Pavement flatness detection method based on vibration sensor

Abstract

The invention discloses a pavement flatness detection method based on a vibration sensor. The method comprises the following steps: acquiring collected vibration sensor data; the vibration sensor data comprises acceleration perpendicular to the ground direction; determining relative variation and direction according to the acceleration and the reference acceleration; judging whether the data of the vibration sensor is a peak value or not according to a preset upper threshold value, a preset lower threshold value, a preset relative variable quantity and a preset direction, and combining adjacent peak values for reporting so as to obtain a complete event; and determining the pavement flatness according to the complete event, and obtaining a flatness detection result. According to the method, the peak values in the vibration signals are detected in a mode of combining multiple groups of threshold values and statistical values, the adjacent peak values are combined into one event, and finally, the pavement flatness is judged by analyzing the event, so that repeated reporting of a single disease can be avoided, the accuracy of a detection result is improved, the requirement on hardware is low, the cost is low, and the method is suitable for popularization and application. And the method has a certain self-adaptive capability to road condition changes.

Description

Technical Field

[0001] The present application relates to the technical field of road surface smoothness detection, and in particular to a road surface smoothness detection method based on a vibration sensor. Background Art

[0002] Road surface smoothness is one of the important indicators of road quality, which directly affects the vehicle's driving stability, riding comfort and the service life of the road. High-quality road surface smoothness can not only reduce vehicle wear and fuel consumption, but also reduce the incidence of traffic accidents and improve the efficiency of the overall transportation system. Therefore, road surface smoothness detection is an important part of road inspection. By evaluating the smoothness of the road surface, road surface defects can be discovered in time to ensure road driving comfort and safety.

[0003] From the hardware level, the existing road surface roughness detection methods mainly include laser ranging, inertial measurement and visual detection. Among them, the laser ranging method uses a laser sensor to measure the height change of the road surface, with high measurement accuracy, and is mainly used for quantitative detection; the inertial measurement method uses inertial sensors such as accelerometers and gyroscopes to measure the vibration and tilt changes of the vehicle during driving, with good real-time and portability, and is widely used for road inspections; the visual detection method uses a camera to collect road surface images and uses image processing algorithms to analyze the road surface roughness, which is currently relatively less used.

[0004] From a system level perspective, existing methods mainly complete data collection and processing based on the edge, and upload the detected abnormal data to the cloud; some methods also upload the original data directly and complete all processing based on the cloud. Among them, the cloud-based method requires uploading the full amount of vibration data, and in actual applications, since the maintenance department usually needs on-site pictures or videos of bumpy sections, the full amount of image data will also be uploaded, which has extremely high requirements for the transmission network and may incur high network costs. Therefore, its application scope is much smaller than the edge-based method. In the edge-based method, spectrum analysis first needs to complete the transformation from spatial domain to frequency domain. Its computational complexity is slightly higher than that of the spatial domain method, and the transformation requires a period of data as input, and cannot directly process single-point data, so spatial domain algorithms are still needed as a supplement.

[0005] In addition, during the patrol driving process, the road conditions vary widely, and a single threshold may not be able to adapt to all patrol scenarios. Existing solutions generally lack support for dynamic data and multiple thresholds, and there is room for optimization.

[0006] In recent years, with the development of sensor technology and data processing technology, there are various cross-applications between different devices and methods. Some vibration sensors rely on laser ranging to improve accuracy, and some visual algorithms can also be used for vibration detection, showing a trend of mutual integration. Summary of the invention

[0007] To this end, the present application provides a road surface flatness detection method based on a vibration sensor to solve the problems of high hardware requirements, inaccurate detection and high cost in the prior art road surface flatness detection method.

[0008] In order to achieve the above objectives, this application provides the following technical solutions:

[0009] A road surface flatness detection method based on a vibration sensor, comprising:

[0010] Step 1: Acquire collected vibration sensor data; the vibration sensor data includes acceleration in a direction perpendicular to the ground;

[0011] Step 2: Determine the relative change amount and direction according to the acceleration and the reference acceleration;

[0012] Step 3: judging whether the vibration sensor data is a peak value according to the preset upper threshold, lower threshold, the relative change amount and the direction;

[0013] Step 4: If the vibration sensor data is a peak value, determine whether an event is currently in progress. If an event is currently in progress, add the peak value to the event, and use the peak value occurrence time as the event end time; if no event is currently in progress, create an event, and use the peak value occurrence time as the event start time and event end time;

[0014] Step 5: Determine the event duration according to the event start time and the event end time, and determine whether the event duration exceeds the preset time. If it exceeds the preset time, report the event; if it does not exceed the preset time, determine whether it exceeds the peak merging time; if it exceeds the peak merging time, report the event; if it does not exceed the peak merging time, wait for the next vibration sensor data to arrive and obtain the complete event;

[0015] Step 6: Determine the road surface flatness according to the complete event and obtain a flatness detection result.

[0016] Optionally, in step 2, the relative change calculation formula is:

[0017] a rel =|ag|

[0018] Among them, a rel represents the relative change, a represents the acceleration, and g represents the reference acceleration.

[0019] Optionally, in step 2, when the direction is determined according to the acceleration and the reference acceleration, specifically: when the acceleration is greater than the reference acceleration, the direction is 1; when the acceleration is less than or equal to the reference acceleration, the direction is 0.

[0020] Optionally, step 3 is specifically as follows: when the relative change is greater than the upper threshold, the vibration sensor data is determined to be a peak value; when the relative change is less than the lower threshold, the vibration sensor data is determined to be a normal value.

[0021] Optionally, step 3 further includes:

[0022] If the vibration sensor data is not a peak value or a normal value, calculating a standard score of the relative change according to a mean value and a standard deviation;

[0023] Whether the vibration sensor data is a peak value is determined according to the standard score and the statistical threshold.

[0024] Optionally, the complete event includes peak vibration duration, average amplitude, maximum amplitude and peak density.

[0025] Optionally, the method further includes: fusing the flatness detection result with the image-based road surface damage detection result to obtain a fusion result, and attributing the damage according to the fusion result.

[0026] Optionally, it also includes: determining a vehicle driving path according to the flatness detection result and a lane line detection algorithm.

[0027] Optionally, the vibration sensor is a laser rangefinder or an inertial measurement unit.

[0028] Optionally, when the vibration sensor is an inertial measurement unit, the z-axis angular velocity and the y-axis angular velocity are used to correct the driving state of the patrol vehicle.

[0029] Compared with the prior art, this application has at least the following beneficial effects:

[0030] 1. The present application provides a road surface roughness detection method based on a vibration sensor. The method uses the input data of the vibration sensor, adopts a combination of multiple sets of thresholds and statistical values ​​to detect the peak value in the vibration signal, and merges adjacent peak values ​​into one event. Finally, the road surface roughness is determined by analyzing the event. The method has a certain degree of adaptability to changes in road conditions, is compatible with various types of vibration sensors, and has low hardware requirements. By merging and analyzing vibration peaks, repeated reporting of a single disease can be avoided, and the accuracy of the detection results is improved. It has a small amount of calculation and high real-time performance. It can be run on vehicle-mounted edge devices with low overall cost.

[0031] 2. Fusion of vibration sensor data and image data can improve the overall effect of road inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more intuitively illustrate the prior art and the present application, several exemplary drawings are given below. It should be understood that the specific shapes and structures shown in the drawings should not generally be regarded as limiting conditions for implementing the present application; for example, those skilled in the art are capable of easily making conventional adjustments or further optimizations to the addition / reduction / attribution division, specific shapes, positional relationships, connection methods, dimensional ratios, etc. of certain units (components) based on the technical concepts and exemplary drawings disclosed in the present application.

[0033] Figure 1 A flow chart of a road surface flatness detection method based on a vibration sensor provided in Example 1 of the present application;

[0034] Figure 2 A peak detection flow chart provided in Example 1 of the present application;

[0035] Figure 3 A flowchart of adjacent peak merging is provided for Embodiment 1 of the present application;

[0036] Figure 4 This is a schematic diagram of the structure of the patrol vehicle-mounted equipment provided in Example 2 of the present application. DETAILED DESCRIPTION

[0037] The present application is further described below in detail through specific embodiments in conjunction with the accompanying drawings.

[0038] In the description of this application: unless otherwise specified, the meaning of "plurality" is two or more. The terms "first", "second", "third", etc. in this application are intended to distinguish the objects referred to, and do not have special meanings in terms of technical connotations (for example, they should not be understood as emphasizing the importance or order, etc.). Expressions such as "including", "comprising", "having", etc. also mean "not limited to" (certain units, components, materials, steps, etc.).

[0039] The terms such as "upper", "lower", "left", "right", "middle", etc. used in this application are usually used to facilitate intuitive understanding by comparing with the accompanying drawings, and are not absolute limitations on the positional relationship in the actual product. Without departing from the technical concept disclosed in this application, changes in these relative positional relationships should also be regarded as the scope of this application.

[0040] Embodiment 1

[0041] See also Figure 1 This embodiment provides a road surface flatness detection method based on a vibration sensor, comprising:

[0042] S1: Acquire the collected vibration sensor data; the vibration sensor data includes acceleration in a direction perpendicular to the ground;

[0043] Specifically, the method provided in this embodiment uses vibration sensor data as input, and the vibration sensor can be any device that can measure distance, speed and acceleration changes, including a laser rangefinder or an inertial measurement unit (IMU). The collected vibration sensor data needs to directly contain, or be able to obtain acceleration in the direction perpendicular to the ground through conversion.

[0044] S2: Determine the relative change amount and direction according to the acceleration and the reference acceleration;

[0045] Specifically, the relative change calculation formula is:

[0046] a rel =|ag| (1)

[0047] In formula (1), a rel represents the relative change, a represents the acceleration, and g represents the reference acceleration.

[0048] When the direction is determined based on the acceleration and the reference acceleration, specifically: when the acceleration is greater than the reference acceleration, the direction is 1; when the acceleration is less than or equal to the reference acceleration, the direction is 0, which can be expressed as:

[0049]

[0050] In formula (2), s represents direction.

[0051] S3: judging whether the vibration sensor data is a peak value according to the preset upper threshold, lower threshold, relative change amount and direction;

[0052] For details, please refer to Figure 2 , when the relative change is greater than the upper threshold, the vibration sensor data is judged as a peak value; when the relative change is less than the lower threshold, the vibration sensor data is judged as a normal value, which can be expressed as:

[0053]

[0054] In formula (3), t h represents the upper threshold, t l Indicates the lower threshold.

[0055] In this step, if the judgment result is "peak" or "normal", the result is returned directly; otherwise, a is calculated. rel The standard score (ZScore) is calculated as follows:

[0056]

[0057] In formula (4), z a represents the standard fraction, μ rel , σ rel Respectively represent a rel The mean and standard deviation over a period of time.

[0058] According to the standard score and statistical threshold, it is judged whether the vibration sensor data is a peak value, which is expressed by the formula:

[0059]

[0060] In formula (5), t z Indicates the statistical threshold.

[0061] During driving, when the road conditions change, μ rel and σ rel It will also be continuously updated, so that the detection method has a certain degree of adaptability and avoids large-scale false alarms caused by changes in road conditions; at the same time, the fixed threshold can limit the overall output of the method and avoid erroneous results caused by unlimited data offset.

[0062] S4: If the vibration sensor data is a peak value, determine whether an event is currently in progress. If an event is currently in progress, add the peak value to the event, and use the peak value occurrence time as the event end time; if no event is currently in progress, create an event, and use the peak value occurrence time as the event start time and event end time;

[0063] S5: Determine the event duration according to the event start time and the event end time, and judge whether the event duration exceeds the preset time. If it exceeds the preset time, report the event; if it does not exceed the preset time, judge whether it exceeds the peak merging time; if it exceeds the peak merging time, report the event; if it does not exceed the peak merging time, wait for the next vibration sensor data to arrive and obtain the complete event;

[0064] Specifically, when the inspection vehicle encounters road defects and is bumpy, a series of vibration peaks are often generated. In order to avoid multiple reports of the same defect, it is necessary to merge adjacent peaks and report them in the form of events. In this embodiment, steps S4 and S5 are for merging adjacent peaks. The merging process is as follows: Figure 3 shown.

[0065] S6: Determine the road surface flatness according to the complete event and obtain a flatness detection result.

[0066] When a complete event is formed through peak merging, all peak signals are included in the event, so information such as vibration duration, average amplitude, maximum amplitude and peak density can be statistically obtained. With this information, the flatness of the road surface can be determined, thus forming a quantitative flatness test result.

[0067] The present embodiment provides a method for detecting the smoothness of a road surface based on a vibration sensor, which also includes: fusing the smoothness detection result with the image-based road surface defect detection result to obtain a fusion result, and attributing the defect according to the fusion result. Specifically: In actual use, the smoothness detection and the image-based road surface defect detection are often performed simultaneously, so the two can be fused to improve the detection effect and perform defect attribution. For example, when defects such as potholes, speed bumps, and height differences of manhole covers that may cause vehicle bumps are detected in the image, the time when the defect is detected can be compared with the vibration detection time for correction. When no defect is detected in the image, but a vibration peak appears, there may be bumps and subsidence on the road surface.

[0068] The present embodiment provides a road surface roughness detection method based on a vibration sensor, further comprising: determining a vehicle driving path according to a roughness detection result and a lane line detection algorithm. Specifically, the roughness detection combined with the lane line detection algorithm can determine the current vehicle driving path, thereby filtering the bumps caused by the wheels running over the lane line. This method can effectively avoid false alarms when using a more sensitive sensor and threshold setting.

[0069] It should be noted that when using IMU as a vibration sensor, since the output data of IMU includes the acceleration and angular velocity of the x, y, and z axes, in addition to using the z-axis acceleration as the main input, other data can also be used to assist in determining the driving status of the patrol vehicle (i.e., operating status correction) to improve the accuracy of peak detection and reduce false alarms. For example, when the z-axis angular velocity is greater than a specific threshold, it means that the vehicle is turning; when the y-axis angular velocity is greater than a specific threshold, it means that the vehicle is accelerating or braking. At this time, the vibration sensor data can be multiplied by an additional coefficient to adapt to the vehicle status.

[0070] When installing the IMU device, the positive direction of the x-axis needs to point to the direction of travel of the patrol vehicle; the positive direction of the y-axis needs to be perpendicular to the direction of travel and point to the right side of the vehicle; the positive direction of the z-axis needs to point to the ground.

[0071] This embodiment provides a road surface flatness detection method based on a vibration sensor with the following advantages:

[0072] (1) It can detect the road surface smoothness based on vibration sensor data and has a certain degree of adaptability to changes in road conditions;

[0073] (2) Small amount of computation, high real-time performance, and can be run on vehicle-mounted edge devices;

[0074] (3) Adapt to various types of vibration sensors, with low hardware requirements and low cost;

[0075] (4) By combining and analyzing the vibration peaks, repeated reporting of a single disease can be avoided, with high accuracy;

[0076] (5) It can be integrated with image data to attribute diseases and improve detection results.

[0077] Embodiment 2

[0078] This embodiment provides a road surface flatness detection system based on a vibration sensor, including:

[0079] A vibration sensor data acquisition module, used to acquire the collected vibration sensor data; the vibration sensor data includes acceleration in a direction perpendicular to the ground;

[0080] A relative change calculation module, used to determine the relative change amount and direction according to the acceleration and the reference acceleration;

[0081] A peak value judgment module, used to judge whether the vibration sensor data is a peak value according to a preset upper threshold, a lower threshold and the relative change amount;

[0082] An event judgment module is used to judge whether an event is currently ongoing if the vibration sensor data is a peak value. If an event is currently ongoing, the peak value is added to the event and the peak value occurrence time is used as the event end time; if no event is currently ongoing, an event is created and the peak value occurrence time is used as the event start time and event end time;

[0083] The peak merging module is used to determine the event duration according to the event start time and the event end time, and to determine whether the event duration exceeds the preset time. If it exceeds the preset time, the event is reported; if it does not exceed the preset time, it is determined whether it exceeds the peak merging time; if it exceeds the peak merging time, the event is reported; if it does not exceed the peak merging time, it waits for the next vibration sensor data to arrive and obtains the complete event;

[0084] The flatness detection module is used to determine the road surface flatness according to the complete event and obtain a flatness detection result.

[0085] The system provided by this embodiment is further described below through a specific example. Figure 4As shown. The IMU is built into the edge computing platform (i.e., the road surface roughness detection system). The computing platform is installed in the back row of the car and fixed to the body to ensure that the body vibration can be transmitted to the IMU. A monocular camera is installed on the roof to capture the road ahead. Both IMU data and image data are processed by the computing platform, and the disease information is detected and stored on the built-in hard disk.

[0086] Before the inspection begins, the inspection vehicle is started and the system is left to stand for 1 minute to determine the z-axis reference acceleration and obtain the initial mean and standard deviation.

[0087] During the inspection, the system will continuously collect images and IMU data for disease detection and flatness detection. At the same time, the system will cache the image data of the last three minutes. When road diseases are detected in the image, a report will be directly generated and stored. When the flatness detection generates a complete event, the corresponding image detection results will be queried according to the start and end time of the event. If the query is successful, the two times will be aligned and the flatness detection results will be summarized in the existing report.

[0088] If no matching record is found in the image detection results, the corresponding image is found from the cached image data based on the three times of the start, end and maximum amplitude in the event, and a separate report is generated and stored on the hard disk.

[0089] After the inspection is completed, you can connect to the computing platform via USB or Wifi to export the inspection report.

[0090] For the specific definition of a road surface roughness detection system based on a vibration sensor, please refer to the above definition of a road surface roughness detection method based on a vibration sensor, which will not be repeated here.

[0091] The technical features of the above embodiments may be arbitrarily combined (as long as there is no contradiction in the combination of these technical features). To make the description concise, not all possible combinations of the technical features in the above embodiments are described; these embodiments that are not explicitly written should also be considered to be within the scope of this specification.

[0092] The above is a relatively specific and detailed description of the present application through general description and specific embodiments. It should be understood that, based on the technical concept of the present application, several conventional adjustments or further innovations can be made to these specific embodiments; however, as long as they do not deviate from the technical concept of the present application, the technical solutions obtained by these conventional adjustments or further innovations also fall within the scope of protection of the claims of the present application.

Claims

1. A road surface roughness detection method based on a vibration sensor, characterized in that: include: Step 1: Obtain the collected vibration sensor data; The vibration sensor data includes acceleration in a direction perpendicular to the ground; Step 2: Determine the relative change amount and direction according to the acceleration and the reference acceleration; Step 3: judging whether the vibration sensor data is a peak value according to the preset upper threshold, lower threshold, the relative change amount and the direction; Step 4: If the vibration sensor data is a peak value, determine whether an event is currently in progress. If an event is currently in progress, add the peak value to the event, and use the peak value occurrence time as the event end time; If there is no event currently in progress, an event is created and the peak occurrence time is used as the event start time and event end time; Step 5: Determine the event duration according to the event start time and event end time, and determine whether the event duration exceeds the preset time. If it exceeds the preset time, report the event; If the preset time is not exceeded, determine whether the peak merging time is exceeded; If the peak merging time is exceeded, an event is reported; If the peak merging time is not exceeded, wait for the next vibration sensor data to arrive and obtain the complete event; Step 6: Determine the road surface flatness according to the complete event and obtain a flatness detection result.

2. The road surface flatness detection method based on a vibration sensor according to claim 1, characterized in that: In step 2, the relative change calculation formula is: a rel =[a-g| Among them, a rel represents the relative change, a represents the acceleration, and g represents the reference acceleration.

3. The road surface flatness detection method based on a vibration sensor according to claim 1, characterized in that: In step 2, when the direction is determined according to the acceleration and the reference acceleration, specifically: when the acceleration is greater than the reference acceleration, the direction is 1; when the acceleration is less than or equal to the reference acceleration, the direction is 0.

4. The road surface flatness detection method based on a vibration sensor according to claim 3 is characterized in that: The step 3 is specifically as follows: when the relative change is greater than the upper threshold, the vibration sensor data is determined to be a peak value; when the relative change is less than the lower threshold, the vibration sensor data is determined to be a normal value.

5. The road surface flatness detection method based on a vibration sensor according to claim 4 is characterized in that: The step 3 further includes: If the vibration sensor data is not a peak value or a normal value, calculating a standard score of the relative change according to a mean value and a standard deviation; Whether the vibration sensor data is a peak value is determined according to the standard score and the statistical threshold.

6. The road surface flatness detection method based on a vibration sensor according to claim 1, characterized in that: The complete event includes the vibration duration, average amplitude, maximum amplitude and peak intensity of the peak.

7. The road surface flatness detection method based on a vibration sensor according to claim 1, characterized in that: Also includes: The flatness detection result is fused with the image-based road surface disease detection result to obtain a fusion result, and the disease is attributed according to the fusion result.

8. The road surface flatness detection method based on a vibration sensor according to claim 1, characterized in that: Also includes: The vehicle driving path is determined according to the flatness detection result and the lane line detection algorithm.

9. The road surface flatness detection method based on a vibration sensor according to claim 1, characterized in that: The vibration sensor is a laser rangefinder or an inertial measurement unit.

10. The road surface flatness detection method based on a vibration sensor according to claim 9, characterized in that: When the vibration sensor is an inertial measurement unit, the z-axis angular velocity and the y-axis angular velocity are used to correct the driving state of the patrol vehicle.

Images