Novel machine-soil non-contact filling compactness determination method and application thereof

Abstract

The application discloses a new type of non-contact filling compaction state judgment method of machine soil, which comprises the following steps: monitoring the acceleration signal of the vibrating wheel of the vibrating roller through a sensor; collecting on-site photos when an industrial camera is installed at the bottom of the vibrating roller; detecting the displacement state of the filling surface in the process of the vibrating roller running by using digital image technology on the obtained photos; generating a "time-displacement" curve, and obtaining the acceleration time-history curve of multiple measuring points in the same photo; deriving the time-displacement curve in the horizontal direction of the measuring points obtained in the above steps to obtain the velocity time-history curve and the acceleration time-history curve of the filling surface; processing the horizontal acceleration time-history curves in the two directions obtained in the above steps by using band-pass filtering to process the frequency similar signals of the acceleration signal of the vibrating wheel; and as the compaction degree of the filling increases, the acceleration amplitude of the vibration wave in the propagation process attenuates more slowly with the increase of the horizontal distance.

Description

Technical Field

[0001] This invention belongs to the field of continuous compaction technology for high-speed railway subgrade. Specifically, this invention relates to a method for determining the compaction state of a novel non-contact soil-machine fill material. Background Technology

[0002] As the mileage of high-speed railway subgrades in my country increases, the amount of subgrade filling projects also rises. Therefore, efficiently and accurately determining the compaction state of high-speed railway subgrades is a pressing issue. To ensure the safety of high-speed railway subgrades, current domestic and international technical indicators are mostly based on experience, conventional testing methods using destructive testing of the fill material itself, and continuous testing methods using indirect non-destructive testing. Therefore, this invention proposes a continuous non-destructive testing method based on the properties of the soil itself.

[0003] Currently, the main methods for controlling the compaction quality of high-speed railway subgrade are as follows:

[0004] Empirical methods: 1. Compaction pass method: Determine whether the roadbed is compacted by observing the compaction conditions on site and the number of compaction passes of the vibratory roller; 2. Wheel track method: Determine whether the roadbed is compacted by observing the roller tracks on site until the tracks disappear; 3. Direct measurement method: Dry density method: Directly measure the compaction degree of the roadbed by methods such as sand filling and water filling.

[0005] Mechanical index control method: 1. K30: The compaction degree of the subgrade is represented by the plate load method K30; 2. Evd: The compaction degree of the subgrade is represented by the measurement of dynamic deformation modulus.

[0006] Continuous compaction index: The compaction index is calculated by arranging sensors on the vibratory roller and monitoring the position of the eccentric block in the vibratory roller.

[0007] The empirical methods used above to judge compaction are greatly influenced by human factors. Conventional testing indicators are destructive tests, and point-to-area control is not suitable for large-area measurement. They are tests conducted after vibration compaction. Mechanical and continuous compaction indicators, on the other hand, represent compaction degree through theoretically related indirect indicators, which are theoretical and involve correlation calculations.

[0008] In addition, such as the prior art, Chinese patent application number: CN2011104196154, publication number: CN102519965A, discloses an online detection method for roadbed compaction based on machine vision, the steps of which are: (1) setting up a visual measurement system: a compaction sampling point measurement system consists of a CCD camera, a sampling marker feature point and a marker post; (2) image processing: the CCD camera collects the compaction field image, fits the three "+" shapes of the marker post, and extracts the image coordinates of the center of the sampling marker feature point and the major and minor axis parameters of the ellipse; (3) settlement calculation: the vertical distance between the center of the sampling marker feature point and the selected reference point of the marker post is calculated, and the settlement of this compaction operation is obtained; (4) roadbed compaction calculation: the settlement is substituted into the settlement-compaction mathematical model.

[0009] Chinese Patent Application No.: CN2019109113135, Publication No.: CN111024922 A discloses a continuous detection system and method for the compaction quality of high-speed railway subgrade, comprising: a data acquisition device, a data processing device, a display device, a management device, and a central control device; the data acquisition device is installed on a road roller to collect information and send the collected information to the data processing device for processing and analysis; the data processing device is installed on the road roller to receive the information collected by the data acquisition device, process and analyze the information to obtain processed and analyzed data, and determine the subgrade compaction quality based on the processed and analyzed data; the display device is installed on the road roller to display the processed and analyzed data; the management device remotely obtains the subgrade compaction quality status in real time based on the processed and analyzed data uploaded to the management device; the central control device connects to and controls the data acquisition device, the data processing device, the display device, and the management device; the gradation identification unit includes a camera and an image recognizer, the camera is connected to the image recognizer, and the image recognizer identifies the subgrade surface image collected by the camera to obtain the gradation information of the subgrade filler particles.

[0010] Chinese Patent Application No.: 201110162528.5, Publication No.: CN102288286 A, a method for analyzing and evaluating the accuracy of gearbox measurement points using a vibration accelerometer. The invention utilizes a vibration accelerometer, with rotational speed signals as the primary control, to synchronously acquire vibration and rotational speed signals from the gearbox; extracts time-domain feature values ​​for analysis and evaluation, and plots RMS-Peak diagrams and box plots; calculates the order spectrum for analysis and evaluation; and analyzes the results using time-domain and frequency-domain analysis methods to form the accuracy analysis and evaluation results of the gearbox measurement points. This invention enables order analysis of the gearbox housing vibration signals, and through time-domain and frequency-domain feature analysis, obtains the optimal location that better characterizes various fault features, reduces measurement errors, determines the installation location of the vibration sensor, and also ensures the accuracy of fault diagnosis.

[0011] Chinese Patent Application No.: 201310343049.2, Publication No.: CN10312046 A, Intelligent and Dynamic Control System and Method for Compaction Quality of Rockfill Dams. The invention discloses an intelligent and dynamic control system for the compaction quality of rockfill dams, including a data acquisition device installed on the inner frame of the vibratory roller; a data receiving and transmitting device installed in an observation room; and a data analysis device for remote quality control. The method for controlling the compaction quality of rockfill dams using the above system involves: 1. Obtaining the CMV value corresponding to the dry density ρ that meets the design requirements through experiments; 2. Acquiring the acceleration data of each vibratory roller during on-site construction; 3. Performing FFT transformation on the acceleration data after data integration and digital filtering to obtain the fundamental wave data and second harmonic data; 4. Calculating the real-time CMV value using the formula CMV=C×A²Ω÷AΩ; 5. Comparing the real-time CMV value with the CMV value corresponding to the dry density ρ that meets the design requirements; thereby achieving intelligent and dynamic control of the compaction quality of the rockfill dam. This invention ensures both construction quality and accelerates the construction progress.

[0012] Chinese Patent Application No.: 201610223271.2, Publication No.: CN105915594 A, Real-time Monitoring Device for the Stiffness of Earth-Rock and Asphalt Concrete Dam Materials During Compaction. This invention belongs to the field of quality control in earth-rock dam construction. It provides a new real-time monitoring device for the compaction quality of earth-rock dam materials and asphalt concrete core wall dam materials, which can improve the accuracy of the compaction quality assessment of the compacted dam materials, enhance the convenience of construction quality control when the two types of dam materials coexist in asphalt core wall earth-rock dam construction, and help ensure the construction quality of high earth-rock dams. Therefore, the technical solution adopted by this invention is a real-time monitoring device for the stiffness of earth-rock and asphalt concrete dam materials during compaction, comprising six parts: a dam material stiffness monitoring module, a positioning module, a controller unit, a power supply module, a vehicle-mounted frequency-hopping network Internet Protocol (IP) communication module, and a remote frequency-hopping network IP communication module. This invention is mainly applied to the quality control of earth-rock dam construction.

[0013] The aforementioned existing technologies all rely on the vibration response characteristics of the vibratory rollers and construction machinery to determine the compaction degree and mechanical faults of the roadbed, and are indirect methods for identifying these issues. They do not involve continuous monitoring of the materials being assessed. In contrast, this invention can directly determine the compaction quality of the roadbed by continuously and in real-time monitoring the vibration response curves of particles on the roadbed surface. Summary of the Invention

[0014] The purpose of this invention is to determine the compaction state of fill material by using an external device installed on a vibratory roller to monitor the displacement of local vibration areas on the ground.

[0015] The above-mentioned objective of the present invention is achieved through the following technical solution:

[0016] A novel non-contact method for determining the compaction state of soil-fill materials includes the following steps:

[0017] Step 1: Install the triaxial acceleration sensor on the frame of the front wheel of the vibratory roller to monitor the acceleration signal of the vibratory roller wheel in real time;

[0018] Step 2: Install an industrial camera in the two red areas at the bottom of the vibratory roller. Install an anti-shake system between the vibratory roller and the camera to prevent blurry photos due to instability of the vibratory roller during construction. At the same time, configure a high-power LED light source to ensure that the camera has a sufficient shooting environment.

[0019] Step 3: Use digital image technology to detect the displacement state of the filler surface during the movement of the vibratory roller using the photos obtained in step (2);

[0020] Step 4: Generate a "time-displacement" curve, and use it to obtain the acceleration time history curves of multiple measurement points in the same photograph;

[0021] Step 5: Differentiate the horizontal time-displacement curves of the measuring points obtained in Step 4 above to obtain the velocity time history curves and acceleration time history curves of the packing surface;

[0022] Step 6: The horizontal acceleration time history curves obtained above in the two directions are processed by using bandpass filtering to obtain signals with frequencies similar to the acceleration signal of the vibrating wheel;

[0023] Step 7: As the compaction degree of the packing increases, the acceleration amplitude of the vibration wave during propagation decays more slowly with the increase of horizontal distance. Therefore, since the amplitude of the vibrating wheel is fixed during operation, the compaction degree of the packing can be determined by the acceleration amplitude obtained in Step 6 above.

[0024] This invention also discloses a novel non-contact method for judging the compaction state of soil-fill material, which is applied to the construction process of high-speed railway subgrade.

[0025] Beneficial effects:

[0026] The method for evaluating the compaction state of fill material during vibratory roller construction, produced using this invention, achieves the following:

[0027] (1) Detecting vibration signals on the roadbed surface without damaging the roadbed;

[0028] (2) Real-time roadbed vibration signal detection during compaction construction;

[0029] (3) The detection of the compacted area changes with the position of the inspection vehicle, without affecting the normal construction progress.

[0030] The production method proposed in this invention is simple to operate, highly feasible, and a highly efficient real-time detection method for continuous compaction. This invention completes the detection of compaction quality during the construction phase and adjusts construction conditions in areas with poor compaction, greatly improving the quality inspection efficiency of high-speed railway subgrade construction and resulting in significant economic and social benefits. Attached Figure Description

[0031] Figure 1 This is a signal from the vibrating wheel;

[0032] Figure 2 This indicates the installation location for the vibratory roller and camera.

[0033] Figure 3 This is a geometric diagram of the reference point before and after deformation;

[0034] Figure 4(a) is the time history curve of the acceleration at the measuring point; Figure 4 (b) is the acceleration time history curve of measuring point two;

[0035] Figure 5 (a) is the time history curve of acceleration at the measurement point after filtering; Figure 5 (b) is the time history curve of acceleration at measurement point two after filtering;

[0036] Figure 6 (a) is a reference image; Figure 6 (b) is the image after deformation. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to specific embodiments. The embodiments given are only for illustrating the present invention and are not intended to limit the scope of the present invention.

[0038] This invention provides a novel non-contact method for determining the compaction state of soil-fill material, comprising the following steps:

[0039] Step 1: Install the triaxial acceleration sensor on the frame of the front wheel of the vibratory roller to monitor the acceleration signal of the vibratory roller's vibratory wheel in real time;

[0040] The acceleration time history signal collected by the acceleration sensor on the vibratory roller serves two purposes: 1. Real-time detection of the vibration amplitude in the vertical direction of the vibratory roller, such as... Figure 1 The acceleration amplitude is approximately 4g. This serves as the source of the vibration, providing a basis for calculating the amplitude of the acceleration time history curve obtained through image processing techniques and for determining the acceleration amplitude attenuation rate. If a mechanical failure occurs in the vibratory roller, causing a change in the amplitude of the vibration source, but without real-time monitoring, this will affect subsequent calculation results. 2. The acceleration time history curve can be used to calculate the number of complete cycles of the vibration wave per second to obtain the operating frequency of the vibratory roller, providing a basis for bandpass filtering of the subsequently fitted velocity time history curve.

[0041] Step 2: Install an industrial camera in the two marked areas on the bottom of the vibratory roller. Install an anti-shake system between the vibratory roller and the camera to prevent blurry photos due to instability of the vibratory roller during construction. At the same time, configure a high-power LED light source to ensure that the camera has a sufficient shooting environment.

[0042] Step 3: Use digital image processing technology to detect the displacement state of the filler surface during the movement of the vibratory roller using the photos obtained in Step 2; the principle is as follows, see below. Figure 3As shown: The digital image technology employs a deformation measurement method based on digital image processing. It calculates the displacement components of each point by accurately tracking the coordinate changes of each point of interest within the calculation area of ​​the "reference image" in the "deformed image." To accurately and reliably track a desired displacement, a square reference image sub-region of size (2M+1) pixels × (2M+1) pixels (where M is the local coordinate size of each data point in the local displacement field) is selected centered on that point (when the sub-region is large enough, it can be uniquely and accurately identified). The displacement components of that point are determined by finding the target image sub-region in the deformed image that has the highest similarity to this sub-region.

[0043] To evaluate the similarity between reference and target image sub-regions, this invention uses the zero-mean normalized sum of squared difference criterion (ZNSSD) as the correlation function to evaluate the similarity between image sub-regions before and after deformation. The purpose is to determine whether the observed target point within the area captured by the camera is the same point before and after deformation. The process mainly consists of three stages: 1. The difference between the grayscale value of the target judgment point and the reference area. The smaller the difference, the more uniform the grayscale value is within the reference area. ,in To obtain the average gray value of the reference image sub-region, the gray values ​​of each pixel are squared, summed, and divided by the number of pixels. The difference in gray values ​​between the deformed image regions is calculated similarly. 2. Normalization is performed on the reference image region by dividing the difference in gray values ​​from the first stage by (2M+1) pixels in the reference image sub-region. 2 The difference between the grayscale value of each pixel and the average value of a sub-region of the reference image. The specific formula is: The reference area for the deformed image can be obtained similarly. 3. Divide each of the above pixels into (2M+1) 2 Each pixel in the image undergoes a first-step difference calculation and normalization process. The difference between the pixel value before and after image deformation is calculated by summing the values ​​before and after deformation. The total difference between each pixel and the mean gray value in the image is then obtained. The similarity Cznssd(P) between the total difference before and after deformation is calculated by subtracting the total interpolation of the image before deformation. The specific calculation formula is shown in formula (1). If the image is undeformed, theoretically, the photos taken by the two cameras are completely identical, then Cznssd(P) = 0. Therefore, this principle can be used to determine whether the regions before and after deformation captured by the two cameras are the same region by checking whether Cznssd(P) approaches zero.

[0044] (1)

[0045] Here, P is a parameter vector describing the deformation state of a sub-region of the image. and Reference images and deformed image grayscale value, , These are the average grayscale values ​​of the reference image sub-region and the target image sub-region, respectively. This allows us to obtain the x and y (horizontal) displacement of the roadbed surface. (The instantaneous displacement value at the same point is determined by the absolute difference between the x and y coordinates of the two images).

[0046] Step 4: In Step 2, the camera uses a high sampling frequency. Using the first photo as the original image from Step 3 above, and the second through last photos as the distorted images, see the appendix. Figure 6 This allows us to obtain the displacement values ​​in the x and y directions with the coordinates (x, y) of the first photograph as the zero point. Based on these results, we can obtain the (x, y) "time-displacement" curve, and from this, we can obtain the acceleration time history curves of multiple measurement points in the same photograph.

[0047] Step 5: Differentiate the horizontal displacement time history curves of the measuring points obtained in Step 4 to obtain the velocity and acceleration time history curves of the filler surface. The displacement time history curves of the subgrade surface particles are obtained by calculating the displacement difference between each photo and the first photo in Step 4. Differentiating the displacement time history curves composed of data points at different times yields the velocity and acceleration time history curves of the subgrade filler particles. For example, if the displacement difference between the second and first photos is 0.01m, the displacement difference between the third and first photos is 0.02m, and so on, then the coordinate points are (0, 0), (0.001, 0.01), 0.002, 0.02, etc., where the horizontal axis represents time and the vertical axis represents the magnitude of displacement.

[0048] See Figure 4 As shown, Figure 4 In this embodiment, the acceleration time history curves of the two cameras are obtained through image recognition and step 5.

[0049] Step 6: The horizontal acceleration time history curves obtained in the two directions are processed using bandpass filtering to match the frequency of the vibration wheel acceleration signal. Since bandpass filtering is a mature technology, the process involves performing a Fourier transform on the acceleration time history curves acquired in Step 1 to obtain a spectrum, and determining the frequency corresponding to the maximum acceleration amplitude, which is the vibration frequency of the vibration wheel. In this embodiment, it is 30Hz. Therefore, the lower limit of the bandpass filter cutoff frequency in this step is 20Hz, and the upper limit is 40Hz.

[0050] (7) As the compaction degree of the packing increases, the acceleration amplitude of the vibration wave during propagation decays more slowly with the increase of horizontal distance. Therefore, since the amplitude of the vibrating wheel is fixed during operation, the compaction degree of the packing can be determined by the acceleration amplitude obtained in step 6 above: the maximum amplitude of a single period of the acceleration time history curve obtained in step 6 is extracted, for example... Figure 4 In (a), the acceleration amplitude of the first cycle is 0.55g, while the amplitude of the acceleration time history curve of the vibrating wheel is 4g. Therefore, the attenuation rate of the acceleration amplitude can be calculated to be 87%. If the attenuation rate of the acceleration amplitude is calculated for another region and found to be 80%, then the compaction degree of this region is greater than that of the region with the aforementioned attenuation rate of 87%. The relative compaction degree value can be obtained using this method.

[0051] By employing industrial camera image recognition and signal processing technologies such as filtering, relative compaction degree can be determined without affecting the normal operation of the road roller. The technologies mentioned in the background all involve direct contact between the equipment and the roadbed, which has a certain impact on both the compaction degree measuring equipment and the road surface. This invention obtains relative compaction degree in a non-contact manner to evaluate the compaction degree of the roadbed, reducing the failure probability of the detection equipment and the impact on the roadbed surface.

[0052] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A novel non-contact method for determining the compaction state of soil-fill materials, comprising the following steps: Step 1: Install the triaxial acceleration sensor on the frame of the front wheel of the vibratory roller to monitor the acceleration signal of the vibratory roller wheel in real time; Step 2: Install an industrial camera in the two marked areas on the bottom of the vibratory roller. Install an anti-shake system between the vibratory roller and the camera to prevent blurry photos due to instability of the vibratory roller during construction. At the same time, configure a high-power LED light source to ensure that the camera has a sufficient shooting environment. Step 3: Use digital image technology to detect the displacement state of the filler surface during the movement of the vibratory roller using the photos obtained in step (2); The digital image technology employs a deformation measurement method based on digital image processing. It calculates the displacement components of each point of interest within the calculation area of ​​the "reference image" by accurately tracking the coordinate changes of each point in the "deformed image". The zero-mean normalized least square distance correlation function is used as the correlation function to evaluate the similarity between sub-regions of the images before and after deformation. The Zero-mean normalized sum of squaredifference criterion (ZNSSD) mainly consists of three stages:

1. The difference between the gray values ​​of the target judgment point and the reference area. The smaller the difference, the more uniform the gray values ​​are within the reference area. ,in To obtain the average gray value of the reference image sub-region by squaring the gray value of each pixel, summing the results, and dividing by the number of pixels, the difference in gray values ​​of the deformed image region is calculated in the same way; 2. The reference image region is normalized by dividing the difference in gray values ​​from the first stage by (2M+1) pixels in the reference image sub-region. 2 The difference between the gray value of each pixel and the average value of a sub-region of the reference image is calculated using the following formula: The reference area for the deformed image can be obtained similarly; 3. Divide each of the above pixels into (2M+1) 2 Each pixel in the image undergoes a first step of difference calculation and normalization. The difference between each pixel and the mean gray value in the image is obtained by summing the values ​​before and after image deformation. The similarity Cznssd(P) between the total difference between the images before and after deformation is obtained by subtracting the total difference between the images after deformation. The specific calculation formula is as follows: If the image is undistorted, theoretically the photos taken by the two cameras are completely identical, then Cznssd(P) = 0. Therefore, this principle can be used to determine whether the areas before and after distortion captured by the two cameras are the same area by whether Cznssd(P) approaches zero. Step 4: Generate a "time-displacement" curve, and use it to obtain the acceleration time history curves of multiple measurement points in the same photograph; Step 5: Differentiate the horizontal time-displacement curves of the measuring points obtained in Step 4 above to obtain the velocity time history curves and acceleration time history curves of the packing surface; Step 6: The horizontal acceleration time history curves obtained above in the two directions are processed by using bandpass filtering to obtain signals with frequencies similar to the acceleration signal of the vibrating wheel; Step 7: As the compaction degree of the packing increases, the acceleration amplitude of the vibration wave during propagation decays more slowly with the increase of horizontal distance. Therefore, since the amplitude of the vibrating wheel is fixed during operation, the compaction degree of the packing can be determined by the acceleration amplitude obtained in Step 6 above.

2. The novel non-contact method for determining the compaction state of soil-fill material according to claim 1, characterized in that: Step 1 further includes the following: The acceleration time history curve signal collected by the triaxial acceleration sensor on the vibratory wheel has two functions: (1) to detect the vibration amplitude in the vertical direction of the vibratory roller in real time; (2) to obtain the working frequency of the vibratory roller by calculating the number of whole cycles of the vibration wave within the range of one second, and to provide a basis for bandpass filtering of the speed time history curve obtained by subsequent fitting.

3. The novel non-contact method for determining the compaction state of soil-fill material according to claim 1, characterized in that: Step 4 further includes the following: using the first photo as the original image and the second to last photos as the deformed images in step 3 above, the displacement values ​​in the x and y directions with the (x,y) coordinates of the first photo as the zero point are obtained, and the (x,y) "time-displacement" curve is obtained from the results, and the acceleration time history curve of multiple measurement points in the same photo is obtained from these results.

4. The novel non-contact method for determining the compaction state of soil-fill material according to claim 1, characterized in that: Step 5 further includes the following: by using the displacement difference between each photo in step 4 and the first photo, the displacement time history curve of the subgrade surface particles is obtained, and the velocity time history curve and acceleration time history curve of the filler surface are obtained by differentiating the acceleration time history curve composed of data points at different times.

Images