Hydraulic rammer digital quality control system based on cloud platform

Abstract

The application relates to the technical field of industrial automatic control, in particular to a hydraulic rammer digital quality management and control system based on a cloud platform, which comprises a first determination module configured to acquire time sequence pressure values of a hydraulic cylinder of a rammer in a hitting cycle to determine an effective ramming energy coefficient of a target space grid of a current work area; a second determination module configured to determine a rebound damping coefficient of the target space grid; a construction module configured to construct a state feature sequence of the target space grid by using the effective ramming energy coefficient and the rebound damping coefficient of the target space grid; and a control module configured to control ramming work of the rammer in the current work area by using the state feature sequence. Through the technical scheme, the ramming work process of the rammer can be digitally controlled, and the ramming work efficiency and the ramming effect on a land plot are considered.

Description

Technical Field

[0001] This application relates to the field of industrial automatic control technology, and in particular to a digital quality control system for hydraulic compactors based on a cloud platform. Background Technology

[0002] In engineering construction such as roadbed compaction, hydraulic compactors are widely used to improve the density of deep soil and the bearing capacity of foundations; related technologies usually rely on manual on-site sampling and testing or on the subjective experience of engineering operators to judge the degree of foundation compaction.

[0003] Traditional quality assessment products and physical testing methods often use sand cone or ring sampler methods to obtain soil compaction data at specific sampling points. Traditional sampling and testing tools can reflect the construction quality level of local areas to a certain extent and play a basic data support role in the final project completion and acceptance stage.

[0004] However, the methods of manual physical sampling or subjective experience judgment have the problems of extremely limited detection space and serious data feedback lag; physical sampling inspection cannot cover the entire vast work area, resulting in undetected blind spots with safety hazards such as substandard compaction indicators, and experience judgment based on observation or operational feel is easily interfered with by human factors.

[0005] During the construction process of hydraulic compactors, related technologies are prone to causing damage to the original stress structure of the soil due to local over-compaction, or uneven settlement of the foundation after project delivery due to local under-compaction. This not only increases the rework costs and the risk of overall project delays, but also affects the overall structural safety of the roadbed and the long-term service life of highway bridges. Therefore, it is necessary to implement digital quality control of compactors to balance the efficiency and compaction effect of compaction operations on the land. Summary of the Invention

[0006] To achieve digital quality control of hydraulic compactors, this application provides a cloud-based digital quality control system for hydraulic compactors, comprising: a first determining module configured to acquire the temporal pressure value of the hydraulic cylinder of the compactor during the impact cycle, and the temporal acceleration value of the compactor's hammer during the impact cycle, and to extract the pressure and acceleration values ​​within a target time window from the hammer's contact with the ground to its rebound and departure from the ground, in order to determine the effective compaction energy coefficient of the target spatial grid falling into the current working area within the target time window; and a second determining module configured to acquire the soil rebound after impact with a predetermined target in the target spatial grid. The system includes a displacement timeline of vertical displacement during recovery, which is used to determine the rebound damping coefficient of the target spatial grid; a construction module configured to use the effective compaction energy coefficient and rebound damping coefficient of the target spatial grid to determine the state characteristic values ​​of the target spatial grid after compaction, and to construct the state characteristic sequence of the target spatial grid using the state characteristic values ​​of the target spatial grid at different compaction times; and a control module configured to use the evolution gradient of the state characteristic sequence to determine the compaction convergence value, and to control the compaction operation of the compactor in the current working area through the cloud platform based on the compaction convergence values ​​of different spatial grids in the current working area.

[0007] This enables digital quality control of the compactor based on the cloud platform.

[0008] Optionally, the effective impact energy coefficient of the target spatial grid is determined as follows: logarithmically transform the ratio of the acceleration value within the target time window to the reference acceleration constant to obtain the logarithmic result; determine the first product of the logarithmic result and the acceleration value, and integrate the first product with time as the variable within the target time window to obtain the time integral value; determine the second product of the pressure peak within the target time window and the duration of the target time window, and use the ratio of the time integral value to the second product as the effective impact energy coefficient.

[0009] In this way, the product of the pressure peak and the duration is introduced for normalization, eliminating the baseline deviation caused by differences in the operating gear of different equipment and the hardness of the geological conditions, and ensuring that the extracted effective impact energy coefficient has universal applicability across working conditions.

[0010] Optionally, the rebound damping coefficient of the target spatial grid is determined as follows: Based on the displacement timeline, the peak moment of the rebound peak and the stable moment of soil recovery are extracted, and the rebound time difference between the peak moment and the stable moment is determined; the instantaneous displacement value, peak rebound displacement value, and stable displacement value corresponding to the displacement timeline are extracted, and the normalized displacement integral value is determined by combining the instantaneous displacement value, peak rebound displacement value, and stable displacement value; using the ratio of the pressure peak value to the reference pressure constant, the exponential decay value negatively correlated with the pressure peak value is determined, the normalized displacement integral value is divided by the rebound time difference, and the result of the division is multiplied by the exponential decay value to obtain the rebound damping coefficient of the target spatial grid.

[0011] By clearly defining the rebound range from the peak to the stable point, the interference of plastic deformation during the tamping hammer's descent can be effectively eliminated, allowing for the extraction of dynamic characteristics specifically for the soil's elastic recovery phase. Introducing an exponential decay value related to the pressure peak to nonlinearly modulate the integral results objectively reflects the physical reality of the nonlinear increase in interparticle friction under high-pressure stress, thus improving the characterization accuracy of the rebound damping coefficient under various working conditions.

[0012] Optionally, the normalized displacement integral value is determined as follows: the difference between the instantaneous displacement value and the stable displacement value is taken as the first displacement difference, and the difference between the peak rebound displacement value and the stable displacement value is taken as the second displacement difference; the ratio of the first displacement difference to the second displacement difference is taken as the displacement ratio, and the square of the displacement ratio is integrated over time over the values ​​between the peak time value and the stable time value, with time as the variable, to obtain the normalized displacement integral value.

[0013] In this way, by using stable displacement values ​​as a reference benchmark to construct a difference system, the absolute coordinate measurement error caused by uneven initial ground elevation is reduced. After squaring the displacement ratio and integrating it in the time dimension, the small shaking characteristics in the later stage of elastic oscillation can be amplified, making the calculated normalized displacement integral value more sensitive to the dissipation process of pore water pressure inside the soil.

[0014] Optionally, the post-compaction state characteristic value is determined in the following way: using the effective compaction energy coefficient of the target space grid at different compaction times, the average coefficient of variation is determined to characterize the variation range of the effective compaction energy coefficient; the first sum of the rebound damping coefficient and the first preset positive number is determined, and the second sum of the natural constant and the average coefficient of variation is determined; the ratio between the effective compaction energy coefficient and the first sum is multiplied by the second sum to obtain the post-compaction state characteristic value.

[0015] In this way, the effective impact energy coefficient and rebound damping coefficient are integrated through a ratio relationship, which combines information from two dimensions: input energy and soil resistance. The average coefficient of variation is used to capture the energy absorption stability during continuous construction.

[0016] Optionally, the average coefficient of variation is determined as follows: for the first effective impact energy coefficient and the second effective impact energy coefficient corresponding to adjacent impacts in different impact numbers of the target space grid, the first difference between the first effective impact energy coefficient and the second effective impact energy coefficient is determined, and the ratio of the absolute value of the first difference to the second effective impact energy coefficient is taken as the single relative change rate; the sum of the single relative change rates calculated within the preset backtracking window is taken as the ratio to the preset backtracking step size as the average coefficient of variation; the length of the backtracking window is equal to the preset backtracking step size.

[0017] Optionally, the compaction convergence value is determined as follows: The total number of compactions since the start of the operation is obtained; when the total number of compactions exceeds a preset sequence length, the latest consecutive state feature values ​​corresponding to the state feature sequence are obtained; the attenuation weight coefficient value is determined by combining the preset backtracking step size and the sequence length value; the absolute difference between adjacent state feature values ​​is determined by combining the consecutive state feature values; the absolute difference is multiplied by the attenuation weight coefficient value and then weighted and summed to obtain the weighted numerator; the sum of the consecutive state feature values ​​is used as the denominator, and the ratio of the weighted numerator to the denominator is used as the compaction convergence value.

[0018] In this way, by using the attenuation weight coefficient value to give higher weight to recent feature differences, the compaction convergence value can respond extremely sensitively to the current compaction saturation trend of the grid, providing a reliable quantitative criterion for the shutdown scheduling of the cloud platform.

[0019] Optionally, the decay weight coefficient value is determined by: determining the second difference between the sequence length value and the preset backtracking step size, and using the sum of the second difference and the second preset positive number as the decay weight coefficient value.

[0020] This ensures that the weight allocation is directly related to the algebraic relationship between the macroscopic length of the sequence analysis and the local backtracking step size, guaranteeing that the system can adaptively adjust the weight distribution when adjusting the backtracking step size under different soil and geological conditions. The addition of a second preset positive number provides a basic bias guarantee, avoiding the risk of the algorithm chain breaking when the difference is zero.

[0021] Optionally, the control module is also configured to perform the following processing steps: determine the number of consecutive times the compaction convergence value is lower than the set safety threshold; if the number of consecutive times is greater than the preset number, mark the target spatial grid as having reached a compacted state in the cloud platform; control the tamping machine through the cloud platform to continue compacting the spatial grids in the current work area that have not reached a compacted state; when all spatial grids in the current work area are marked as having reached the target compacted state, control the tamping machine through the cloud platform to stop tamping the current work area.

[0022] Optionally, the displacement timeline of the vertical displacement of the predetermined target during the elastic recovery of the soil after impact is obtained by the following method: by using a camera deployed at a stationary reference position, continuous image frames of the soil undergoing elastic recovery after being impacted by the compactor are acquired in the target space grid, and the vertical displacement of the target in the continuous image frames is tracked by optical flow method to obtain the displacement timeline.

[0023] The technical solutions provided by the embodiments of this application may include the following beneficial effects: by fusing pressure data and acceleration data to extract the effective compaction energy coefficient, and combining the obtained displacement time line to analyze the rebound damping coefficient, the mechanical response characteristics of deep soil under transient impact can be comprehensively and dynamically reflected. Based on the state characteristic sequence constructed by the effective compaction energy coefficient and the rebound damping coefficient, the compaction degree of local grids can be quantified. The global automated command issuance is realized through the cloud platform, which improves the operating efficiency of mechanical equipment and avoids the engineering problems of local under-compaction or over-compaction damaging the soil structure.

[0024] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0025] Figure 1 This is a schematic diagram illustrating the structure of a cloud-based digital quality control system for a hydraulic compactor, according to an exemplary embodiment. Figure 2 This is a schematic diagram illustrating the change in compaction convergence value in an embodiment of this application. Detailed Implementation

[0026] To implement digital quality control of the compaction process of a hydraulic rammer, this application provides a cloud-based digital quality control system 1000 for hydraulic rammers. Figure 1 This is a schematic diagram illustrating the structure of a cloud-based digital quality control system 1000 for a hydraulic compactor, according to an exemplary embodiment. Figure 1As shown, the cloud-based digital quality control system 1000 for hydraulic compactors includes: a first determining module 1100, a second determining module 1200, a construction module 1300, and a control module 1400.

[0027] The first determining module 1100 is configured to acquire the time-series pressure value of the hydraulic cylinder of the compactor during the impact cycle, and the time-series acceleration value of the compactor's hammer during the impact cycle. The time-series pressure value of the hydraulic cylinder is acquired in real time by a high-frequency hydraulic sensor array, while the time-series acceleration value is recorded by an industrial-grade piezoelectric accelerometer installed inside the hammer. The first determining module 1100 captures the pressure and acceleration values ​​within a target time window from the hammer's contact with the ground to its rebound and departure from the ground.

[0028] The first determining module 1100 can determine the effective impact energy coefficient of the target spatial grid falling into the current working area within the target time window; the current working area can be divided into several independent target spatial grids of uniform size through the positioning system.

[0029] The second determining module 1200 is configured to acquire the displacement timeline of the vertical displacement of the predetermined target when the soil undergoes elastic recovery after being hit, targeting a predetermined target in the target space grid; the predetermined target may be composed of a high-contrast black and white calibration plate sprayed on the outside of the compactor base guard plate, so that the optical equipment can continuously track it.

[0030] The second determining module 1200 determines the rebound damping coefficient of the target space grid through the displacement time line; the rebound damping coefficient is used to quantify the magnitude of the internal frictional resistance experienced by the soil when it releases elastic energy outward after being subjected to severe compression.

[0031] Module 1300 is configured to determine the state characteristic values ​​of the target spatial grid after impact by utilizing the effective impact energy coefficient and rebound damping coefficient of the target spatial grid. The state characteristic value is a dimensionless comprehensive evaluation index, and the magnitude of the state characteristic value directly maps the degree of compaction of the particle arrangement inside the target spatial grid after a specific impact.

[0032] The construction module 1300 uses the state characteristic values ​​of the target spatial grid at different compaction times to construct the state characteristic sequence of the target spatial grid; the state characteristic sequence contains the entire process evolution record of the soil gradually transitioning from a loose state to a firm state within a complete construction cycle.

[0033] The control module 1400 is configured to determine the compaction convergence value using the evolution gradient of the state feature sequence. The evolution gradient represents the decay rate of soil density as the number of blows increases. Based on the compaction convergence values ​​of different spatial grids in the current working area, the control module 1400 controls the compaction operation of the compactor in the current working area through the cloud platform.

[0034] In one embodiment, the effective impact energy coefficient of the target spatial grid is determined as follows: logarithmically transforming the ratio of the acceleration value within the target time window to a reference acceleration constant to obtain a logarithmic result; determining the first product of the logarithmic result and the acceleration value, integrating the first product with time as a variable within the target time window to obtain a time integral value; determining the second product of the pressure peak within the target time window and the duration of the target time window, and using the ratio of the time integral value to the second product as the effective impact energy coefficient.

[0035] In the process of acquiring dynamic data using sensors, acceleration signals may be accompanied by strong high-frequency oscillation noise. Directly calculating the original acceleration may cause the energy value to diverge significantly. Logarithmic transformation can be used to nonlinearly smooth out extreme peaks.

[0036] By multiplying the acceleration characteristics by the pressure value and then performing an integral operation, the discrete time-domain signal can be transformed into a macroscopic parameter representing the total work done. By introducing the product of the pressure peak value and the time difference as a denominator, the output energy of different equipment models and under different geological conditions can be standardized.

[0037] When the equipment operates on a hard foundation, the rebound time is short. The energy scale can be adaptively adjusted by the time difference, and the obtained effective impact energy coefficient can objectively reflect the net work efficiency of the tamping hammer on the soil.

[0038] The effective impact energy coefficient of the target space grid can be determined in the following way: ,in The first element of the target space grid represents the first element. Effective impact energy coefficient of a single impact. and The numbers represent the order of the numbers. The initial moment of contact with the ground and the moment of rebound and separation during the second impact. Within the target time window The pressure value at any moment, Within the target time window acceleration at any moment For reference acceleration constant, The peak pressure within the target time window.

[0039] As can be seen from the above formula for calculating the effective impact energy coefficient, the numerator term integrates the nonlinearly smoothed acceleration factor with the real-time pressure value, which maps the core physical process of converting impact kinetic energy into soil plastic strain energy. If high-frequency oscillations cause extremely large instantaneous acceleration absolute values, the logarithmic function will limit the gain to a reasonable range.

[0040] The denominator is the product of the pressure extreme value and the time of action. When the hydraulic system pressure suddenly increases or the ground surface is too soft, causing the compaction time to be prolonged, the corresponding increase in the denominator can offset the expansion of the absolute value of the numerator. The close linkage between parameters ensures that the calculated effective compaction energy coefficient maintains dimensional consistency and numerical stability in a physical sense.

[0041] In one embodiment, the rebound damping coefficient of the target spatial grid is determined as follows: The peak moment of the rebound peak and the stable moment of soil recovery are extracted based on the displacement timeline; the rebound time difference between the peak moment and the stable moment is determined; the instantaneous displacement value, peak rebound displacement value, and stable displacement value corresponding to the displacement timeline are extracted; the normalized displacement integral value is determined by combining the instantaneous displacement value, peak rebound displacement value, and stable displacement value; the exponential decay value, which is negatively correlated with the pressure peak, is determined using the ratio of the pressure peak value to the reference pressure constant; the normalized displacement integral value is divided by the rebound time difference; and the result of the division is multiplied by the exponential decay value to obtain the rebound damping coefficient of the target spatial grid.

[0042] The deformation process of soil after being subjected to impact load includes two different physical stages: plastic compression and elastic recovery. By extracting the data segment from the peak moment to the steady moment, we can focus on the recovery range with spring-like and damping characteristics.

[0043] For example, the peak moment occurs 15 seconds after impact, and the stable moment occurs 80 seconds after impact. By extracting features such as instantaneous displacement values, the oscillation pattern of the soil within this 65-second rebound time difference can be characterized.

[0044] The formula for calculating the rebound damping coefficient is as follows: , Indicates the rebound damping coefficient; This represents the normalized displacement integral value; Indicates the difference in rebound time; Represents the natural constant; Indicates peak pressure; This represents the reference pressure constant.

[0045] The reference pressure constant is used to achieve dimensionless processing of the pressure peak value. For example, it can be determined in advance based on historical pressure information, which will not be described in detail in the embodiments of this application.

[0046] In one embodiment, the normalized displacement integral value is determined as follows: the difference between the instantaneous displacement value and the stable displacement value is taken as the first displacement difference; the difference between the peak rebound displacement value and the stable displacement value is taken as the second displacement difference; the ratio of the first displacement difference to the second displacement difference is taken as the displacement ratio; and the square of the displacement ratio is integrated over time, with time as the variable, between the peak time value and the stable time value, to obtain the normalized displacement integral value.

[0047] Due to the undulating terrain and the sinking of the equipment tracks at the construction site, the initial reference elevation captured by the camera before each impact may not be fixed. If the absolute displacement coordinates are used directly for integration, various slow equipment drift errors may be introduced into the calculation process.

[0048] The first displacement difference is obtained by subtracting the stable displacement value from the instantaneous displacement value. A local relative coordinate system can be dynamically established for the current impact, thereby shielding the systematic bias caused by global elevation changes.

[0049] Constructing the displacement ratio and performing a squaring operation can amplify the small damped fluctuations in the later stages of elastic oscillation. Since the displacement amplitude is weaker in the later stages, it is easily overwhelmed by numerical truncation errors when performing integration in the time dimension. The squaring operation can give nonlinear weights to the small residual oscillations, making the calculated normalized displacement integral value more sensitive to the dissipation rate of pore water pressure after soil particle rearrangement.

[0050] In one embodiment, the post-compaction state characteristic value is determined by: using the effective compaction energy coefficient of the target space grid at different compaction times, determining the average coefficient of variation used to characterize the variation range of the effective compaction energy coefficient; determining the first sum of the rebound damping coefficient and a first preset positive number; determining the second sum of the natural constant and the average coefficient of variation; and multiplying the ratio between the effective compaction energy coefficient and the first sum by the second sum to obtain the post-compaction state characteristic value.

[0051] The soil contains a complex stress relaxation mechanism. Data from a single impact may not be sufficient to comprehensively assess the compaction effect. The average coefficient of variation can macroscopically characterize whether the energy absorption efficiency has entered a plateau period as the number of impacts increases. For example, the effective compaction energy coefficient fluctuates greatly in the first three impacts, indicating that soil particles are undergoing dislocation and rearrangement.

[0052] The first sum is obtained by adding the rebound damping coefficient to the first preset positive number and placing it in the denominator. The purpose is to ensure that the algebraic operation of the whole fraction does not result in division by zero when the soil is close to a saturated and dense state, causing the damping coefficient to approach zero infinitely. The second sum is multiplied by the ratio structure at the front end, so that the state characteristic value not only integrates the current energy and resistance relationship, but also has the function of remembering the historical trend.

[0053] The formula for calculating the state eigenvalues ​​can be: , State eigenvalues To effectively improve the impact energy coefficient The rebound damping coefficient The first preset positive number natural constant is the average coefficient of variation.

[0054] The first preset positive number is used to avoid the denominator being 0. For example, it can be equal to a positive integer smaller than the rebound damping coefficient. As the construction progresses, the soil becomes more and more compacted, and the effective compaction energy coefficient will gradually stabilize, causing the average coefficient of variation to continuously decrease and approach zero. The rebound damping coefficient will also change accordingly due to the increase in the elastic modulus. The state characteristic value takes into account different parameters in the compaction process and can better characterize the compaction state of the soil where the processed spatial grid is located.

[0055] In one embodiment, the average coefficient of variation is determined as follows: for adjacent impacts of the target spatial grid at different impact frequencies, the first effective impact energy coefficient and the second effective impact energy coefficient are respectively determined; a first difference between the first effective impact energy coefficient and the second effective impact energy coefficient is determined; the ratio of the absolute value of the first difference to the second effective impact energy coefficient is taken as the single relative change rate; the ratio of the sum of the single relative change rates calculated within a preset backtracking window to a preset backtracking step size is taken as the average coefficient of variation; the length of the backtracking window is equal to the preset backtracking step size.

[0056] By analyzing the difference in the effective compaction energy coefficient between two adjacent compaction blows, we can eliminate the gradual trend caused by overall geological changes in the site and focus on the local densification increment directly caused by each compaction blow. For example, the first effective compaction energy coefficient of the 6th compaction blow is 85, and the second effective compaction energy coefficient of the 7th compaction blow is 88. By taking the absolute value of the difference between these two and dividing it by the data of the next compaction blow, we can objectively describe the relative changes caused by the compaction on the soil of the spatial grid.

[0057] By introducing a backtracking window and a backtracking step size, smoothing filters can be used to eliminate misjudgments caused by occasional noise from a single sensor. If a strike happens to hit a lone rock on the ground, it can cause an abnormal spike in the single relative change rate. By summing all single relative change rates within a specific backtracking window and dividing by the backtracking step size, high-frequency transient disturbances can be effectively distributed over the entire time period, ensuring that the calculated average coefficient of variation has good noise resistance.

[0058] The formula for calculating the average coefficient of variation is as follows: , This represents the average coefficient of variation; Indicates the first effective impact energy coefficient; Indicates the second effective impact energy coefficient; This indicates the preset backtracking step size.

[0059] In one embodiment, the compaction convergence value is determined as follows: The total number of tamping blows accumulated since the start of the operation is obtained; when the total number of tamping blows is greater than a preset sequence length value, the latest consecutive multiple state feature values ​​corresponding to the state feature sequence are obtained; the attenuation weight coefficient value is determined by combining a preset backtracking step size and the sequence length value; the absolute difference between adjacent state feature values ​​is determined by combining the consecutive multiple state feature values; the absolute difference is multiplied by the attenuation weight coefficient value and then weighted and summed to obtain a weighted numerator; the sum of the consecutive multiple state feature values ​​is used as the denominator; and the ratio of the weighted numerator to the denominator is used as the compaction convergence value.

[0060] In the initial stage of compaction, the surface soil may experience random massive collapses or lateral slippages. Any characteristic data collected at this time may not be representative. Setting a preset sequence length value, such as requiring a total number of compaction blows of more than 5 before activating the evaluation mechanism, can effectively filter out the interference of data from the initial disordered interference stage. Obtaining the latest continuous multiple state characteristic values ​​for joint calculation can enable compaction operations based on the latest soil mechanical response state.

[0061] The closer the feature changes are to the current moment, the more accurately they reflect the degree to which the soil is close to saturation. Applying a decay weight coefficient to the absolute difference between adjacent state feature values ​​and performing a weighted sum can smooth the time series difference. The decay weight coefficient gives higher attention to feature changes that are closer to the current moment.

[0062] The formula for calculating the compaction convergence value can be: , This represents the compaction convergence value; This represents the attenuation weighting coefficient value; This represents the state feature value at the previous position among a series of consecutive state feature values; It represents the state characteristic value at the last position among multiple consecutive state characteristic values; the numerator of the formula for calculating the compaction convergence value is the weighted result of all adjacent differences accumulated by summing the summation symbol.

[0063] If the current grid soil is still loose, each impact will cause a huge jump in the state characteristic value. At this time, the weighted numerator of the compaction convergence value will remain high, making the calculated compaction convergence value much larger than the shutdown standard. Only when all adjacent differences shrink to a minimum value due to the completion of compaction will the output compaction convergence value smoothly fall into the set threshold safety zone.

[0064] In one embodiment, the decay weight coefficient value is determined by: determining a second difference between the sequence length value and a preset backtracking step size, and using the sum of the second difference and a second preset positive number as the decay weight coefficient value.

[0065] In different engineering geological environments, such as areas with high water content clay fill or areas with coarse gravel fill, the preset backtracking step size can be dynamically fine-tuned to ensure calculation accuracy. By establishing an algebraic relationship between the preset backtracking step size and the sequence length value to obtain the second difference, the weight system can be adaptively updated under various working conditions. When it is necessary to extend the backtracking time, the corresponding weight coefficient value will automatically shrink, thereby making the smoothing effect of long-term data more obvious.

[0066] The second preset positive number provides a minimum safety margin to ensure that the system can maintain the normal operation of the core quality control logic even when faced with extreme parameter settings or abnormal network configurations.

[0067] The formula for calculating the attenuation weighting coefficient is as follows: , Attenuation weighting coefficient value This is the sequence length value. Preset backtracking step size, The second preset positive number.

[0068] As the sequence length increases to accommodate more historical records, the decay weights also expand linearly, achieving adaptive adjustment of the weight coefficient values. Furthermore, the linear linkage mechanism helps reduce the processing pressure on pipeline data generated by heavy compaction machinery at high sampling rates.

[0069] Figure 2 This is a schematic diagram illustrating the change in compaction convergence value in an embodiment of this application, as shown below. Figure 2 As shown in the figure, the horizontal axis represents the cumulative number of operations, and the vertical axis represents the compaction convergence value. In the early stage of compaction, for example, when the cumulative number of operations is 60, the soil particles rearrangement and densification increase caused by each tamping are more intense because the pores inside the deep soil are larger and in a relatively loose state.

[0070] like Figure 2 As shown, in the early stage of compaction, the compaction convergence value obtained by weighted differential gradient in this embodiment of the application exhibits significant high-level wide-range fluctuations. This indicates that the compaction convergence value in this embodiment of the application has characteristic amplification capability and sensitivity to the non-dense state of the soil. As the construction operation continues (e.g., the cumulative number of operations is between 60 and 100), the soil density gradually tends to saturate, while the judgment value curve of the existing baseline algorithm tends to flatten overall, and continues to experience slight repeated oscillations when approaching the baseline threshold.

[0071] In one embodiment, the control module 1400 is further configured to perform the following processing steps: determine the number of consecutive times the compaction convergence value is lower than a set safety threshold; if the number of consecutive times is greater than a preset number, mark the target spatial grid as having reached a compacted state in the cloud platform; control the tamping machine through the cloud platform to continue compacting the spatial grids in the current work area that have not reached a compacted state; when all spatial grids in the current work area are marked as having reached the target compacted state, control the tamping machine through the cloud platform to stop tamping the current work area.

[0072] For example, if the convergence value is less than 0.05 in three consecutive calculations, it can not only filter out misjudgment signals caused by hardware noise, but also ensure that the foundation soil with a truly stable response, thus ensuring that the status marker has engineering credibility.

[0073] The cloud platform can issue scheduling instructions through the Internet of Things link, realizing a fully automated closed loop in the construction process. When it is confirmed that a certain spatial grid has reached a compact state, the grid color label can be changed on the digital construction map, and the operator or the underlying firmware of the autonomous driving system can be guided to move the robotic arm to the next uncomplied area.

[0074] By managing the compaction process of individual grids, the repeated over-compaction behavior that may occur with the compactor is avoided. This not only helps reduce the energy consumption of the hydraulic compactor, but also prevents the re-loosening and structural damage of shallow soil due to excessive mechanical disturbance.

[0075] In one embodiment, the control module 1400 is further configured to perform the following processing steps: acquire the actual surface height information and the expected target surface height information of the current working area; determine the height difference information between the actual surface height information and the target surface height information; determine the spatial grid to be filled in the current working area based on the compaction convergence value and height difference information of different spatial grids in the current working area; and output backfilling instructions through the cloud platform, the backfilling instructions being used to control the backfilling equipment to process the spatial grid to be filled.

[0076] The quality of roadbed construction is assessed not only by the density of deep soil but also, crucially, by the smoothness and elevation of the surface. Under prolonged compaction, certain areas with high moisture content may experience severe volume shrinkage, resulting in an actual surface height significantly lower than the design elevation. By incorporating height difference information for composite evaluation, the system can establish a three-dimensional quality inspection model that considers both internal mechanical properties and external geometric dimensions. This composite evaluation method provides the cloud platform with a global perspective, enabling it to detect three-dimensional defects in every grid cell.

[0077] When the compaction convergence value of a certain grid has met the shutdown requirements, but the height difference still shows as a depression, the cloud platform can automatically identify the defect location that needs to be filled with material; the generated digital backfilling command can be sent to the surrounding loader or grader terminal via the network, which can speed up the turnover efficiency of earthwork filling.

[0078] In one embodiment, after determining the effective impact energy coefficient falling within the target space grid, the first determining module 1100 is further configured to perform the following processing steps: deploying a ring-shaped data buffer within the edge computing gateway of the hydraulic compactor; storing the collected time-series pressure and acceleration values ​​in the ring-shaped data buffer during network communication interruption; filling the data gaps caused by packet loss using a Lagrange polynomial interpolation algorithm to obtain a repaired data sequence when the network connection is restored and data is resynchronized to the cloud platform; and performing curve curvature verification between the repaired data sequence and historical data stored in the cloud platform.

[0079] Infrastructure projects may be far from urban base stations and may face problems such as mobile network signal attenuation and intermittent disconnection. By deploying a ring-shaped data buffer at the hydraulic equipment end, in the event of an unexpected interruption of the long connection with the cloud platform's transmission control protocol, the mechanism of cyclic overlay can be used to ensure that the latest generated time-series waveforms do not disappear out of thin air, thus preserving the most authentic original digital evidence of the construction process.

[0080] The Lagrange polynomial interpolation algorithm is well adapted to dealing with the loss of discrete data caused by network congestion. During data retransmission, if blank holes are found in the timestamp sequence, the interpolation algorithm can use the effective sampling points on both sides of the hole to reconstruct a smooth curve that conforms to the current impact dynamics.

[0081] In one embodiment, the displacement timeline of the vertical displacement of the predetermined target during the elastic recovery of the soil after impact is obtained by: acquiring continuous image frames of the soil elastic recovery stage after being impacted by a compactor in the target space grid using a camera deployed at a stationary reference position; and obtaining the displacement timeline by tracking the vertical displacement of the target in the continuous image frames using optical flow.

[0082] Traditional contact-type mechanical displacement pull-wire sensors may suffer mechanical fatigue fracture or severe zero-point drift under the vibration of hydraulic compactors with accelerations of up to tens of gravitational forces. By using an industrial camera deployed at a stationary reference position for visual capture, non-contact detection of the target can be achieved, avoiding the interference of vibration on the image acquisition process. This not only extends the service life of the detection equipment but also eliminates the unmeasurable interference of the pull-wire's own inertial mass on the high-frequency oscillation waveform.

[0083] After acquiring continuous image frames using high-speed camera technology, the optical flow method can calculate the motion vector field of the target in two-dimensional space by analyzing the instantaneous change rate of pixel grayscale patterns between adjacent frames, and transform the soil elastic recovery process into a displacement timeline, which facilitates the automated control of the compaction process.

[0084] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A cloud-based digital quality control system for hydraulic compactors, characterized in that, include: The first determining module is configured to acquire the timing pressure value of the hydraulic cylinder of the compactor during the impact cycle, and the timing acceleration value of the hammer of the compactor during the impact cycle, and to extract the pressure and acceleration values ​​within the target time window from the time the hammer contacts the ground to the time it rebounds and leaves the ground, so as to determine the effective compaction energy coefficient of the target spatial grid falling into the current working area within the target time window. The second determining module is configured to acquire the displacement timeline of the vertical displacement of the predetermined target when the soil elastically recovers after being hit, and determine the rebound damping coefficient of the target space grid through the displacement timeline. The construction module is configured to use the effective impact energy coefficient and rebound damping coefficient of the target space grid to determine the state characteristic value of the target space grid after impact, and to use the state characteristic value of the target space grid at different impact times to construct the state characteristic sequence of the target space grid. The control module is configured to determine the compaction convergence value using the evolution gradient of the state feature sequence, and control the compaction operation of the tamping machine in the current working area through the cloud platform based on the compaction convergence value of different spatial grids in the current working area.

2. The cloud-based digital quality control system for hydraulic compactors according to claim 1, characterized in that, The effective impact energy coefficient of the target space grid is determined in the following way: Logarithmically transform the ratio of the acceleration value within the target time window to the reference acceleration constant to obtain the logarithmic result; determine the first product of the logarithmic result and the acceleration value, and integrate the first product with time as the variable within the target time window to obtain the time integral value; The second product of the peak pressure within the target time window and the duration of the target time window is determined, and the ratio of the time integral value to the second product is used as the effective impact energy coefficient.

3. The cloud-based digital quality control system for hydraulic compactors according to claim 1, characterized in that, The rebound damping coefficient of the target space grid is determined in the following way: Based on the displacement timeline, the peak moment of reaching the rebound peak and the stable moment of soil recovery are extracted, and the rebound time difference between the peak moment and the stable moment is determined. Extract the instantaneous displacement value, peak rebound displacement value, and stable displacement value corresponding to the displacement time line, and combine the instantaneous displacement value, peak rebound displacement value, and stable displacement value to determine the normalized displacement integral value; By using the ratio of the peak pressure to the reference pressure constant, the exponential decay value, which is negatively correlated with the peak pressure, is determined. The normalized displacement integral value is divided by the rebound time difference, and the result of the division is multiplied by the exponential decay value to obtain the rebound damping coefficient of the target space grid.

4. The cloud-based digital quality control system for hydraulic compactors according to claim 3, characterized in that, The normalized displacement integral value is determined in the following way: The difference between the instantaneous displacement value and the stable displacement value is taken as the first displacement difference, and the difference between the peak rebound displacement value and the stable displacement value is taken as the second displacement difference. The ratio of the first displacement difference to the second displacement difference is used as the displacement ratio. The square of the displacement ratio is integrated over time, taking into account the values ​​between the peak time value and the steady time value, to obtain the normalized displacement integral value.

5. The cloud-based digital quality control system for hydraulic compactors according to claim 1, characterized in that, The post-compaction state characteristics are determined in the following way: By utilizing the effective impact energy coefficient of the target space grid at different impact frequencies, the average coefficient of variation used to characterize the variation range of the effective impact energy coefficient is determined. Determine the first sum of the rebound damping coefficient and the first preset positive number, and determine the second sum of the natural constant and the average coefficient of variation; multiply the ratio between the effective impact energy coefficient and the first sum by the second sum to obtain the state characteristic value after impact.

6. The cloud-based digital quality control system for hydraulic compactors according to claim 5, characterized in that, The average coefficient of variation is determined in the following way: For the first effective impact energy coefficient and the second effective impact energy coefficient corresponding to adjacent impacts in different impact numbers of the target space grid, the first difference between the first effective impact energy coefficient and the second effective impact energy coefficient is determined, and the ratio of the absolute value of the first difference to the second effective impact energy coefficient is taken as the single relative change rate. The average coefficient of variation is the ratio of the sum of the single relative change rates calculated within the preset backtracking window to the preset backtracking step size; the length of the backtracking window is equal to the preset backtracking step size.

7. The cloud-based digital quality control system for hydraulic compactors according to claim 1, characterized in that, The compaction convergence value is determined in the following way: Obtain the total number of tamping blows accumulated since the start of the operation. When the total number of tamping blows is greater than the preset sequence length value, obtain the latest consecutive multiple state feature values ​​corresponding to the state feature sequence. The decay weight coefficient value is determined by combining the preset backtracking step size and sequence length value. The absolute difference between adjacent state feature values ​​is determined by combining multiple consecutive state feature values. The absolute difference is multiplied by the decay weight coefficient value and then weighted and summed to obtain the weighted numerator. The sum of multiple consecutive state characteristic values ​​is used as the denominator, and the ratio of the weighted numerator to the denominator is used as the compaction convergence value.

8. The cloud-based digital quality control system for hydraulic compactors according to claim 7, characterized in that, The decay weight coefficient value is determined as follows: the second difference between the sequence length value and the preset backtracking step size is determined, and the sum of the second difference and the second preset positive number is used as the decay weight coefficient value.

9. The cloud-based digital quality control system for hydraulic compactors according to claim 1, characterized in that, The control module is also configured to perform the following processing steps: Determine the number of consecutive times the compaction convergence value is continuously lower than the set safety threshold. If the number of consecutive times is greater than the preset number, mark the target spatial grid as having reached a compacted state in the cloud platform. The cloud platform controls the compactor to continue compacting the spatial grids in the current work area that have not yet reached the compaction state. When all spatial grids in the current work area are marked as having reached the target compaction state, the cloud platform controls the compactor to stop compacting the current work area.

10. The cloud-based digital quality control system for hydraulic compactors according to claim 1, characterized in that, The displacement-time curve of the vertical displacement of the predetermined target during the elastic recovery of the soil after impact was obtained through the following method: By using a camera deployed at a static reference position, continuous image frames are acquired of the soil undergoing elastic recovery after being struck by a compactor in the target spatial grid. The vertical displacement of the target in the continuous image frames is tracked using optical flow to obtain a displacement timeline.

Images