicle passing through a certain road surface block during a certain period of time, i = 1, 2, 3... n;

[0102] The speed data can be obtained through the speed sensors on the autonomous driving vehicles.

[0103] 2) The vehicle-mounted computing unit calculates the average speed and variance of the vehicle speed, and filters the vehicles that meet the requirements according to the average speed and variance;

[0104] 3) The vehicle-mounted lidar obtains lidar scan data and stores it in the vehicle-mounted storage unit;

[0105] 4) The vehicle-mounted computing unit analyzes the lidar scan data and uploads it to the central server through the vehicle-mounted communication unit;

[0106] Among them, the vehicle-mounted computing unit analyzes the lidar scan data, including:

[0107] Establish a road surface block elevation point cloud through the vehicle-mounted lidar scan data;

[0108] Estimate the International Roughness Index (IRI) of the block according to the longitudinal elevation difference, judge whether there are ruts according to the lateral elevation difference, and judge whether there are local protrusions or depressions according to the local elevation difference.

[0109] 5) The central server performs a fusion analysis on the analysis results of the lidar scan data;

[0110] For the quantitative calculation index: Eliminate the larger 5% and the smaller 5% of the data, and calculate the average value of the remaining vehicles as the representative value of the block index;

[0111] For the qualitative analysis index: When the analysis results of 85% of the vehicles support the same conclusion, then adopt this conclusion.

[0112] 6) The data storage unit of the central server stores the analysis results.

[0113] The road surface technical condition detection method provided by this embodiment: Divide the road surface into blocks according to lanes and lengths, let numerous autonomous vehicles drive in the road surface blocks, and scan the road surface through the vehicle-mounted lidar to establish a road surface block elevation point cloud to detect deformation diseases of the road surface, which can quantitatively track the occurrence and development trend of deformation diseases of the road surface in real time, timely discover potential safety hazards, avoid the occurrence of serious accidents, and there is no need to invest special manpower, equipment and time in the road surface technical condition detection work, the cost is greatly reduced, and at the same time, road traffic control can be reduced, the number of road operation times and scope can be reduced, and the number of operation accidents will be significantly reduced.

[0114] The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all belong to the protection scope of the present invention.

Claims

1. A method for detecting the technical condition of road surfaces and bridges, characterized in that, The method includes the following steps: 1) Obtain positioning information and vehicle speed data of numerous autonomous vehicles; 2) Calculate the average value and variance of the vehicle speed by the in-vehicle computing unit, and screen the vehicles that meet the requirements according to the average value and variance; 3) Obtain acceleration data, camera image data, and lidar scan data respectively by the in-vehicle acceleration sensor, camera, and lidar, and store them in the in-vehicle storage unit; 4) Analyze the acceleration data, camera image data, and lidar scan data by the in-vehicle computing unit, and upload them to the central server through the in-vehicle communication unit; Among them, the analysis of the acceleration data by the in-vehicle computing unit includes: Use Fourier transform to convert the road surface acceleration data from the time domain to the frequency domain, and extract the first three-order frequencies denoted as f1, f2, and f3; Use a band-stop filter to filter the bridge deck acceleration data, and the stopband frequency ranges are respectively: [f1×0.95, f1×1.05], [f2×0.95, f2×1.05], [f3×0.95, f3×1.05]; The filtered data is transformed from the time domain to the frequency domain by Fourier transform, and the first three-order frequencies are extracted and denoted as 5) The central server performs fusion analysis on the analysis results of the acceleration data, camera image data, and lidar scan data respectively; Among them, the fusion analysis of the acceleration data analysis result by the central server includes: Eliminate the larger 5% and smaller 5% of the data, and calculate the mean of the remaining vehicles to obtain h1, h2, and h3 as the characteristic frequencies of a certain section of the bridge during a certain period; 6) The data storage unit of the central server stores the analysis results.

2. The pavement and bridge technical condition detection method according to claim 1, characterized in that The positioning information of numerous autonomous vehicles includes: vehicle positioning information for analyzing bridge dynamic characteristics and vehicle positioning information for analyzing road surface diseases; Among them, the vehicle positioning information for analyzing bridge dynamic characteristics includes: number a vehicle passing through the bridge within a certain period as i, i = 1, 2, 3... n, and obtain the positioning information of vehicle i; The vehicle positioning information for analyzing road surface diseases includes: divide the road surface into blocks according to lanes and lengths, denoted as itk, where t is the sequence number in the length direction, t = 1, 2, 3... n, k is the lane number, k = 1, 2, 3, 4, and i is the number of a vehicle passing through a certain road surface block within a certain period, i = 1, 2, 3... n.

3. The pavement and bridge technical condition detection method according to claim 2, characterized in that Obtaining acceleration data by the in-vehicle acceleration sensor includes: Judge the start and end times of the vehicle passing through the bridge through the positioning information, respectively intercept the acceleration data of the vehicle on the bridge deck and on the road surface, and filter the acceleration data.

4. The pavement and bridge technical condition detection method according to claim 1, wherein, The analysis of the camera image data by the in-vehicle computing unit includes: Obtain the video stream data of the in-vehicle camera, and identify road surface diseases based on image recognition technology.

5. The pavement and bridge technical condition detection method according to claim 4, wherein Identifying road surface diseases based on image recognition technology includes: Construct a road surface disease picture data set; Train an image recognition algorithm based on the data set; Apply the image recognition algorithm to the in-vehicle camera video.

6. The pavement and bridge technical condition detection method according to claim 1, characterized in that The analysis of the lidar scan data by the in-vehicle computing unit includes: Establish a road surface block elevation point cloud through the in-vehicle lidar scan data; Estimate the international roughness index IRI of the block according to the longitudinal elevation difference, judge whether there are ruts according to the lateral elevation difference, and judge whether there are local protrusions or depressions according to the local elevation difference.

7. The pavement and bridge technical condition detection method according to claim 1, characterized in that, The fusion analysis of the camera image data analysis result by the central server includes: When the vehicle data analysis results of 85% support the same conclusion, then adopt this conclusion, and the conclusion includes whether there are diseases and the types and sizes of diseases.

8. The pavement and bridge technical condition detection method according to claim 1, characterized in that, The central server conducts a fusion analysis on the lidar scan data analysis results, including: Quantitative calculation index: Exclude the larger 5% and the smaller 5% of the data, and calculate the mean value of the remaining vehicles as the representative value of the index of this block; Qualitative analysis index: When the vehicle data analysis results of 85% support the same conclusion, then adopt this conclusion.

Images