A method and system for monitoring and feedback control of the construction quality of layered compaction of lime-soil
By installing vibration and moisture content sensors on road rollers, combining RTK positioning and Kriging interpolation algorithms to generate spatial distribution maps, and using short-term rainfall forecasts for feedforward control, the problems of continuous monitoring of soil-cement layer compaction and quality control during rainy season construction were solved, achieving intelligent quality monitoring and economic benefits.
Patent Information
- Application Number
- CN202611108733.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-24
- Publication Date
- 2026-11-10
AI Technical Summary
Traditional methods for testing the compaction of lime-soil layers cannot achieve continuous monitoring of moisture content and compaction degree across the entire area, cannot generate spatial distribution maps, and cannot promptly address uneven rainfall distribution during rainy season construction, leading to potential quality hazards and economic losses.
Vibration sensors and moisture sensors are used to continuously detect moisture content on the roller. RTK positioning and Kriging interpolation algorithms are combined to generate spatial distribution maps of moisture content and compaction degree. Feedforward control is performed through short-term rainfall forecasts, and intelligent quality control is achieved by combining feedback adjustment, generating a three-dimensional digital twin map of compaction degree.
It enables continuous full-area detection and intelligent control of the quality of soil-lime layer compaction, avoiding potential quality problems during rainy season construction, reducing economic losses, and providing full life-cycle data traceability support.
Smart Images

Figure CN123067367A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of foundation treatment and intelligent construction technology for building engineering, specifically to a method and system for real-time monitoring of compaction quality, assessment of spatial distribution of moisture content, and hybrid control of meteorological feedforward and feedback during the layered spreading and compaction of lime-soil in a large-area stockpile. Background Technology
[0002] Lime-stabilized soil (lime-stabilized soil) is widely used in foundation treatment for large-area stockpiles and roadbeds due to its good mechanical properties and strong water stability. Its strength development depends on the physicochemical reaction between lime and soil, and is closely related to the moisture content and compaction work during rolling. In large-area layered construction, accurate detection and effective control of compaction quality directly affect the bearing capacity and long-term performance of the project.
[0003] Traditional compaction testing uses the ring cutter method or sand cone method, taking only 3 points per 100 square meters. The test results are delayed by several hours and are destructive. More importantly, during the rainy season, rainfall distribution is significantly uneven in space. Traditional sampling tests cannot capture the spatial variations in moisture content and compaction, and many potential problems in the sampling gap area are missed, causing quality risks.
[0004] Moisture content is a decisive factor affecting compaction effectiveness: too high a moisture content can lead to "springy soil" or even siltation, while too low a moisture content makes it difficult for particles to compact. Current technologies still mainly rely on sampling and drying methods, which cannot generate spatial distribution maps of moisture content. Although there are continuous compaction testing methods based on vibration response, they mainly focus on the degree of compaction and fail to simultaneously address the evaluation of the spatial distribution of moisture content.
[0005] Existing intelligent compaction systems all rely on feedback control during the compaction process, resulting in a passive response. They often fail to respond adequately to sudden external disturbances such as rainfall, and by the time an anomaly is detected, the soil has already been soaked, requiring extensive rework and causing significant quality and economic losses. Furthermore, they lack zoning-layer 3D traceability mechanisms, and construction data is stored in isolated tables, making it difficult to query data based on spatial location and layer, which is detrimental to final acceptance and subsequent operation and maintenance. Summary of the Invention
[0006] To address the aforementioned shortcomings of existing technologies, this application provides a method and system for monitoring and feedback control of the construction quality of layered compaction of lime-soil, particularly suitable for large-scale lime-soil construction during the rainy season in coastal reclamation areas. This application enables continuous monitoring of moisture content and compaction degree across the entire area during the layered compaction of lime-soil in large-area stockpiles; automatic generation of a spatial distribution map of moisture content; feedforward-feedback hybrid control based on short-term rainfall forecasts; and three-dimensional traceability of the compaction quality of different zones and layers. It offers advantages such as continuous monitoring, intelligent control, and digital traceability, and effectively addresses the risk of quality control failure caused by uneven rainfall distribution during rainy season construction.
[0007] To achieve the above objectives, the technical solution adopted in this application is as follows:
[0008] A method for monitoring and feedback control of construction quality in layered compaction of lime-soil, such as Figure 2 As shown, the method includes the following steps:
[0009] S1. Install vibration sensors and moisture content sensors on the vibratory drum of the road roller, and select multiple points in the construction area to collect the vibration response signal, surface moisture content of the lime-soil, and compaction value obtained by detection at each point, and establish a mapping calibration model between moisture content, vibration response and compaction degree.
[0010] S2. When the road roller is compacting the lime-soil surface layer, the vibration sensor and the moisture content sensor continuously collect vibration data and moisture content data, and simultaneously record the spatial coordinates of each collection point; interpolate the moisture content of each collection point to generate a spatial distribution field map of the moisture content of the compacted layer, and substitute the vibration response of each collection point into the mapping calibration model to calculate the compaction degree value, thereby generating a spatial distribution field map of the compaction degree.
[0011] S3. Compare the spatial distribution map of moisture content with the preset optimal moisture content range point by point, automatically identify and mark the areas where the moisture content exceeds the optimal moisture content range, and issue corresponding warnings.
[0012] S4. Obtain short-term rainfall forecast data for the construction area; when the forecast rainfall exceeds a preset threshold, issue a feedforward control command; and during the compaction operation, issue a feedback control command to the roller based on the deviation between the real-time detected moisture content or compaction degree and the target value for feedback adjustment.
[0013] Furthermore, in S1, the vibration sensor is a triaxial accelerometer, the moisture content sensor is a near-infrared moisture content sensor; the compaction value is obtained by sand cone method; the mapping calibration model is a ternary mapping model with moisture content and vibration response characteristics as independent variables and compaction degree as dependent variable.
[0014] Furthermore, in step S2, the spatial coordinates of each acquisition point are recorded by the RTK positioning module, and the vibration data, moisture content data, and spatial coordinates are uploaded to the edge computing terminal via the wireless transmission module; the interpolation calculation of the moisture content adopts the Kriging interpolation algorithm.
[0015] Furthermore, the spatial distribution map of moisture content and the spatial distribution map of compaction are displayed in pseudo-color format, with qualified areas, areas above the upper limit and areas below the lower limit marked with different colors.
[0016] Furthermore, in step S3, when the moisture content at the collection point is higher than the upper limit of the optimal moisture content range, a warning instruction to suspend compaction and allow drying is issued; when the moisture content at the collection point is lower than the lower limit of the optimal moisture content range, a warning instruction to supplement watering is issued.
[0017] Furthermore, in S4, the feedforward control command includes: issuing a work stoppage warning to the roller operator and site management personnel; automatically calculating the required area of waterproof tarpaulin to be covered based on the current construction area, and generating tarpaulin laying path guidance.
[0018] Furthermore, in step S4, after the rainfall ends, the moisture content distribution of the lime-soil layer is collected and evaluated again. Only when all the moisture content test values fall within the optimal moisture content range is an instruction to allow resumption of work issued.
[0019] Furthermore, the feedback control command includes at least one of adjusting the vibration frequency of the roller, adjusting the compaction speed, or increasing the number of compaction passes; the method also includes: dividing the construction area into multiple construction zones, storing the spatial distribution maps of moisture content, compaction degree, and compaction trajectory of each construction zone and each lime-soil layer in a spatiotemporal manner, generating a three-dimensional compaction degree digital twin map, wherein the three-dimensional compaction degree digital twin map supports traceability query of compaction quality by location and lime-soil layer.
[0020] A quality monitoring and feedback control system for layered compaction of lime-soil, used to implement the method described above, includes:
[0021] The sensing and detection group includes a vibration sensor and a moisture content sensor installed on the vibrating wheel of the road roller, used to collect vibration data and moisture content data;
[0022] The positioning and transmission group includes a positioning module and a wireless transmission module. The positioning module is used to record the spatial coordinates of each collection point, and the wireless transmission module is used to upload the collected data and spatial coordinates.
[0023] The edge computing terminal, connected to the positioning and transmission group, is used to receive the data and generate a spatial distribution map of moisture content and a spatial distribution map of compaction, identify areas with excessive moisture content, and is equipped with a meteorological interface for obtaining short-term rainfall forecast data.
[0024] And a display terminal connected to the edge computing terminal.
[0025] Furthermore, the display terminal includes a cab display terminal installed in the cab of the road roller and a mobile terminal for the site management personnel; the edge computing terminal is connected to the cab display terminal and the mobile terminal for the site management personnel via a wireless network.
[0026] Compared with existing technologies, the method and system for monitoring and feedback control of the construction quality of layered compaction of lime-soil proposed in this application have the following beneficial effects:
[0027] By integrating vibration and moisture content sensors onto the vibratory drum of a road roller, the sensors continuously scan the entire compacted surface as the roller advances. An interpolation algorithm generates spatial distribution maps of moisture content and compaction degree from discrete sampling points. Compared to traditional sparse sampling of 3 points per 100 square meters, this application covers the entire compacted surface, fundamentally avoiding the omission of quality risks in areas with excessive moisture content. It completely solves the technical problem of traditional sampling failing to capture spatial variations under uneven rainfall distribution, upgrading quality inspection from discrete point sampling to a comprehensive surface survey.
[0028] By introducing short-term rainfall forecasts as feedforward control signals, proactive warnings of work stoppages and guidance on waterproof tarpaulin covering are issued before rainfall arrives. After rainfall, the system proactively assesses whether the moisture content has returned to the allowable range to determine whether work can resume. Simultaneously, feedback adjustment capabilities are retained during the compaction process. Feedforward control proactively prevents external rainfall disturbances, while feedback control provides immediate correction for internal construction deviations. The two work together to overcome the passive shortcomings of traditional pure feedback control, which often fails to respond in time, effectively reducing quality risks and economic losses during rainy season construction.
[0029] By storing data such as moisture content field maps, compaction field maps, rolling trajectories, and number of rolling passes of each construction zone and each lime-soil layer in a spatiotemporal manner, a three-dimensional digital twin map of compaction degree is formed, realizing the full life cycle traceability of compaction quality. This provides complete and traceable data support for project acceptance and operation and maintenance management, and has good economic and social benefits. Attached Figure Description
[0030] Figure 1 This is a schematic diagram of the system structure and data flow of this application;
[0031] Figure 2 This is a flowchart of the control method of this application.
[0032] In the diagram: 1. Road roller; 2. Vibratory wheel; 3. Sensor detection group; 4. Positioning and transmission group; 5. Edge computing terminal; 6. Display terminal; 7. Mobile terminal; 8. Service platform. Detailed Implementation
[0033] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the application and are not intended to limit the application. The specific numerical values recorded herein are merely exemplary preferred values, and those skilled in the art can adjust them within the scope of the concept of this application according to actual application needs.
[0034] like Figure 1 As shown, this embodiment provides a monitoring and feedback control system for the construction quality of layered compaction of lime-soil. It mainly includes a sensor detection group and a positioning and transmission group located on the vehicle, and an edge computing terminal 5 located at the edge computing end. It works in conjunction with a display terminal and an external meteorological service platform. The entire system can be divided according to data flow into a perception layer (sensor detection group 3), a transmission and positioning layer (positioning and transmission group 4), a calculation and decision-making layer (edge computing terminal 5), and a display and execution layer (display terminal 6, mobile terminal 7, road roller 1), with an external meteorological service platform (service platform 8) serving as the feedforward information source.
[0035] Roller 1 is the main body of the construction machinery. A single-drum vibratory roller is preferred, with a working mass of not less than 20t. The diameter and width of the steel drum meet the requirements for large-area compaction of lime-soil layers, and it possesses automatic frequency and amplitude modulation functions. Selecting a roller with a working mass of not less than 20t ensures sufficient compaction work for large areas and thick lime-soil layers. The automatic frequency and amplitude modulation function is the hardware foundation for implementing feedback control; that is, the frequency and speed modulation commands issued by the edge computing terminal 5 must be executed by the roller with this function.
[0036] The vibratory drum 2 is a key working device of the road roller 1, installed at the front of the roller. It contains an eccentric vibration mechanism that generates directional vibration driven by a hydraulic motor. The vibration frequency is adjustable, and the excitation force can reach 300kN-400kN. Sensor mounting bases are pre-installed on the outer wall of the vibratory drum. During compaction, the vibratory drum directly contacts and vibrates the lime-soil layer. Its vibration response characteristics (such as acceleration amplitude, dominant frequency, and harmonic ratio) change depending on the compaction state of the lime-soil beneath it—the denser the lime-soil, the stronger the reaction force on the vibratory drum, and the richer the harmonic components in the vibration response. This is the physical basis for indirectly reflecting the degree of compaction through vibration response.
[0037] The sensing group 3 is an integrated package of a triaxial accelerometer and a near-infrared moisture sensor, mounted on a pre-reserved base on the inner wall of the vibrating roller 2. The triaxial accelerometer, with a range of ±50g, is used to collect the triaxial vibration acceleration signal during the interaction between the vibrating roller and the soil. The use of a triaxial sensor, rather than a single axis, allows for complete acquisition of the vertical and horizontal vibration responses of the vibrating roller, improving the accuracy of vibration feature extraction. The near-infrared moisture sensor employs a reflective measurement principle, with a measurement depth of 0-10cm and an accuracy of ±1.5%, used to collect the volumetric moisture content of the soil surface. The non-contact near-infrared reflective measurement allows for real-time and continuous acquisition of moisture content during compaction, avoiding the lag of traditional drying methods. The signal output of the sensing group 3 is connected to the positioning and transmission group 4 via a shielded cable. The shielded cable is used to suppress interference from roller vibration and the electromagnetic environment on the weak sensing signal.
[0038] The positioning and transmission group 4 is an integrated package of an RTK high-precision positioning module and a wireless data transmission module, installed inside the cab of the road roller 1. The RTK positioning module has a horizontal positioning accuracy of ±2.5cm +1ppm, used to record the three-dimensional spatial coordinates of each acquisition point in real time; such high positioning accuracy is a prerequisite for generating accurate spatial distribution maps and achieving coordinate-based traceability. The wireless data transmission module adopts 4G full-network communication and supports the MQTT protocol. It packages the vibration signal and moisture content signal collected by the sensor detection group 3 with the coordinate signal collected by the positioning and transmission group 4 and uploads them to the edge computing terminal 5 in real time. The use of the lightweight IoT communication protocol MQTT can reduce communication overhead while ensuring real-time performance and adapt to the network conditions at the construction site.
[0039] Edge computing terminal 5, deployed in the on-site management office, is an industrial-grade embedded industrial control computer running a Linux operating system. It incorporates a three-dimensional mapping calibration model for moisture content, vibration response, and compaction degree, a Kriging interpolation algorithm library, a short-term rainfall forecast API interface, and a digital twin map database. Edge computing terminal 5 is the core of the entire system's computing and decision-making processes. The use of edge computing instead of cloud computing aims to decentralize data processing to the construction site, reducing communication latency and ensuring real-time early warning and feedback. Edge computing terminal 5 accesses the service platform 8 via the internet to obtain short-term rainfall forecast data for the construction area for the next 1-3 hours. The data output terminals of edge computing terminal 5 are connected to display terminal 6 and mobile terminal 7 via wireless networks.
[0040] The display terminal 6 in the cab preferably uses a 10.1-inch industrial-grade touch screen, installed in the cab of the road roller 1. It features high brightness and anti-glare capabilities, supports real-time pseudo-color image rendering and voice broadcast warnings, and is used to display real-time spatial distribution maps of moisture content, compaction degree, over-limit area warning signs, rainfall forecast warnings, and tarpaulin coverage path guidance maps. The high brightness and anti-glare functions are designed to adapt to the strong light environment of open-air construction; the voice broadcast ensures that the operator can receive timely warnings even when focused on driving.
[0041] The mobile terminal 7 of the on-site management personnel is wirelessly connected to the edge computing terminal 5 via a WeChat mini program or a dedicated APP. It is used to receive various early warning information and construction progress pushed by the edge computing terminal 5, and supports remote viewing of the spatial distribution map of moisture content and compaction degree. This allows management personnel to keep abreast of the construction quality status in real time and monitor it remotely even when they are not on-site.
[0042] The meteorological service platform 8 provides the edge computing terminal 5 with an API data interface for short-term (1-3 hour) rainfall forecasts for the construction area via the Internet. The forecast content includes the forecast rainfall amount (mm), rainfall intensity (mm / h), and the start and end times of the rainfall. The data update frequency is no less than once every 30 minutes. The reason for using 1-3 hour short-term (near-term) forecasts instead of medium- and long-term forecasts is that short-term forecasts have higher accuracy in predicting the time and intensity of local severe convective rainfall, which can provide sufficient lead time for feedforward control without too many false alarms.
[0043] During system operation, such as Figure 1 As shown in the data flow diagram, the vibration and moisture content signals collected by the sensing and detection group 3 and the coordinate signals collected by the positioning and transmission group 4 are uploaded to the edge computing terminal 5 via the wireless transmission module of the positioning and transmission group 4. After completing data processing, field map generation, over-limit identification, meteorological feedforward judgment and feedback control decision, the edge computing terminal 5 pushes the results to the display terminal 6 and the mobile terminal 7 respectively, and sends feedback adjustment commands to the control system of the road roller 1, thus forming a complete closed loop from perception to calculation to decision-making to execution to feedback.
[0044] Based on the above system, the construction process of the real-time monitoring and feedback control method for the construction quality of large-area stockpile ash-soil layered compaction is as follows.
[0045] S1. Establish a mapping calibration model.
[0046] After the lime-soil paving is completed and before compaction begins, the sensor detection group 3 is installed on the reserved base of the vibratory roller 2, and the positioning and transmission group 4 is installed in the cab of the roller 1. Representative points (preferably 5) are selected within the construction area. These points should cover the distribution range of lime-soil moisture content from low to high to ensure the calibration model has good applicability throughout the entire moisture content range. At each point, the vibration response signal and moisture content reading of the sensor detection group 3 are recorded simultaneously, and the compaction degree at that point is determined using the sand cone method.
[0047] Data collected from five locations is entered into edge computing terminal 5. Edge computing terminal 5 automatically fits and generates a ternary mapping calibration model of moisture content, vibration response, and compaction degree. This model is in the form W(moisture content) = f(A,F,H) and D(compaction degree) = g(A,F,W), where A is the acceleration amplitude, F is the dominant frequency, and H is the harmonic ratio. Specifically, the first mapping relationship W = f(A,F,H) is used to assist in verifying the moisture content using vibration response characteristics in subsequent continuous testing. The second mapping relationship D = g(A,F,W) uses vibration response characteristics and moisture content as independent variables and compaction degree as the dependent variable, and is used to calculate the compaction degree in real time from continuously collected vibration signals and moisture content. Model fitting can be achieved using multiple regression, support vector regression, or neural networks. Five (or more) sets of measured vibration characteristic-moisture content-compaction degree data are used as training samples to fit the parameters of the above mapping relationship. Some measured points can be retained to verify the model accuracy.
[0048] Through the above calibration, a ternary mapping model is established with moisture content and vibration response characteristics as independent variables and compaction degree as the dependent variable. This provides a basis for subsequent real-time conversion of compaction degree using continuously acquired vibration signals, which is the technical premise for this application to achieve continuous detection instead of sampling detection. It is understood that the number of calibration points is not limited to 5; 3-5 or more representative points can be selected based on the moisture content distribution range of the construction area to ensure the accuracy of the calibration model. Furthermore, for lime-soil with different mix proportions and material sources, separate calibration is recommended to ensure the accuracy of the conversion results.
[0049] S2. Generate a spatial distribution field map.
[0050] When the road roller is compacting the lime-soil surface, the sensing and detection group 3 continuously collects vibration and moisture content data at a fixed frequency (preferably 1Hz or higher), while the positioning and transmission group 4 simultaneously records the three-dimensional spatial coordinates (X, Y, Z) of each collection point. As the road roller moves continuously during compaction, the collection points gradually cover the entire compacted surface, thus achieving continuous scanning detection of the compacted surface. The collected data is uploaded in real time to the edge computing terminal 5 via the wireless transmission module of the positioning and transmission group 4.
[0051] After receiving data from discrete collection points, the edge computing terminal 5 performs Kriging interpolation calculations on the moisture content of each collection point to generate a spatial distribution field map of moisture content covering the entire compaction surface. At the same time, it substitutes the vibration response characteristics of each collection point into the ternary mapping calibration model established by S1 to calculate the real-time compaction value of each point, and then interpolates the compaction value to generate a spatial distribution field map of compaction.
[0052] Kriging interpolation is an optimal unbiased estimation method based on spatial statistics. Its basic principle is as follows: for any unsampled location on the compaction surface, the estimated moisture content (or compaction degree) is taken as the weighted sum of the measured values from several surrounding sampled points. The weight of each sampled point is determined based on the spatial correlation (described by the variogram) between the sampled point and the point to be estimated—the closer the sampled point is to the point to be estimated and the stronger the spatial correlation, the greater its weight. By solving the set of weight coefficient equations that minimizes the estimation variance and ensures unbiased estimation, Kriging interpolation can obtain a continuous, smooth, and error-minimized distribution of moisture content and compaction degree across the entire compaction surface from discrete sampled points. This objectively and accurately reflects the spatial variation of quality, which is the key algorithmic support for overcoming the limitation of traditional sampling inspection in capturing spatial inhomogeneity.
[0053] The generated spatial distribution maps of moisture content and compaction degree are displayed as pseudo-color images on the operator's cab terminal of display terminal 6. Qualified areas are represented in green, areas above the upper limit in red, and areas below the lower limit in blue. The roller operator can view the current surface compaction quality distribution in real time, intuitively determining which areas are qualified, which are too wet, and which are too dry, thus allowing for targeted subsequent compaction.
[0054] S3. Automatic identification and early warning of areas with excessive moisture content.
[0055] Edge computing terminal 5 compares the spatial distribution map of moisture content generated by S2 with the preset optimal moisture content range for lime-soil construction (optimal moisture content ±2%, i.e., the range determined by the lower and upper limits) point by point, automatically identifying areas with moisture content below the lower limit, within the range, and above the upper limit. For excessively wet areas above the upper limit, the system issues a warning command to pause compaction and allow drying to prevent the formation of springy soil from continuing compaction in an excessively wet state; for excessively dry areas below the lower limit, the system issues a warning command to supplement watering, and compaction is resumed only after the moisture content returns to within the range.
[0056] The warning information is displayed on the driver's cab terminal of display terminal 6 in the form of a heat map overlay, and is simultaneously pushed to the mobile terminal of the on-site management personnel on mobile terminal 7. This allows operators and managers to address the over-limit areas in a targeted manner, avoiding quality hazards caused by continuing to compact the vehicle when the moisture content is unqualified. By automating the identification and warning of over-limit conditions, this application transforms moisture content control from relying on manual experience to objective and quantitative judgment based on comprehensive data.
[0057] S4, Hybrid control of feedforward and feedback.
[0058] S41. Feedforward control based on short-term rainfall forecast.
[0059] Edge computing terminal 5 requests short-term rainfall forecast data for the construction area in the next 1-3 hours from service platform 8 every 30 minutes. The system accesses the API interface of service platform 8 via wireless network to obtain short-term rainfall forecast data for the construction area in real time. When the forecast rainfall exceeds a preset threshold (preferably 10mm), edge computing terminal 5 automatically triggers a feedforward control command. The threshold is preferably set at 10mm because 10mm is equivalent to moderate rainfall, which is sufficient to significantly increase the moisture content of the paved, uncovered lime-soil surface layer, affecting the compaction quality. This threshold can be adjusted according to the lime-soil mix ratio, paving thickness, and local rainfall characteristics. The feedforward control commands include:
[0060] (1) Issue a work stoppage warning to the roller operator and site management personnel;
[0061] (2) Based on the current construction area boundary and the already paved area, automatically calculate the area of waterproof canvas that needs to be covered and generate canvas laying path guidance. The guidance content is displayed on display terminal 6 and mobile terminal 7 to guide on-site personnel to complete the protective coverage with the shortest path and highest efficiency before the rain arrives.
[0062] (3) After the rainfall ends, the system will automatically call the moisture content detection function of S2 to assess whether the moisture content of the lime-soil layer has recovered to the allowable range for construction. Only when all the moisture content detection values are in the optimal moisture content range will work be allowed to resume and an instruction to resume work will be issued. If the moisture content is still outside the range, continue drying or take other treatment measures until the conditions for resuming work are met.
[0063] By introducing short-term rainfall forecasts as feedforward control signals, this application can proactively halt construction for protection before rainfall arrives, rather than passively dealing with the situation after the soil has been soaked by rainwater and its moisture content exceeds the standard. This overcomes the shortcomings of traditional pure feedback control, which cannot respond in time, and is especially suitable for construction scenarios with sudden and severe convective weather during the rainy season.
[0064] S42, Adjustment of feedback control.
[0065] During the compaction operation, the edge computing terminal 5 monitors the moisture content and compaction degree data in real time. When the moisture content or compaction degree deviates from the target value, the edge computing terminal 5 automatically sends adjustment commands to the roller control system of the roller 1, including at least one of the following: adjusting the vibration frequency, adjusting the compaction speed, or increasing the number of compaction passes (or issuing a suggestion to increase the number of compaction passes). For example, when the compaction degree of a certain area is detected to be lower than the design value, the edge computing terminal 5 sends a feedback adjustment command to the control system of the roller 1 to increase the vibration frequency from 30Hz to 35Hz to increase the compaction work; or to decrease the compaction speed from 3km / h to 2.5km / h to prolong the vibration time per unit area and improve the compaction effect; after several compaction passes, the test is checked again. If the standard is still not met, the number of compaction passes is increased until the compaction degree of the area reaches the design requirements. The above three feedback methods can be used individually or in combination.
[0066] The combination of feedforward control and feedback control constitutes the feedforward-feedback hybrid control strategy of this application: feedforward control provides proactive prevention against disturbances from the external environment (rainfall), while feedback control provides immediate correction of real-time deviations during construction. The two work together to achieve intelligent management and control of the entire process of lime-soil compaction construction quality.
[0067] S5. Generate a digital twin map of compaction degree.
[0068] Large-area stockpile construction is typically divided into multiple construction zones, each requiring the construction of multiple layers of lime-soil. After each construction zone and each layer of lime-soil is compacted, the edge computing terminal 5 stores the spatial distribution map of the moisture content, spatial distribution map of the compaction degree, heat map of the compaction trajectory, and distribution map of the number of compaction passes of that layer according to spatiotemporal association. It then stores these maps in a digital twin map database using an index structure of "zone number-layer number-date" to form a three-dimensional (spatial X, spatial Y, layer Z) compaction degree digital twin map.
[0069] In this index structure, the partition number identifies the location of the construction area on the plane, the layer number identifies the backfill layer to which the data belongs, and the date identifies the construction time. These three elements together constitute a unique spatiotemporal index for each data record. Based on this index structure, the digital twin map supports traceability queries of compaction quality at any location and for any backfill layer. During project completion and acceptance, the map can be used to query compaction quality data for any coordinate location and any backfill layer. After the project is put into use, if abnormal settlement occurs in a certain area, the historical construction quality of each layer in that area can be traced back, thereby quickly locating the problematic layer and its cause. Therefore, this application achieves full lifecycle traceability of compaction quality, providing complete and traceable data support for project acceptance and subsequent operation and maintenance management.
[0070] In summary, the system of this application forms a complete intelligent closed loop during operation: the sensor detection group 3 and the positioning and transmission group 4 complete the perception and transmission of vibration, moisture content, and coordinates; the edge computing terminal 5 completes data processing, dual-field map generation, over-limit identification, meteorological feedforward judgment, feedback decision-making, and digital twin storage; the display terminal 6 and the mobile terminal 7 complete information display and early warning push; and the road roller 1 completes the execution of compaction parameters based on feedback instructions. Therefore, this application achieves continuous detection, intelligent control, and digital traceability of the quality of layered compaction of lime-soil, effectively addressing the risk of quality control failure caused by uneven rainfall distribution during rainy season construction.
[0071] To further illustrate the implementation effect of this application, we take the rainy season construction of a second-phase stockpile project in a coastal reclamation area as an example. The foundation treatment requirements for this project are: 50cm of original trench soil mixed with 5% lime and compacted by rolling + layered backfilling > 60cm of 7% lime-stone soil + 20cm of 7% lime soil. The single layer of lime soil compaction area reaches tens of thousands of square meters. The construction is carried out during the rainy season, with frequent local strong convective rainfall.
[0072] During construction, the entire site was divided into several construction zones. Before compaction, five representative points were selected in each zone to simultaneously collect triaxial acceleration signals, near-infrared moisture content, and sand cone compaction degree. The edge computing terminal then fitted a ternary mapping calibration model of the soil source. During compaction, a sensor detection group installed on the vibratory roller continuously collected vibration and moisture content data at 1Hz. The RTK module simultaneously recorded the coordinates and uploaded them to the edge computing terminal via 4G. The terminal used Kriging interpolation to generate a pseudo-color field map of moisture content and compaction degree in real time, which was simultaneously displayed on the terminal in the driver's cab and the mobile phone of the management personnel. Green indicated acceptable conditions, red indicated excessive moisture, and blue indicated excessive dryness. For the identified excessively moist areas, the system issued a warning to pause compaction and allow the soil to dry; for the excessively dry areas, a warning to supplement watering was issued.
[0073] About one hour before a severe convective rainfall event, the edge computing terminal received a forecast rainfall exceeding 10mm from the meteorological platform. This triggered feedforward control: a work stoppage warning was issued to operators and managers, and the required area of waterproof tarpaulin to be covered in the current zone was calculated, generating a laying path guide. Based on this, the tarpaulin was covered before the rainfall, preventing the already laid soil layer from being soaked. After the rainfall, the system automatically re-detected the moisture content, issuing a resumption command only after all areas had reached the optimal moisture content range. During compaction, for areas with low compaction, the system automatically increased the vibration frequency from 30Hz to 35Hz and reduced the compaction speed from 3km / h to 2.5km / h, suggesting additional compaction passes if necessary. After each layer was compacted, the system archived various site maps according to "zone number-layer number-date," generating a 3D digital twin map for final acceptance and traceability.
[0074] Application results show that, after adopting this application, areas with excessive moisture content can be detected and dealt with in a timely and comprehensive manner, avoiding missed detections in traditional sampling and testing; sudden rainfall during the rainy season can be proactively protected before it arrives, avoiding large-scale rework; and quality traceability can be carried out at any location and layer during final acceptance, verifying the effectiveness of this application in continuous detection, intelligent control, and digital traceability, as well as its economic and social benefits.
[0075] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.
Claims
1. A method for monitoring and feedback control of construction quality in layered compaction of lime-soil mixtures, characterized in that, The method includes the following steps: S1. Install vibration sensors and moisture content sensors on the vibratory drum of the road roller, and select multiple points in the construction area to collect the vibration response signal, surface moisture content of the lime-soil, and compaction value obtained by detection at each point, and establish a mapping calibration model between moisture content, vibration response and compaction degree. S2. When the road roller is compacting the lime-soil surface layer, the vibration sensor and the moisture content sensor continuously collect vibration data and moisture content data, and simultaneously record the spatial coordinates of each collection point; interpolate the moisture content of each collection point to generate a spatial distribution field map of the moisture content of the compacted layer, and substitute the vibration response of each collection point into the mapping calibration model to calculate the compaction degree value, thereby generating a spatial distribution field map of the compaction degree. S3. Compare the spatial distribution map of moisture content with the preset optimal moisture content range point by point, automatically identify and mark the areas where the moisture content exceeds the optimal moisture content range, and issue corresponding warnings. S4. Obtain short-term rainfall forecast data for the construction area; when the forecast rainfall exceeds a preset threshold, issue a feedforward control command; and during the compaction operation, issue a feedback control command to the roller based on the deviation between the real-time detected moisture content or compaction degree and the target value for feedback adjustment.
2. The method according to claim 1, characterized in that, In step S1, the vibration sensor is a triaxial accelerometer, and the moisture content sensor is a near-infrared moisture content sensor; the compaction value is obtained by sand cone method; the mapping calibration model is a ternary mapping model with moisture content and vibration response characteristics as independent variables and compaction degree as dependent variable.
3. The method according to claim 1, characterized in that, In step S2, the spatial coordinates of each acquisition point are recorded by the RTK positioning module, and the vibration data, moisture content data and spatial coordinates are uploaded to the edge computing terminal via the wireless transmission module; the interpolation calculation of the moisture content adopts the Kriging interpolation algorithm.
4. The method according to claim 1 or 3, characterized in that, The spatial distribution map of moisture content and the spatial distribution map of compaction are displayed in pseudo-color format, with qualified areas, areas above the upper limit and areas below the lower limit marked with different colors.
5. The method according to claim 1, characterized in that, In step S3, when the moisture content at the collection point is higher than the upper limit of the optimal moisture content range, a warning instruction to suspend compaction and allow drying is issued; when the moisture content at the collection point is lower than the lower limit of the optimal moisture content range, a warning instruction to supplement watering is issued.
6. The method according to claim 1, characterized in that, In S4, the feedforward control command includes: issuing a work stoppage warning to the roller operator and site management personnel; automatically calculating the required area of waterproof tarpaulin to be covered based on the current construction area, and generating tarpaulin laying path guidance.
7. The method according to claim 1 or 6, characterized in that, In step S4, after the rainfall ends, the moisture content distribution of the lime-soil layer is collected and evaluated again. Only when all the moisture content test values fall within the optimal moisture content range is an instruction to allow resumption of work issued.
8. The method according to claim 1, characterized in that, The feedback control command includes at least one of adjusting the vibration frequency of the road roller, adjusting the compaction speed, or increasing the number of compaction passes; the method further includes: dividing the construction area into multiple construction zones, storing the spatial distribution maps of moisture content, compaction degree, and compaction trajectory of each construction zone and each lime-soil layer in a spatiotemporal manner, generating a three-dimensional compaction degree digital twin map, wherein the three-dimensional compaction degree digital twin map supports traceability query of compaction quality by location and lime-soil layer.
9. A quality monitoring and feedback control system for layered compaction of lime-soil, used to implement the method as described in any one of claims 1 to 8, characterized in that, include: The sensing and detection group includes a vibration sensor and a moisture content sensor installed on the vibrating drum of the road roller, used to collect vibration data and moisture content data; The positioning and transmission group includes a positioning module and a wireless transmission module. The positioning module is used to record the spatial coordinates of each collection point, and the wireless transmission module is used to upload the collected data and spatial coordinates. The edge computing terminal, connected to the positioning and transmission group, is used to receive the data and generate a spatial distribution map of moisture content and a spatial distribution map of compaction, identify areas with excessive moisture content, and is equipped with a meteorological interface for obtaining short-term rainfall forecast data. And a display terminal connected to the edge computing terminal.
10. The system according to claim 9, characterized in that, The display terminal includes a cab display terminal installed in the cab of the road roller and a mobile terminal for the site management personnel; the edge computing terminal is connected to the cab display terminal and the mobile terminal for the site management personnel via a wireless network.
