Comprehensive analysis method for compaction quality of lime treated soft soil roadbed based on intelligent compaction

Abstract

The invention provides a comprehensive analysis method for compaction quality of a lime treated soft soil roadbed based on intelligent compaction. The method comprises the following steps: S1, acquiring vibration modulus data corresponding to a plurality of detection units respectively; s2, calculating a compaction degree passing rate and a stability passing rate based on the vibration modulus data corresponding to the plurality of detection units, obtaining a uniformity passing rate by adopting a multi-scale variation structure analysis method, and calculating an abnormal aggregation index through a DBSCAN clustering algorithm; s3, determining the weight of each index, wherein the weight comprises the weight of the compaction degree passing rate, the weight of the uniformity passing rate and the weight of the stability passing rate; s4, calculating a compaction quality value based on each index and the corresponding weight; s5, the roadbed compaction quality grade is judged based on the compaction quality value, and roadbed repairing is conducted based on the roadbed compaction quality grade and the abnormal aggregation index. Compared with the prior art, the method has the advantages of improving the spatial resolution and defect identification precision of uniformity detection of the roadbed compactness and the like.

Description

Technical Field

[0001] This invention relates to the field of comprehensive compaction quality analysis technology, and in particular to a comprehensive compaction quality analysis method for lime-treated soft soil subgrade based on intelligent compaction. Background Technology

[0002] In recent years, intelligent compaction technology has emerged as a leading technology for real-time monitoring and comprehensive evaluation of subgrade compaction quality. It can assess the uniformity of subgrade compaction and guide localized supplementary compaction to improve overall uniformity. However, existing research on intelligent compaction technology is largely limited to improving the correlation between intelligent compaction indices and conventional indices or analyzing local uniformity. There is a lack of research on methods for evaluating the compaction quality of lime-treated soft soil subgrades, particularly failing to construct a multi-dimensional, multi-indicator integrated evaluation system. Specifically, it has not developed an evaluation model integrating the three elements of "compaction degree, uniformity, and stability" for lime-treated soft soil subgrades. Consequently, intelligent compaction evaluation methods struggle to comprehensively reflect the compaction quality of lime-treated soft soil subgrades.

[0003] Existing intelligent compaction quality evaluation methods often employ a strict "full compliance" standard, requiring all monitored indicators to meet preset thresholds before a pass rating is given. In actual engineering projects, situations frequently arise where compaction indicators have not fully met the standards, while uniformity has already shown good performance. Inadequate compaction of lime-treated soft soil subgrades may lead to an overall failure rating, failing to reflect changes in uniformity during the compaction process and hindering refined hierarchical management and targeted optimization of the compaction quality. Furthermore, the evaluation's defect identification accuracy is insufficient, easily leading to misjudgments and omissions due to overlooking density differences. Summary of the Invention

[0004] The purpose of this invention is to provide a comprehensive analysis method for the compaction quality of lime-treated soft soil subgrade based on intelligent compaction, which aims to improve the spatial resolution and defect identification accuracy of subgrade compaction uniformity detection. The objective of this invention can be achieved through the following technical solutions: A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, the method includes the following steps: S1. Obtain the original vibration modulus data of the roadbed and perform data preprocessing to obtain the vibration modulus data corresponding to multiple detection units respectively; S2. Based on the vibration modulus data corresponding to multiple detection units, calculate the compaction pass rate and stability pass rate, and use the multi-scale variation structure analysis method to obtain the uniformity pass rate. Calculate the abnormal clustering index using the DBSCAN clustering algorithm. S3. Determine the weight of each indicator, including the weight of the compaction pass rate, the weight of the uniformity pass rate, and the weight of the stability pass rate; S4. Calculate the compaction quality value based on each indicator and its corresponding weight; S5. Determine the subgrade compaction quality level based on the compaction quality value, and repair the subgrade based on the subgrade compaction quality level and the abnormal aggregation index.

[0005] Furthermore, the specific steps for calculating the compaction pass rate and stability pass rate based on the vibration modulus data corresponding to multiple detection units are as follows: Determine the vibration compaction target value. For the vibration modulus data of the i-th detection unit, determine whether it is not less than the vibration compaction target value. Calculate the percentage of detection units that are not less than the vibration compaction target value among all detection units as the compaction pass rate. Obtain vibration modulus data from two consecutive tests within the same detection unit, calculate the rate of change of the vibration modulus data from the two tests, determine whether the rate of change is less than or equal to a threshold, and calculate the percentage of detection units that are less than or equal to the threshold among all detection units as the stability pass rate.

[0006] Furthermore, the specific steps for obtaining the uniformity pass rate using the multi-scale variation structure analysis method are as follows: Predefined K spatial scales The k-th scale corresponds to the radius A circular window for each detection unit and each scale. Let the set of cells within the window be Calculate the local semivariance ,Will For scale Perform a fitting test and calculate the uniformity pass rate.

[0007] Furthermore, local semivariance for: ; in, Let be the local semivariance of the i-th detection unit. Let p and t be the set of cells within the window of the i-th detection unit, and p and t represent the elements in the set. and Sets Vibrational modulus values ​​corresponding to the internal elements p and t.

[0008] Furthermore, the uniformity pass rate is:

[0009] Where N represents the total number of detection units, For uniformity throughput, This represents the uniformity indicator function for the i-th detection unit, which indicates when the unit meets the uniformity requirement. =1, otherwise =0; .

[0010] Furthermore, the specific steps for calculating the outlier clustering index using the DBSCAN clustering algorithm are as follows: The set of detection units whose uniformity indicator function is 0 and whose vibration modulus data is less than the vibration compaction target value is defined as follows: ,j represents the sequence number of the detection unit whose uniformity indicator function is 0 and less than the target value of vibration compaction, for the set Using the DBSCAN clustering method, L clusters were obtained. Calculate the point density of each cluster. For each cluster point First calculate its nearest neighbor distance. Then calculate the weighted distance. Finally, the average weighted distance of all non-compliant points is calculated to obtain the abnormal clustering index, and the non-compliant points are the elements in set Q.

[0011] Furthermore, the point density of each cluster for: ; in, This represents the convex hull area of ​​each cluster.

[0012] Furthermore, weighted distance for:

[0013] in, Indicates the nearest neighbor distance. It is a tiny constant.

[0014] Furthermore, the anomalous clustering index is: ; in, This represents the abnormal clustering value of point q within a cluster. Each point q within a cluster and each detection unit where the uniformity indicator function is 0 and less than the vibration compaction target value are considered. One-to-one correspondence;

[0015] ; in, This represents the distance from q to the j-th non-compliant point. This represents the average weighted distance of all non-conforming points. Let M be the weighted distance to point q, and M represent the number of non-compliant points.

[0016] Furthermore, the specific steps for roadbed repair based on roadbed compaction quality grade and abnormal aggregation index include: The roadbed corresponding to the detection unit with a roadbed compaction quality grade of medium and an abnormal aggregation index exceeding the abnormal aggregation threshold shall be repaired; When the subgrade compaction quality grade is poor, the subgrade corresponding to all test units shall be repaired.

[0017] Compared with the prior art, the present invention has the following beneficial effects: (1) By introducing a multi-scale variation structure analysis method, the problem of local anomalies being masked by averaging in traditional global semivariogram analysis is solved. Traditional methods use a single global semivariogram function for analysis, which causes local compaction defects to be averaged in the global statistics, resulting in insufficient detection accuracy. This invention addresses the problem by using local semivariogram analysis... The formula, in a radius of Calculate local variability within the sliding window, such that The parameters can sensitively reflect different spatial scales By analyzing local variation features, and then using multi-scale fitting, small-scale compaction uneven areas that were missed by the global analysis are identified, thus significantly improving the spatial resolution of uniformity detection and the accuracy of defect identification.

[0018] (2) An innovative weighted nearest neighbor anomaly index (WNNI) calculation mechanism is proposed, overcoming the limitation of traditional nearest neighbor analysis in failing to distinguish the influence of density differences. Traditional nearest neighbor analysis only considers spatial distance. This approach ignores the density distribution characteristics of defective points, making it impossible to accurately distinguish the degree of anomalies between high-density and low-density areas. This invention addresses this issue through a density-weighted function. Nearest neighbor distance With cluster density Combining these factors makes the weighted distance... It can simultaneously reflect the degree of spatial aggregation and density distribution characteristics, and thus through The calculation yields standardized abnormal clustering values, thus enabling the accurate identification of high-risk abnormal clustering areas that truly require close attention. This avoids misjudgments and omissions caused by neglecting density differences in traditional methods, providing accurate spatial positioning guidance for targeted pressure replenishment.

[0019] (3) The synergistic effect of the two methods has enabled the transformation from "coarse-grained global detection" to "refined local diagnosis". The MSVA method identifies potential problem areas through multi-scale variation analysis, while the WNNI mechanism further quantifies the risk level of abnormal clustering. The combination of the two methods enables the compaction quality evaluation to change from the traditional binary judgment of "qualified / unqualified" to a multi-dimensional and accurate assessment based on spatial variation characteristics and clustering risk. Therefore, it can provide the construction party with a more scientific and accurate basis for quality control. Attached Figure Description

[0020] Figure 1 The flowcharts for S1 and S2 of the present invention are shown below; Figure 2 This is a detailed flowchart of S2 of the present invention; Figure 3 This is a detailed flowchart of S3 of the present invention; Figure 4 This is a flowchart of S5 of the present invention. Detailed Implementation

[0021] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0022] Figure 1 The flowcharts for S1 and S2 of the present invention are shown below; Figure 2 This is a detailed flowchart of S2 of the present invention; Figure 3 This is a detailed flowchart of S3 of the present invention; Figure 4 This is a flowchart of S5 of the present invention.

[0023] This invention relates to the field of intelligent compaction in roadbed engineering, specifically a comprehensive evaluation method for the compaction quality of lime-treated soft soil roadbeds based on intelligent compaction, and particularly to a comprehensive closed-loop system integrating data processing, quality judgment, and construction feedback. This method performs comprehensive data processing on the vibration modulus (Evib) collected in real time by the intelligent compaction system, including preprocessing, spatial allocation, and region division, to calculate the compaction degree pass rate, uniformity pass rate, and stability pass rate. Compared to existing global semivariogram fitting methods, this invention innovatively proposes a multi-scale variation structure analysis method (MSVA) based on local window sliding to capture the compaction uniformity variability of soft soil roadbeds at different spatial scales, thereby achieving multi-level identification of uniformity characteristics and automatic detection of non-uniform regions. Furthermore, based on traditional nearest neighbor spatial analysis, a density-weighted mechanism is introduced, proposing a weighted nearest neighbor anomaly index (WNNI) to measure the degree of abnormal clustering of unqualified detection units in space. To scientifically combine the three dimensions of indicators, the CRITIC method is used to determine the weight of each indicator, and a CQ compaction quality comprehensive evaluation model is proposed. Based on the calculated CQ values, a multi-level grading evaluation standard is established to achieve a comprehensive, scientific, and refined assessment of the subgrade compaction quality. Compared with existing technologies, the beneficial effects of this invention are that it introduces MSVA and WNNI algorithms, enabling the evaluation results to accurately reflect local unevenness and defect aggregation problems in subgrade compaction. Compared with traditional weighted evaluation, it improves the spatial analysis dimension and recognition resolution, helping construction units to carry out targeted optimization control. It also transforms the single, one-sided qualified / unqualified judgment in traditional intelligent compaction evaluation into a multi-level, operable quality assessment, overcoming the limitations of the traditional "fully qualified" standard in actual engineering. It can adapt to the compaction quality grading evaluation needs of intelligent compaction systems, providing a more guiding solution for intelligent compaction quality control, and improving construction efficiency, accuracy, and project quality.

[0024] The steps of this invention include: S1: Input and preprocessing of intelligent compaction data, including the raw data collected by the intelligent compaction system. E vib Data is imported into the data processing platform, where spatial matching, data cleaning, data format standardization, and the compaction area is divided into several detection units. Vibration modulus refers to the dynamic parameter reflecting the stiffness characteristics of soil, measured in real-time by the intelligent compaction system during the vibration compaction process; its unit is MPa.

[0025] S2: Calculation of the pass rate for each individual indicator, including: Compaction degree pass rate ( P comp ); Uniformity pass rate ( P uni): Introducing the "Multi-Scale Variation Structure Analysis (MSVA)" algorithm, the local semivariance is calculated within a sliding window at different spatial scales, and a spatial variation rate map of compaction uniformity is generated to identify local non-uniformity of subgrade materials; Anomaly Clustering Index (WNNI): The DBSCAN clustering algorithm is used to identify dense areas of non-compliance points, and the weighted nearest neighbor anomaly index WNNI is calculated based on point density and nearest neighbor distance to quantitatively assess the spatial clustering risk of weak compaction areas. Stability pass rate ( P stab ).

[0026] S3: Determination of weighting coefficients: The CRITIC method is used to determine the weighting coefficients for compaction degree, uniformity and stability.

[0027] S4: Calculation of compaction quality value (CQ): The CQ value is calculated using a weighted fusion model based on the pass rate and weighting coefficient of each individual indicator.

[0028] S5: Based on the CQ value, the compaction quality is graded and judged. The calculated CQ value is compared with the preset grading standard, and the subgrade compaction quality grade information is output and displayed spatially and provided with construction feedback through GIS.

[0029] In step S1, the preprocessing information for the intelligent compaction data includes: E vib Spatial matching of data with the precise location coordinates of regular detection points, identification and processing of outliers or missing data, and all E vib The data is standardized to a standard format, and the entire compaction area is divided into several detection units.

[0030] In step S2, the compaction pass rate ( P comp The calculation method includes: determining the target value corresponding to the qualified compaction degree. E vib The value represents the detection result of intelligent compaction for the i-th detection unit. ), determine whether it is not less than the target value for vibration compaction[ E vib ]. P comp It is calculated as a percentage of the total number of detection units that passed the test.

[0031] In step S2, the uniformity pass rate ( P uniThe calculation method includes: introducing the "Multiscale Variability Structure Analysis (MSVA)" method, which... E vib The data is divided into a continuous sliding window region, and the compaction area is divided into N detection units. Record the vibration modulus of each unit. E vib Value, the detection result of intelligent compaction of the i-th detection unit ( K spatial scales are predefined. The k-th scale corresponds to the radius A circular window for each detection unit and each scale After setting the cells within the window as a set, calculate the local semivariance. .Will For scale A fitting process is performed to extract variability peaks and discontinuity boundaries by fitting the degree of variation at various scales. Finally, the identification results are used to determine local heterogeneous regions. The larger the size, the more severe the inequality, and the lower the score. P uni The results are obtained by normalizing the scores of all units. The calculation and fitting formulas for the Multiscale Variation Structure Analysis (MSVA) method are as follows:

[0032]

[0033]

[0034]

[0035] In the formula: —The i-th detection unit; —The k-th scale corresponds to the radius of the circular window; —A collection of cells within a window; —Local semivariance.

[0036] Meanwhile, in the non-conforming point clustering discrimination, the WNNI calculation mechanism is introduced, specifically including: using DBSCAN to perform density clustering of non-conforming points, identifying high-incidence areas of anomalies, and based on preset thresholds (such as...). E vib M non-compliant test units were screened out (not meeting the standards), and their set is denoted as Munqualified. Applying DBSCAN to Q yields L clusters. Calculate the point density of each cluster. For each point within a cluster First calculate its nearest neighbor distance. Then define the weighted distance. Finally, the average weighted distance of all non-conforming points is calculated, and the point-level WNNI is defined. When When WNNI > 1, the clustering degree of this point is higher than the average level; if WNNI exceeds the preset threshold T, it is determined that there is significant spatial clustering, and local compression needs to be applied to the hotspot areas. The calculation formula is as follows:

[0037]

[0038]

[0039]

[0040]

[0041] In the formula: —The area of ​​the convex hull of the cluster; —Point density of each cluster; —Nearest neighbor distance; —Weighted distance; —A small constant to prevent division by zero; —The average weighted distance of all non-compliant points.

[0042] In step S2, the stability pass rate ( P stab The calculation method is as follows: calculate the change rate of the intelligent compaction detection value of the same rolling part in two consecutive passes, and determine whether the change rate is less than or equal to 3%. P stab The result is calculated as the percentage of the number of detection units that passed stability tests out of the total number of detection units.

[0043] In step S3, the weighting coefficients are determined using the CRITIC method, the specific steps of which are as follows: 1. Standardize the pass rate data for each individual indicator, transforming the original data matrix... Standardization is performed to obtain the standardized matrix. The calculation formula is:

[0044] In the formula: i—sample number; j—original value number of the index, where j=1 represents the degree of compaction, j=2 represents uniformity, and j=3 represents stability.

[0045] 2. Calculate the standard deviation of each standardized indicator. Assume there are m evaluation samples and n evaluation indicators. For each standardized column (indicator)...j ), calculate its standard deviation The calculation formula is:

[0046] 3. Using the Pearson correlation coefficient formula, calculate the correlation coefficient between any two indicators. j and k Correlation coefficient between The calculation formula is:

[0047] 4. Calculate the conflict information for each indicator. The calculation formula is:

[0048] 5. Calculate the final weights of each indicator through normalization. The calculation formula is: .

[0049] In step S4, the formula for calculating the compaction quality value (CQ) is:

[0050] In the formula: P comp - The compaction pass rate is calculated as the percentage of the area that passes the compaction pass rate out of the total compacted area. The calculation formula is:

[0051] - The uniformity pass rate is calculated as the percentage of the uniform pass area to the compacted area. The calculation formula is:

[0052] - The stability pass rate is calculated as the percentage of the stable pass area to the compacted area. The calculation formula is:

[0053] α, β, γ – are the weighting coefficients of the corresponding indicators.

[0054] In step S5, based on the survey results of the roadbed compaction quality requirements, a grading standard for roadbed compaction quality evaluation is established. The compaction quality value (CQ) is used as the standard for grade classification. Multiple grades of CQ values ​​are determined according to preset thresholds, and the corresponding compaction grade information is output for GIS display and construction strategy generation, as determined in Table 1. Table 1. Correspondence between CQ values, corresponding compaction quality grades, and engineering significance.

[0055] Based on relevant research findings both domestically and internationally, this invention proposes a method for evaluating the compaction quality of lime-treated soft soil subgrades using intelligent compaction technology. A closed-loop evaluation system consisting of three stages—data processing, quality assessment, and construction feedback—is constructed. First, the vibration modulus (…) is collected and processed. E vib The data was preprocessed, spatially interpolated, and region-divided. Based on this, three pass rate indices—compaction degree, uniformity, and stability—were calculated. Secondly, a novel multi-scale structural variation analysis (MSVA) method was introduced. A multi-scale semivariogram function was constructed using a sliding window to dynamically identify compaction uniformity characteristics and quantify local variability at different spatial scales, effectively improving sensitivity and accuracy in identifying uneven regions. Simultaneously, based on traditional nearest neighbor analysis, a density-weighted mechanism was introduced, proposing the Weighted Nearest Neighbor Anomaly Index (WNNI) to identify spatially anomalous clustering characteristics of non-compliant detection units.

[0056] To scientifically integrate the three types of evaluation indicators, this invention uses the CRITIC method to objectively assign their weights, thereby constructing a comprehensive evaluation model for CQ compaction quality. Based on this model, a multi-level grading standard is formulated to classify the compaction quality of lime-treated soft soil subgrade into four levels: "excellent, good, medium, and poor". This makes the evaluation results operable and discriminative, and can guide construction units to carry out differentiated and targeted compaction optimization control.

[0057] The beneficial effects of the intelligent roadbed compaction quality evaluation method provided by this invention are as follows: Achieving refined grading and regional assessment of compaction quality: based on E vib By analyzing the spatial distribution characteristics of the data and integrating the MSVA algorithm to perform variational structure analysis on multi-scale compaction uniformity, the system can accurately identify local compaction defect areas in soft soil subgrades. The compaction quality grade results output by the system can guide the construction team to carry out targeted recompaction operations, effectively avoiding resource waste caused by overcompaction or large-scale rework, thereby improving construction efficiency, reducing costs, and shortening the construction period. This method provides a multi-dimensional integrated quality evaluation mechanism: It breaks through the traditional "qualified / unqualified" judgment model based on a single indicator, constructing a comprehensive evaluation system integrating compaction degree, uniformity, and stability. Furthermore, it introduces the CRITIC method to determine indicator weights, making the evaluation results more scientific, representative, and stable. This integrated mechanism avoids negating the overall compaction effect due to a single indicator's failure, improving the rationality and inclusiveness of the overall quality assessment. Establish an operable CQ quality grade standard to support the closed-loop of construction feedback: Taking the comprehensive compaction quality index CQ as the core variable, construct a multi-level classification standard system, which can meet the requirements of the intelligent compaction system for quality grade division and feedback control. Each grade standard can be bound to the GIS platform to achieve spatial visualization of quality data and optimization of the recompaction path. Further, as the core link of "evaluation - prediction - control", the CQ value can be used to generate construction feedback strategies and assist in formulating personalized construction optimization plans, significantly enhancing the intelligent control ability of the compaction process.

[0058] The present invention provides a comprehensive evaluation method for the compaction quality of lime-treated soft soil subgrade based on intelligent compaction. The core lies in the analysis and processing of the collected intelligent compaction data, rather than the data collection itself. Data collection, as a prerequisite step, is completed in real time by the intelligent compaction system during the compaction process and generates E vib data, and the collection and storage are completed at the construction site, which is the basis for the subsequent data processing of the present invention. The above method not only ensures the whole-process monitoring and data collection of the compaction process in the entire construction process, but also realizes the refined evaluation of the compaction quality of the lime-treated soft soil subgrade and the accurate positioning of weak areas.

[0059] The experimental section of the present invention is 110m in length and 36m in width. The object of the base compaction quality evaluation is a soft soil road section of silty clay treated with 7% lime, with a loose paving thickness of 30cm and a target compaction degree of 96%. The corresponding Evib target value is calculated as 92.85MPa. A BW226DH-5 full-hydraulic single-drum vibratory roller is used for vibratory compaction, and the Evib index is collected by the BOMAP intelligent compaction system.

[0060] In the embodiment of the present invention, referring to the specification "Technical Conditions for Continuous Compaction Control System of Highway Subgrade Filling Engineering" (JTT1127-2017), the following three judgment criteria are adopted: 1. Compaction degree: Judging as a qualified unit with Evib i ≥92.85MPa; 2. Compaction uniformity: Judging as a qualified unit with <00,00341>; 3. Compaction stability: Judging as a qualified unit with the change rate of the same position during adjacent two passes of compaction Evib i ≤?3%;

[0061] The test section is continuously vibratory compacted four times. After each pass of compaction, the full-coverage Evib data is extracted, and the passing rates of the three indexes are calculated. The results are shown in Table 2. <,

[0062] Table 2 Index Passing Rates <,

[0063] The standard deviation is calculated using the CRITIC method based on three dimensions: compaction pass rate, uniformity pass rate, and stability pass rate. Correlation coefficient Conflicting information After normalization, the final weights of each factor are obtained. (Determined by the CRITIC method, this study used...) , , The results are shown in Table 3.

[0064] Table 3 Weights of Each Factor

[0065] Based on the weights of compaction degree, uniformity, and stability, and combined with the formula for calculating compaction quality (CQ), the CQ value for each number of compaction passes is calculated. The calculation process is as follows:

[0066]

[0067]

[0068]

[0069] Based on the survey results of the roadbed compaction quality requirements, a grading standard for roadbed compaction quality evaluation was established. The CQ value was judged in multiple levels according to the preset threshold, and the corresponding compaction level information was output for GIS display and construction strategy generation, as shown in Table 4.

[0070] Table 4. Standards for Grading Subgrade Compaction Quality Grades

[0071] Based on the quantitative evaluation of compaction quality under different rolling passes, the corresponding compaction quality grades were obtained, and the results are shown in Table 5.

[0072] Table 5. CQ values ​​and compaction quality grades corresponding to different number of compaction passes.

[0073] According to the CQ values under each compaction pass number shown in Table 5 and their corresponding quality grades, it shows a fluctuating trend of rising first and then falling. All results are in the "poor" grade (0.3 < CQ < 0.6), indicating that there are significant defects in the compaction quality of the subgrade of the test section under vibration compaction, and further optimization is required. Further analyze the tabular data from three aspects: data trend, index contribution degree, and engineering significance. The passing rate of compaction degree: gradually increases with the increase of compaction pass number, but the absolute value is extremely low (<12%), far lower than the engineering qualified threshold (usually >80%), indicating that vibration compaction has little improvement in the passing rate of compaction degree of lime-treated silty clay subgrade, and lime-treated silty clay subgrade is not suitable for excessive vibration compaction; the passing rate of uniformity: reaches the peak value (70.1%) when compacted 2 times, and then drops sharply to 54.5% (3 times) and then slightly rebounds (55.9%, 4 times), which may be related to the improvement of material distribution in the initial compaction and the crushing of aggregates caused by over-compaction in the later stage; the passing rate of stability: shows an overall upward trend, but its weight is relatively low (24.19%), and its contribution to the CQ value is relatively low (an increase of 0.029).

[0074] In addition, according to the CQ model , the contributions of each index are as follows: Uniformity dominates the overall evaluation, with a weight as high as 49.53%, and the test data shows that its highest passing rate is 70.1%, and the corresponding contribution value is about 0.347 (0.701×0.4953), which is the main reason for the relatively high overall CQ value. In contrast, the passing rate of compaction degree reaches 11.7% when vibration-compacted 4 times, and its contribution value is about 0.031 (0.117×0.2629); the stability index has a slightly smaller weight than the compaction degree, but the highest passing rate is 16.1%, and its contribution value when vibration-compacted 4 times is about 0.039 (0.161×0.2419). Therefore, the overall impact of compaction degree and stability on the CQ value can be relatively ignored.

[0075] In summary, from the contribution of each index, under the current vibration compaction process conditions, the overall compaction quality grade standard of the subgrade is poor (0.3 < CQ < 0.6). The core problem is that the passing rates in the two dimensions of compaction degree and stability are always at a relatively low level, thus reducing the overall evaluation result. In the follow-up, static pressure reinforcement means should be considered to improve the compaction effect and effectively enhance the overall quality of the project.

[0076] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative labor. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention based on the concept of the present invention through logical analysis, reasoning, or limited experiments on the basis of the prior art should be within the protection scope determined by the claims.

Claims

1. A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, characterized in that, The method includes the following steps: S1. Obtain the original vibration modulus data of the roadbed and perform data preprocessing to obtain the vibration modulus data corresponding to multiple detection units respectively; S2. Based on the vibration modulus data corresponding to multiple detection units, calculate the compaction pass rate and stability pass rate, and use the multi-scale variation structure analysis method to obtain the uniformity pass rate. Calculate the abnormal clustering index using the DBSCAN clustering algorithm. S3. Determine the weight of each indicator, including the weight of the compaction pass rate, the weight of the uniformity pass rate, and the weight of the stability pass rate; S4. Calculate the compaction quality value based on each indicator and its corresponding weight; S5. Determine the subgrade compaction quality level based on the compaction quality value, and repair the subgrade based on the subgrade compaction quality level and the abnormal aggregation index.

2. The comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction according to claim 1, characterized in that, The specific steps for calculating the compaction pass rate and stability pass rate based on the vibration modulus data corresponding to multiple detection units are as follows: Determine the vibration compaction target value. For the vibration modulus data of the i-th detection unit, determine whether it is not less than the vibration compaction target value. Calculate the percentage of detection units that are not less than the vibration compaction target value among all detection units as the compaction pass rate. Obtain vibration modulus data from two consecutive tests within the same detection unit, calculate the rate of change of the vibration modulus data from the two tests, determine whether the rate of change is less than or equal to a threshold, and calculate the percentage of detection units that are less than or equal to the threshold among all detection units as the stability pass rate.

3. The comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction as described in claim 1, characterized in that, The specific steps for obtaining the uniformity pass rate using the multi-scale variation structure analysis method are as follows: Predefined K spatial scales The k-th scale corresponds to the radius A circular window for each detection unit and each scale. Let the set of cells within the window be Calculate the local semivariance ,Will For scale Perform a fitting test and calculate the uniformity pass rate.

4. The comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction according to claim 3, characterized in that, Local semivariance for: ； in, Let be the local semivariance of the i-th detection unit. Let p and t be the set of cells within the window of the i-th detection unit, and p and t represent the elements in the set. and Sets Vibrational modulus values ​​corresponding to the internal elements p and t.

5. A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, as described in claim 4, is characterized in that... Uniformity pass rate: ； Where N represents the total number of detection units, For uniformity throughput, This represents the uniformity indicator function for the i-th detection unit, which indicates when the unit meets the uniformity requirement. =1, otherwise =0; 。 6. The comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction according to claim 5, characterized in that, The specific steps for calculating the outlier clustering index using the DBSCAN clustering algorithm are as follows: Detection units with a uniformity indicator function of 0 and vibration modulus data less than the vibration compaction target value are considered as non-conforming points. The set of non-conforming points is thus defined as follows: , j represents the serial number of the detection unit where the uniformity indicator function is 0 and less than the vibration compaction target value. This indicates that the j-th uniformity indicator function is 0 and less than the vibration compaction target value, for the set Using the DBSCAN clustering method, L clusters were obtained. Calculate the point density of each cluster. For each cluster point First calculate its nearest neighbor distance. Then calculate the weighted distance. Finally, the average weighted distance of all non-compliant points is calculated to obtain the abnormal clustering index.

7. A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, as described in claim 6, is characterized in that... Point density of each cluster for: ； in, This represents the convex hull area of ​​each cluster.

8. A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, as described in claim 7, is characterized in that... Weighted distance for: ； in, Indicates the nearest neighbor distance. It is a tiny constant.

9. A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, as described in claim 8, is characterized in that... The abnormal aggregation index is: ； in, This represents the abnormal clustering value of point q within a cluster. Each point q within a cluster and each detection unit where the uniformity indicator function is 0 and less than the vibration compaction target value are considered. One-to-one correspondence; ； ； in, This represents the distance from q to the j-th non-compliant point. This represents the average weighted distance of all non-conforming points. Let M be the weighted distance to point q, and M represent the number of non-compliant points.

10. A comprehensive analysis method for compaction quality of lime-treated soft soil subgrade based on intelligent compaction, as described in claim 9, is characterized in that... The specific steps for subgrade repair based on subgrade compaction quality grade and abnormal aggregation index include: The roadbed corresponding to the detection unit with a roadbed compaction quality grade of medium and an abnormal aggregation index exceeding the abnormal aggregation threshold shall be repaired; When the subgrade compaction quality grade is poor, the subgrade corresponding to all test units shall be repaired.

Images