Compaction quality detection method based on optimization of vibration signal interval cross-correlation

Abstract

The present invention discloses a compaction quality detection method based on the optimization of the cross-correlation of vibration signal intervals. First, the on-site compaction area is grid-divided and sampling points are arranged. The vibration signal is collected by the acceleration sensor on the roller and the compaction degree of each point is measured after compaction. Then, the area is divided, the vibration signal of the specified interval is preprocessed, and the cross-correlation of adjacent intervals is calculated. The signal regularity and the cross-correlation peak reference value are determined, the corresponding ICMV values of the intervals below the reference value are eliminated, and a linear regression model of the actual compaction degree and the ICMV value is established. The interval range is adjusted and iterated, and the model with the highest fitting accuracy is selected as the final compaction quality analysis model. The present invention does not require additional detection means, and uses the cross-correlation of the roller vibration signal to detect sections with large variability; by eliminating the variability value, the accuracy of the linear model for compaction evaluation is improved; at the same time, the present invention also takes into account the influence of the interval size on the variability of the vibration signal.

Description

Technical Field

[0001] The present invention belongs to the technical field of road engineering, and in particular relates to a compaction quality detection method based on optimization of vibration signal interval cross-correlation. Background Art

[0002] With the further development of information technology and the needs of my country's highway construction, intelligent roadbed compaction has been further developed and widely applied. Currently, intelligent compaction mainly evaluates compaction quality using various ICMVs (Intelligent Compaction Measure Values, ICMVs) extracted from roller vibration signals. Linear fitting of these ICMVs with the measured compaction degree can initially generate a compaction quality evaluation model. To accurately evaluate roadbed compaction quality using this model, it is necessary to maximize the test accuracy and data reliability of the test section. However, although the filler used in a certain range of roadbed sections is the same type, the overall roadbed soil is an inhomogeneous mixture, and the filler properties of each section will show certain differences that are uncontrollable. This difference in the properties of the roadbed soil will cause the roller vibration signal to exhibit different vibration responses. Failure to distinguish these differences will lead to large errors in the compaction quality evaluation using the estimated model, thus affecting the accuracy of the compaction quality evaluation. Therefore, it is necessary to propose a method to quickly eliminate vibration signals in intervals with large differences in filler properties to achieve raw data optimization. Summary of the Invention

[0003] The purpose of the embodiment of the present invention is to provide a compaction quality detection method based on the optimization of the interval cross-correlation of vibration signals, which measures the regularity of the roadbed soil characteristics in the overall roadbed interval through the interval cross-correlation of the roller vibration signal, and realizes the rapid elimination of signals with large differences in filler properties, thereby improving the accuracy of roadbed compaction quality detection.

[0004] To solve the above technical problems, the technical solution adopted by the present invention is a compaction quality detection method based on vibration signal interval cross-correlation optimization, comprising the following steps:

[0005] S1. Grid the compaction area on site and arrange compaction sampling points;

[0006] S2. Arrange an acceleration sensor on the vibrating wheel of the roller to roll the roadbed and collect vibration signals during the entire compaction process; after compaction is completed, use the ring knife method to measure the compaction degree of each compaction degree sampling point;

[0007] S3, dividing the compaction strip into regions;

[0008] S4. Preprocessing the vibration signal within the specified interval; after the preprocessing is completed, performing cross-correlation calculation on the vibration signals of adjacent intervals;

[0009] S5. determining the ICMV value within the interval according to the vibration signal;

[0010] S6. Determine the degree of regularity of the signals in each section over the entire roadbed;

[0011] S7. Determine a baseline control value for the peak value of the cross-correlation, identify and mark all intervals below the baseline control value based on a correlation value distribution trace diagram generated by the roller's operating trajectory, and eliminate the corresponding ICMV values; establish a linear regression relationship model between the actual compaction degree within the intervals where the ICMV values are not eliminated and the corresponding ICMV values;

[0012] S8. Repeat the process of S3 to S7, change different interval ranges, determine different linear regression relationship models, and select the relationship model with the highest fitting accuracy as the compaction quality analysis model.

[0013] Furthermore, the side length of the grid divided in S1 is 0.5-0.6 m, and the compaction sampling points are set at the intersections of the grids.

[0014] Furthermore, in S3, the length of the compacted strip is divided into intervals ranging from 2 to 4 m.

[0015] Furthermore, the process of preprocessing the vibration signal in S4 is specifically as follows: first, a window function is used to perform a truncation operation to ensure that the vibration signals extracted from each segment have the same length. After the truncation is completed, the obtained vibration signals are normalized, specifically as follows:

[0016]

[0017] Among them, X i is the value of the interval vibration signal at any time, X min is the minimum value of the interval vibration signal, X max is the maximum value of the interval vibration signal, and Normalization is the normalization result.

[0018] Furthermore, the cross-correlation calculation of the vibration signals in adjacent intervals in S4 is specifically as follows:

[0019]

[0020] Where: x i (t) is the vibration signal in this interval; x i+1 (t) is the vibration signal of the adjacent interval on the right side of this interval; x i-1(t) is the vibration signal of the adjacent interval on the left side of the interval; t represents the time variable; T is the sampling time; τ is the lag time of the signal, R i (τ) is the cross-correlation function of the adjacent interval to the right of the specified interval; R i '(τ) is the cross-correlation function of the adjacent interval to the left of the specified interval.

[0021] Furthermore, the regularity of the signal in each section in S6 over the entire roadbed is R r The specific determination method is:

[0022] R r =(R i(max) (τ)+R' i(max) (τ)) / 2 (3)

[0023] Where: R i(max) (τ) is the cross-correlation function R between the i-th interval and the interval to the right of the interval i The maximum value of (τ); R' i(max) (τ) is the cross-correlation function R between the i-th interval and the interval to the left of the interval i The maximum value of '(τ); R r is the average value of the peak value of the cross-correlation function in each interval;

[0024] Among them, the R of the first interval and the last interval r They are: R r =R 1(max) (τ), R r =R' n(max) (τ), where n is the total number of intervals.

[0025] Furthermore, the reference control value R of the cross-correlation peak l R is the peak value average value of the cross-correlation function in each interval r Take the overall average R a 0.8 times:

[0026] R l =0.8R a (4)

[0027] Where: R a is the overall average; R l is the baseline control value.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The present invention provides a quality detection method based on the optimization of the cross-correlation of the vibration signal interval of the roller. By analyzing and eliminating the vibration signals in the intervals with large differences in filler properties, the error caused by uneven soil quality is effectively reduced, thereby significantly improving the accuracy of the roadbed compaction quality evaluation. This method does not require additional complex detection equipment, simplifies the data processing process and improves work efficiency. At the same time, it takes into account the spatial variability of the roadbed compaction quality, and finds the optimal interval value based on the approximate variability range of the compaction quality (2 to 4m), ensuring the influence of the interval size on the detection accuracy. The present invention does not require additional detection means, and detects the sections with large variability from the cross-correlation of the vibration signal of the roller; after eliminating the variability value, the evaluation accuracy of the linear model for the degree of compaction is improved; at the same time, the influence of the interval size on the variability of the vibration signal is considered. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0031] Figure 1 This is a schematic diagram of the field test interval division of this embodiment;

[0032] Figure 2 is an abnormal value interval evaluation diagram of this embodiment;

[0033] Figure 3 is a linear fitting diagram of ICMV and compaction degree in this embodiment;

[0034] Figure 4 is the fitting accuracy of CMV and compaction degree in this embodiment; (a) is the original data, and (b) is the data after removing outliers;

[0035] Figure 5 is the fitting accuracy of CMV and compaction degree in this embodiment; (a) is the original data, and (b) is the data after removing outliers; DETAILED DESCRIPTION

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0037] This embodiment provides a compaction quality detection method based on optimization of the cross-correlation of vibration signal intervals. First, the roadbed is divided into regions in the test section. Each interval is set with points where the compaction degree of each point is measured using the ring knife method. The overall compaction value of the area is calculated using the average of the measured compaction degrees of each point in the area. The mutual differences in filler properties are evaluated by the cross-correlation of the roller vibration signals in each interval of the roadbed section. In theory, when filler properties are similar, rolling the roadbed under the same working conditions will produce similar vibration signals. Intervals with significant differences in vibration signals from surrounding intervals are eliminated to achieve data optimization and improve the accuracy of roadbed compaction quality detection. However, since the roadbed compaction quality has spatial variability, the selection of the interval range will affect the compaction effect to a certain extent. Therefore, based on the approximate variability range of roadbed compaction quality, i.e., 2 to 4 meters, the optimal interval value is found within this interval.

[0038] This embodiment is specifically carried out according to the following steps:

[0039] S1. Divide the on-site compaction area into grids and arrange compaction sampling points. In order to make the test value of the interval compaction as accurate as possible, the grid size is set to 0.5-0.6m. The intersection of each grid is designated as the compaction sampling point, which serves as the basic location point for measurement and analysis. Figure 1 shown.

[0040] S2. Place an acceleration sensor on the vibrating wheel of the roller and perform roadbed compaction according to the original operating conditions. Collect vibration signals throughout the compaction process. After compaction is complete, use the knife-ring method to measure the compaction degree at each sampling point. The overall compaction quality of a section is equal to the average compaction degree of each sampling point.

[0041] In some possible implementations, an HCF4000-A1 accelerometer is used to collect vibration signals during the compaction process. Depending on the specific application requirements, other accelerometers with corresponding capabilities may be used as alternatives to ensure accurate and effective acquisition of vibration characteristic data during the compaction process.

[0042] S3. Divide the compaction strip into regions. The range of variation in compaction quality typically lies between 2 and 4 meters. Selecting an interval that is too small (e.g., less than 2 meters) is of limited value for evaluating compaction quality, while selecting an interval that is too large may result in reduced detection accuracy. Therefore, this embodiment selects 2 meters as the initial analysis interval and gradually increases the interval size in increments of 0.5 meters to determine the impact of different interval sizes on the accuracy of compaction quality assessment.

[0043] S4. Process the vibration signal within the specified interval. First, use a window function to perform a truncation operation on it, ensuring that the vibration signal extracted from each segment has the same length. The key to this step is that the selected truncation segment must cover all compaction sampling points within the segment to ensure data integrity and representativeness. After truncation, the obtained vibration signal is normalized to eliminate the impact of signal amplitude differences and ensure the accuracy of subsequent analysis. Specifically:

[0044]

[0045] Subsequently, a cross-correlation calculation is performed on the vibration signals in adjacent intervals. This calculation quantifies the similarity or degree of association between two vibration signals. Specifically, the peak of the cross-correlation function reflects the highest cross-correlation between the two signals, that is, the maximum degree of match between the two signals at a given relative displacement. The cross-correlation function between the vibration signal in a given interval and the left and right intervals is calculated as follows:

[0046]

[0047] Where: x i (t) is the vibration signal in this interval; x i+1 (t) is the vibration signal of the adjacent interval on the right side of this interval; x i-1 (t) is the vibration signal of the adjacent interval on the left side of the interval; t represents the time variable (indicating that the vibration signal changes with the change of the time variable); T is the sampling time; τ is the lag time of the signal, R i (τ) is the cross-correlation function between the specified interval and the adjacent interval on the right; R i '(τ) is the cross-correlation function between the specified interval and the adjacent interval on the left.

[0048] S5. Calculate the ICMV value based on the divided intervals. The ICMV value is a traditional continuous compaction quality evaluation value. The calculation methods of various ICMV values are inconsistent. You can select a specific ICMV value for calculation as needed.

[0049] The improved compaction quality value (ICMV value) is calculated based on the interval division. The ICMV value is a traditional continuous compaction quality evaluation indicator. Its calculation method and specific type vary according to the specific application scenario and needs. In actual application, the appropriate ICMV calculation method can be selected for evaluation according to engineering requirements.

[0050] S6. Calculate the average value of the cross-correlation function peak values of each interval and the adjacent intervals before and after as the cross-correlation representative value to represent the regularity of the signal in the interval in the entire roadbed.

[0051] R r =(R i(max)(τ)+R' i(max) (τ)) / 2 (3)

[0052] Where: R i(max) (τ) is the cross-correlation function R between the i-th interval and the interval to the right of the interval i The maximum value of (τ); R' i(max) (τ) is the cross-correlation function R between the i-th interval and the interval to the left of the interval i The maximum value of '(τ); the first interval takes the peak value of the cross-correlation function between the interval and the second interval as the representative value of the interval, and the last interval takes the peak value of the cross-correlation function between the interval and the second to last interval as the representative value of the interval: R r =R 1(max) (τ), R r =R' n(max) (τ), where n is the total number of intervals.

[0053] S7, the cross-correlation representative value R of each interval in the road section r Take the overall average R a 0.8 times R l As the benchmark control value, the relevant numerical distribution trace diagram (such as Figure 2 As shown), identify and mark R r Lower than the above benchmark control value R l All intervals of:

[0054] R l =0.8R a (4)

[0055] Where: R a is the overall average; R l is the baseline control value.

[0056] These areas are considered to be areas that do not meet the compaction quality requirements, and the corresponding intelligent compaction measurement values (ICMV values) measured by the vibration signals are removed from the subsequent analysis to avoid adverse effects on the overall evaluation results due to local low-quality compaction points. Finally, the actual compaction degree in the remaining areas that meet the requirements is used to perform linear regression analysis with the corresponding ICMV values to establish a linear relationship model between the two, such as Figure 3 As shown:

[0057] Compaction value = A(ICMV) + B

[0058] This implementation evaluates the cross-correlation of roller vibration signals within intervals, eliminating intervals with significant differences in subgrade soil properties. This provides a more accurate and reliable method for evaluating compaction quality. This process improves the accuracy and reliability of the evaluation results and provides a theoretical basis and technical support for optimizing compaction operations in engineering practice.

[0059] S8. Repeat steps S3 to S6, changing different interval ranges to obtain different linear relationship models, and select the linear regression model with the highest fitting accuracy as the compaction quality analysis model.

[0060] Example

[0061] The test data for a 40m long section of the Leiyi Expressway was analyzed using single-strip compaction. The analysis intervals were 2m, 2.5m, 3.0m, 3.5m, and 4.0m. The compaction test interval was 0.5m. The ICMV used here is the most commonly used CMV value in compaction systems, as shown in the following formula:

[0062]

[0063] Where: c is the coefficient, c = 200; A2 is the acceleration amplitude corresponding to 2 times the fundamental frequency; A0 is the acceleration amplitude corresponding to the fundamental frequency; the acquired time series signal is transformed into a frequency domain signal by Fourier series transform to obtain the acceleration amplitude corresponding to the fundamental frequency and twice the fundamental frequency.

[0064] Table 1 Test results for intervals of 2m

[0065]

[0066]

[0067] Depend on Figure 4 (a) and (b) show that after removing outliers, the fitting accuracy is higher than the original linear model.

[0068] Table 2 Test results for interval 2.5m

[0069]

[0070]

[0071] Depend on Figure 5 As shown in (a) and (b), after removing outliers, the improved model has higher fitting accuracy than the original linear model. Using a 2.0m interval for the segmentation is better than using a 2.5m interval. However, due to limited test data, expanding the segmentation would result in insufficient data. These test results fully demonstrate that this invention can effectively remove outlier vibration signal segments in the roadbed area to improve model accuracy.

[0072] Each embodiment in this specification is described in a related manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiment is generally similar to the method embodiment, so the description is relatively simple. For related parts, refer to the description of the method embodiment.

[0073] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention are included in the scope of protection of the present invention.

Claims

1. A compaction quality detection method based on optimization of vibration signal interval cross-correlation, characterized in that: The following steps are involved: S1. Grid the compaction area on site and arrange compaction sampling points; S2. Arrange an acceleration sensor on the vibrating wheel of the roller to roll the roadbed and collect vibration signals during the entire compaction process; after compaction is completed, use the ring knife method to measure the compaction degree of each compaction degree sampling point; S3, dividing the compaction strip into regions; S4. Preprocessing the vibration signal within the specified interval; after the preprocessing is completed, performing cross-correlation calculation on the vibration signals of adjacent intervals; The cross-correlation calculation of the vibration signals in adjacent intervals is specifically as follows: Where: x i (t) is the vibration signal in this interval; x i+1 (t) is the vibration signal of the adjacent interval on the right side of this interval; x i-1 (t) is the vibration signal of the adjacent interval on the left side of the interval; t represents the time variable; T is the sampling time; τ is the lag time of the signal, R i (τ) is the cross-correlation function between the specified interval and the adjacent interval on the right; R i '(τ) is the cross-correlation function between the specified interval and the adjacent interval on the left; S5. Determine a compaction quality measurement value within the interval based on the vibration signal; S6. Determine the degree of regularity of the signals in each section over the entire roadbed; The regularity of the signals in each section over the entire roadbed is R r The specific determination method is: R r =(R i(max) (t)+R' i(max) (t)) / 2 (2) Where: R i(max) (τ) is the cross-correlation function R between the i-th interval and the interval to the right of the interval i The maximum value of (τ); R' i(max) (τ) is the cross-correlation function R between the i-th interval and the interval to the left of the interval i The maximum value of '(τ); R r is the average value of the peak value of the cross-correlation function in each interval; Among them, the R of the first interval and the last interval r They are: R r =R 1(max) (τ), R r =R' n(max) (τ), where n is the total number of intervals; S7. Determine a reference control value for the peak value of the cross-correlation, identify and mark all intervals below the reference control value based on a correlation value distribution trace diagram generated by the roller operation trajectory, and remove the corresponding compaction quality measurement values; and establish a linear regression relationship model between the actual compaction degree within the interval of the compaction quality measurement value not removed and the corresponding compaction quality measurement value; The reference control value R of the cross-correlation peak l R is the peak value average value of the cross-correlation function in each interval r Take the overall average R a 0.8 times: R l =0.8R a (3) Where: R a is the overall average; R l is the benchmark control value; S8. Repeat the process of S3 to S7, change different interval ranges, determine different linear regression relationship models, and select the relationship model with the highest fitting accuracy as the compaction quality analysis model.

2. The compaction quality detection method based on vibration signal interval cross-correlation optimization according to claim 1 is characterized in that: The side length of the grid divided in S1 is 0.5-0.6 m, and the compaction sampling points are set at the grid intersections.

3. The compaction quality detection method based on vibration signal interval cross-correlation optimization according to claim 1 is characterized in that: In S3, the length of the compacted strip is divided into intervals ranging from 2 to 4 m.

4. The compaction quality detection method based on vibration signal interval cross-correlation optimization according to claim 1 is characterized in that: The process of preprocessing the vibration signal within the specified interval in S4 is specifically as follows: first, a window function is used to perform a truncation operation to ensure that the vibration signals extracted from each segment have the same length. After the truncation is completed, the obtained vibration signals are normalized, specifically as follows: Among them, X i is the value of the interval vibration signal at any time, X min is the minimum value of the interval vibration signal, X max is the maximum value of the interval vibration signal, and Normalization is the normalization result.

Images