Construction control method for in-situ permeation curing of road base and road base

Abstract

The application discloses a construction control method for in-situ permeation and solidification of a road base and a road base. The method realizes real-time sensing of resistivity distribution of each spatial position of an aggregate layer through an electrode array, combines a porosity prior distribution model, adopts Bayesian inversion to calculate an equivalent water-cement ratio deviation of each spatial grid to determine an actual permeation state of slurry of each grid, and further takes a square sum of deviations of each grid as an optimization target, takes spraying pressure, advancing speed and an online water-cement ratio as decision variables, obtains a differentiated optimal parameter combination of each grid through rolling optimization solution, and continuously executes closed-loop control in the spraying process, so that the slurry is effectively prevented from flowing and being blocked due to non-uniform distribution of porosity, the inhomogeneous natural stone slag can meet the construction requirements of the road base without complex pretreatment, the traditional road base construction is freed from the dependence on aggregate grading and centralized mixing, and the construction cost is significantly reduced.

Description

Technical Field

[0001] This invention belongs to the field of road construction technology, and more specifically, relates to a construction control method for in-situ infiltration and curing of road base courses and the road base course itself. Background Technology

[0002] The road base course is a crucial component of the pavement structure, playing a vital role in distributing loads and transferring stress. Traditional road base course construction typically employs cement-stabilized crushed stone or lime-fly ash stabilized crushed stone structures. The conventional approach involves centrally mixing cement, aggregates, and water according to the designed mix proportions at a mixing plant, then transporting the mixture to the construction site for paving and compaction. While this method is technically mature and widely used, it has inherent limitations. Firstly, the construction of the mixing plant and the transportation of the mixture significantly increase construction costs and energy consumption. Secondly, this process has extremely strict requirements on aggregate gradation, necessitating the use of high-quality crushed stone within a specific particle size range, thus limiting the use of low-grade materials.

[0003] Furthermore, with the continuous advancement of infrastructure construction, a large amount of natural stone waste is generated during various engineering constructions. This stone waste is usually centrally stockpiled or landfilled, which not only occupies valuable land resources, but the dust and leachate generated during the stockpiling process may also pollute the surrounding environment. However, due to the stringent requirements for aggregate gradation in traditional road base construction, this natural stone waste is difficult to use directly as aggregate, resulting in a long-term low level of resource utilization.

[0004] To address the aforementioned issues, in recent years, those skilled in the art have attempted to utilize waste soil and construction waste to prepare road base materials, thereby alleviating the pressure of waste disposal and reducing material costs. For example, patent CN116003080A discloses a recycled road base mixture based on solidified engineering waste soil and its application. The treatment method is simple and can realize the resource reuse of waste and the sustainable development of cities. However, it still relies on a centralized mixing mode, requiring complex pretreatment and curing processes; it has strict requirements for soil gradation, making it difficult to guarantee material uniformity; the curing cycle is long, and the construction cost is high. Patent CN116065435A discloses a treatment method for construction waste road base, which effectively improves the quality of road base made from construction waste, exhibiting excellent properties such as impermeability, frost resistance, and impact resistance. However, it requires the addition of multiple curing materials, resulting in high costs; the construction process is cumbersome and energy-intensive; strict requirements are placed on the particle size of construction waste; and the uniformity of the curing material distribution is difficult to control. Patent CN121248182A discloses a method for preparing road base curing agents using construction waste, which reduces landfill pollution and achieves resource recycling by consuming large amounts of construction waste; it uses polypropylene fiber and water-soluble resin materials in synergy to improve the strength, toughness, and durability of the curing agent. However, its curing agent preparation process is extremely complex, requiring more than nine kinds of raw materials and various specialized equipment; the production cycle is long and energy consumption is high; it is not suitable for natural stone chips; and the strength is difficult to control. Summary of the Invention

[0005] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a construction control method for in-situ permeation and curing of road base courses, as well as the road base course itself. This method uses an electrode array to sense the resistivity distribution at various spatial locations of the aggregate layer in real time. Combined with a priori porosity distribution model, Bayesian inversion is used to calculate the equivalent water-cement ratio deviation of each spatial grid to clarify the actual permeation state of the slurry in each grid. Then, minimizing the sum of squared deviations of each grid is used as the optimization objective. Spraying pressure, travel speed, and online water-cement ratio are used as decision variables. A rolling optimization solution is employed to obtain the differentiated optimal parameter combination for each grid. Closed-loop control is continuously executed during the spraying process, effectively solving the problems of slurry fingering and clogging caused by non-uniform porosity distribution. This allows heterogeneous natural stone chips to meet the requirements of road base course construction without complex pretreatment, eliminating the dependence on aggregate gradation and centralized mixing in traditional road base course construction and significantly reducing construction costs.

[0006] To achieve the above objectives, according to one aspect of the present invention, a construction control method for in-situ infiltration and curing of road base courses is provided, characterized by comprising the following steps: S100: Obtain the control benchmark dataset before construction. The control benchmark dataset includes the gradation data of natural stone chips, mud content data, target water-cement ratio, upper and lower limits of operating parameters, and preset deviation thresholds. Establish a priori porosity distribution model. S200: A subgrade layer is laid on the treated soil subgrade, and an aggregate layer is laid on the subgrade layer; an electrode array is arranged in the aggregate layer along the construction width direction, and a multi-channel resistivity response signal is obtained through the electrode array; the spatial distribution of resistivity of the aggregate layer is determined according to the resistivity response signal. S300: Combining the resistivity spatial distribution with the porosity a priori distribution model, obtain the porosity distribution data and equivalent water-cement ratio distribution data in each spatial grid of the aggregate layer, and calculate the equivalent water-cement ratio deviation data based on the equivalent water-cement ratio distribution data and the target water-cement ratio; S400: Establish an optimization objective function based on the equivalent water-cement ratio deviation data; use the optimization objective function and operating parameters to perform rolling optimization to obtain the optimal set of operating parameters; S500: Prepare cement slurry according to the optimal operating parameter set and spray it on the surface of the aggregate layer. Repeat steps S300 and S400 during the spraying process to obtain an updated operating parameter set and continue spraying until the equivalent water-cement ratio deviation data of each spatial grid is within the preset deviation threshold. S600: Standard maintenance shall be carried out on the road surface after the cement slurry spraying is completed.

[0007] Further, in step S100, the a priori porosity distribution model is established in the following manner: The gradation data and mud content data of the natural stone chips used in the current construction are obtained. A random stacking algorithm is used to generate a random stacking structure of multiple natural stone chip particles in a three-dimensional virtual space. The random stacking algorithm determines the number ratio of particles of each size according to the gradation data and simulates the natural stacking process of particles through random placement and gravity settling. Calculate the porosity and permeability tensors for each virtual packing structure; The prior distribution model is determined by the statistical distribution of the porosity and permeability tensors of all virtual structures.

[0008] Further, in step S200, the thickness of the subbase layer is 20cm~25cm. The thickness of the aggregate layer is 18cm~20cm; the aggregate layer is initially compacted after paving to maintain an initial porosity of 20%~35%.

[0009] Further, in step S200, the electrode array is multiple and is respectively located in the upper, middle and lower parts of the aggregate layer. The upper electrode array is located 3cm to 5cm below the surface of the aggregate layer, the middle electrode array is located at 1 / 2 of the thickness of the aggregate layer, and the lower electrode array is located 3cm to 5cm above the bottom of the aggregate layer. At least four rows of electrodes are arranged along the construction width direction at each depth location. Each row of electrodes contains at least eight pairs of electrodes. Each pair of electrodes alternates as an excitation electrode and a measurement electrode. Each pair of electrodes is evenly distributed along the construction width direction, and the distance between adjacent pairs of electrodes is 10cm to 20cm.

[0010] Further, in step S300, the acquisition of the porosity distribution data and the equivalent water-cement ratio distribution data specifically involves: Using the porosity a priori distribution model as a priori constraint and the resistivity spatial distribution as the observation input, Bayesian inversion is used to calculate point by point on the discrete spatial grid to obtain the porosity distribution data φᵢ and the equivalent water-cement ratio distribution data Wᵢ for each spatial grid, where i is the spatial grid number, i=1,2,…,M, M is the total number of spatial grids, M≥8; The equivalent water-cement ratio deviation data is the difference between the equivalent water-cement ratio and the target water-cement ratio for each spatial grid: Where W0 is the target water-cement ratio.

[0011] Further, in step S400, the optimization objective function is the sum of squares of the equivalent water-cement ratio deviations of each spatial grid: Where J is the objective function value and ΔWᵢ is the equivalent water-cement ratio deviation of the i-th spatial grid; The operating parameters include spraying pressure p1, travel speed p2, and online water-cement ratio p3; The rolling optimization solution is as follows: the optimization objective function J and the running parameters p1, p2, p3 are used as decision variables, and the upper and lower limits of the running parameters are used as constraints. Solve for the optimal control sequence that minimizes J in the prediction time domain, and output the optimal parameter combination data for each spatial grid according to the spatial grid index {p 1i p 2i p 3i}, where p 1i p is the optimal spray pressure value for the i-th spatial grid. 2i p is the optimal travel speed value for the i-th spatial grid. 3i Let be the optimal online water-cement ratio for the i-th spatial grid.

[0012] Furthermore, among the upper and lower limits of the operating parameters, the upper and lower limits of the spraying pressure p1 are 0.2 to 0.8 MPa, the upper and lower limits of the traveling speed p2 are 1 to 4 m / min, and the upper and lower limits of the online water-cement ratio p3 are 0.4 to 1.6.

[0013] Further, in step S500, repeating steps S300 and S400 to obtain an updated set of operating parameters and continuing spraying specifically involves: The multi-channel resistivity response signal at the current moment is continuously acquired through the electrode array. Step S300 is executed based on the current resistivity response signal to obtain the equivalent water-cement ratio deviation data of each spatial grid. Then, step S400 is executed to obtain the updated optimal operating parameter set of each spatial grid. The spraying pressure and travel speed values ​​from the updated optimal operating parameter set are output in real time to adjust the spraying status; if the online water-cement ratio value in the updated optimal operating parameter set is inconsistent with the water-cement ratio of the currently prepared cement slurry, spraying is stopped and the cement slurry is re-prepared, and then spraying continues according to the updated parameter set.

[0014] According to a second aspect of the present invention, a road base course is provided, which is prepared by applying the aforementioned in-situ penetration and curing road base course construction control method.

[0015] Furthermore, the road base layer comprises, from bottom to top: an earthen roadbed, a subbase laid on the earthen roadbed, and an aggregate layer located on the subbase. The aggregate layer is cured by cement slurry penetration to form a cured layer.

[0016] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: 1. This invention uses an electrode array to sense the resistivity distribution at various spatial locations in real time, combines Bayesian inversion to calculate the equivalent water-cement ratio deviation, and uses a rolling optimization solution for differentiated control. This replaces the traditional construction mode of fixed-parameter spraying combined with post-drilling core sampling. It solves the problems of fingering and clogging caused by the inability to know the slurry penetration status in real time in existing in-situ infiltration curing construction. At the same time, it avoids the high cost burden caused by the stringent requirements for aggregate gradation and mud content and complex pretreatment in traditional centralized mixing schemes. It significantly improves the uniformity of road base curing quality and construction efficiency, and reduces material pretreatment and rework costs.

[0017] 2. This invention discretizes the aggregate layer into spatial grids and uses the minimum sum of squares of the equivalent water-cement ratio deviations of each grid as the optimization objective for rolling optimization. This enables independent, real-time, and precise control of spraying pressure, travel speed, and online water-cement ratio at each spatial location. It solves the problem of mismatch between slurry distribution and pore characteristics caused by using uniform spraying parameters at each spatial location in traditional methods. This achieves the effect of making the equivalent water-cement ratio of each spatial grid approach the target water-cement ratio and achieving uniform penetration across the entire cross-section.

[0018] 3. This invention constructs a closed-loop control link that integrates real-time sensing by an electrode array, Bayesian inversion calculation, rolling optimization decision-making, and dynamic adjustment of the actuator. This ensures that quality control is integrated throughout the entire construction process rather than relying on post-construction testing. It can effectively guarantee the solidification quality of natural stone chips under various gradation conditions, enabling even heterogeneous natural stone chips with gradation deviations from the ideal curve and mud content exceeding conventional aggregate control standards to meet the requirements of road base construction, thus achieving efficient resource utilization of engineering stone chips. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the construction control method for in-situ penetration and curing of road base layers according to the present invention; Figure 2 This is a schematic diagram of the spraying and penetration curing process in the construction control method for in-situ penetration curing of road base layers of the present invention; Figure 3 This is a schematic diagram of the construction structure of the road base layer in the construction control method of in-situ permeation and curing road base layer of the present invention.

[0020] Figure labels: 1-Cement slurry solidification layer; 2-Natural stone slag cushion layer; 3-Earth subgrade; 4-Natural stone slag; 5-Cement slurry penetration area; 6-Cement bonded interlocking structure; 7-Spraying equipment; 8-Cement slurry; 9-Cement slurry penetration direction; 10-Electrode array; 11-Control unit. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0022] This invention provides a construction control method for in-situ infiltration and curing of road base courses and a road base course itself. The road base course, from bottom to top, comprises: an earthen subgrade 3, a cushion layer 2 laid on the earthen subgrade, and an aggregate layer located on the cushion layer. The aggregate layer, after being infiltrated and cured by cement grout 8, forms a cured layer 1. The construction control method flow is as follows: Figure 1 As shown, it includes the following steps: S100: Obtain the control baseline dataset before construction. The control baseline dataset includes the gradation data, mud content data, target water-cement ratio, upper and lower limits of operating parameters, and preset deviation thresholds of natural stone slag 4. Establish a porosity a priori distribution model. The porosity a priori distribution model is established in the following way: Obtain the gradation data and mud content data of the natural stone chips 4 used in the current construction, and use a random stacking algorithm to generate a random stacking structure of multiple natural stone chip particles in a three-dimensional virtual space. The random stacking algorithm determines the number ratio of particles of each particle size according to the gradation data, and simulates the natural stacking process of particles through random placement and gravity settling. Calculate the porosity and permeability tensors for each virtual packing structure; The prior distribution model is determined by the statistical distribution of the porosity and permeability tensors of all virtual structures. S200: A cushion layer 2 is laid on the treated subgrade 3, and an aggregate layer is laid on the cushion layer; an electrode array 10 is arranged in the aggregate layer along the construction width direction, and a multi-channel resistivity response signal is obtained through the electrode array 10, and the spatial distribution of resistivity of the aggregate layer is determined according to the resistivity response signal; the thickness of the cushion layer is 20cm~25cm. The aggregate layer has a thickness of 18cm to 20cm. After paving, the aggregate layer is initially compacted to maintain an initial porosity of 20% to 35%. Multiple electrode arrays are arranged at the upper, middle, and lower parts of the aggregate layer. The upper electrode array is located 3cm to 5cm below the surface of the aggregate layer, the middle electrode array is located at half the thickness of the aggregate layer, and the lower electrode array is located 3cm to 5cm above the bottom of the aggregate layer. At least four rows of electrode arrays are arranged along the construction width direction at each depth location. Each row of electrode arrays contains at least eight electrode pairs, with each electrode pair alternately serving as both an excitation electrode and a measurement electrode. The electrode pairs are evenly spaced along the construction width direction, with a spacing of 10cm to 20cm between adjacent electrode pairs. S300: Combining the resistivity spatial distribution with the porosity a priori distribution model, obtain the porosity distribution data and equivalent water-cement ratio distribution data in each spatial grid of the aggregate layer. Calculate the equivalent water-cement ratio deviation data based on the equivalent water-cement ratio distribution data and the target water-cement ratio. The acquisition of the porosity distribution data and equivalent water-cement ratio distribution data specifically involves: Using the porosity a priori distribution model as a priori constraint and the resistivity spatial distribution as the observation input, Bayesian inversion is used to calculate point by point on the discrete spatial grid to obtain the porosity distribution data φᵢ and the equivalent water-cement ratio distribution data Wᵢ for each spatial grid, where i is the spatial grid number, i=1,2,…,M, M is the total number of spatial grids, M≥8; The equivalent water-cement ratio deviation data is the difference between the equivalent water-cement ratio and the target water-cement ratio for each spatial grid: Where W0 is the target water-cement ratio; S400: Establish an optimization objective function based on the equivalent water-cement ratio deviation data; perform rolling optimization using the optimization objective function and operating parameters to obtain the optimal set of operating parameters; the optimization objective function is the sum of squares of the equivalent water-cement ratio deviations of each spatial grid. Where J is the objective function value, ΔWᵢ is the equivalent water-cement ratio deviation of the i-th spatial grid, i is the spatial grid number, i=1,2,…,M, and M is the total number of spatial grids; The operating parameters include spraying pressure p1, travel speed p2, and online water-cement ratio p3; The rolling optimization solution is as follows: the optimization objective function J and the running parameters p1, p2, p3 are used as decision variables, and the upper and lower limits of the running parameters are used as constraints. Solve for the optimal control sequence that minimizes J in the prediction time domain, and output the optimal parameter combination data for each spatial grid according to the spatial grid index {p 1i p 2i p 3i}, where p 1i p is the optimal spray pressure value for the i-th spatial grid. 2i p is the optimal travel speed value for the i-th spatial grid. 3i Let be the optimal online water-cement ratio for the i-th spatial grid.

[0023] Among the upper and lower limits of the operating parameters, the upper and lower limits of the spraying pressure p1 are 0.2 to 0.8 MPa, the upper and lower limits of the traveling speed p2 are 1 to 4 m / min, and the upper and lower limits of the online water-cement ratio p3 are 0.4 to 1.6. S500: Prepare cement slurry 8 according to the optimal operating parameter set and spray it onto the surface of the aggregate layer. During the spraying process, repeat steps S300 and S400 to obtain an updated operating parameter set and continue spraying until the equivalent water-cement ratio deviation data of each spatial grid is within the preset deviation threshold. The specific steps of repeating steps S300 and S400 to obtain an updated operating parameter set and continue spraying are as follows: The multi-channel resistivity response signal at the current moment is continuously acquired through the electrode array 10. Based on the current resistivity response signal, step S300 is executed to obtain the equivalent water-cement ratio deviation data of each spatial grid. Then, step S400 is executed to obtain the updated optimal operating parameter set of each spatial grid. The spraying pressure and travel speed values ​​in the updated optimal operating parameter set are output in real time to adjust the spraying status; if the online water-cement ratio value in the updated optimal operating parameter set is inconsistent with the water-cement ratio of the currently prepared cement slurry 8, the spraying is stopped and the cement slurry 8 is re-prepared, and then the spraying continues according to the updated parameter set. S600: Standard maintenance shall be carried out on the road surface after the cement slurry spraying is completed.

[0024] This invention acquires the spatial resistivity distribution of the aggregate layer in real time at various spatial locations using an electrode array. Combined with a priori porosity distribution model, Bayesian inversion is used to calculate the equivalent water-cement ratio deviation of each spatial grid. Then, an optimization objective function is established with the goal of minimizing the sum of squared deviations, and rolling optimization is used to obtain the optimal operating parameter set for each spatial grid. This effectively enables independent, real-time, and precise control of spraying pressure, travel speed, and online water-cement ratio at each spatial location. This solves the problem of uneven grout flow along large pores caused by the non-uniform spatial distribution of porosity in natural stone chips during traditional "aggregate first, grout later" in-situ permeation curing construction. The present invention addresses the technical bottlenecks of flow and blockage in small pore areas. While existing centralized mixing solutions attempt to improve material uniformity through pretreatment, they remain limited by reliance on mixing plants and stringent gradation requirements. This invention, starting from the fundamental level of construction control logic, transforms quality control from "relying on material pre-uniformity" to "achieving uniform curing through process regulation." This fundamentally avoids the stringent requirements on aggregate gradation and mud content, significantly reducing material pretreatment costs and construction energy consumption. Simultaneously, through a closed-loop control link, it reduces the standard deviation of the cured layer strength by more than 50%, significantly improving the uniformity and controllability of road base construction quality.

[0025] The present invention will be described below with reference to specific embodiments.

[0026] This embodiment takes a test section of a highway as an example. The natural stone aggregate has normal gradation and mud content between 5% and 8%. The test section is located on a newly built highway with a total length of 100m, a road width of 12m, three lanes in one direction, a road base design thickness of 20cm, and a design strength of 4.0MPa (28d unconfined compressive strength).

[0027] The natural stone aggregate was sourced from excavated spoil from a tunnel near the test section, and its lithology was limestone. Screening and mud content tests were conducted according to the relevant provisions of the "Specifications for Testing Aggregates in Highway Engineering". The screening test employed a dry sieving method, passing the natural stone aggregate sequentially through square-hole sieves of 40mm, 20mm, 10mm, 5mm, 2.5mm, and 0.075mm to obtain the gradation curve. The gradation data for the natural stone aggregate used in this embodiment are as follows: particle size > 40mm: 15%; particle size 20mm~40mm: 30%; particle size 5mm~20mm: 35%; particle size < 5mm: 20%; with a mud content (<0.075mm) of 6%. P·O 42.5 grade ordinary Portland cement was used.

[0028] A control baseline dataset is acquired before construction, and a priori porosity distribution model is established. Before construction, the gradation data and mud content data of natural stone chips, the target water-cement ratio W0, the upper and lower limits of operating parameters, and the preset deviation threshold ε are entered into the control unit. The control unit 11 is a vehicle-mounted intelligent control device installed on the spraying equipment 7, which integrates a data acquisition module, a Bayesian inversion calculation module, a rolling optimization solution module, and an equipment control module. It is used to receive the resistivity signal of the electrode array 10, perform inversion and optimization calculations, and output operating parameter control commands.

[0029] The target water-cement ratio W0 was determined through indoor mix proportion tests before construction, as follows: Collect natural stone chips from the same source used at the construction site and prepare aggregate samples according to the above gradation. Compact the aggregate samples in standard molds to prepare multiple sets of parallel specimens. Prepare cement slurry with water-cement ratios of 0.6, 0.8, 1.0, 1.2, and 1.4 respectively. The cement slurry is prepared by mixing P·O 42.5 cement with water and sprayed onto the surface of each specimen to allow natural penetration. After curing under standard conditions (temperature 20±2℃, relative humidity ≥95%) for 28 days, conduct unconfined compressive strength tests.

[0030] The test results show that the strength is 3.2 MPa when the water-cement ratio is 0.6, 4.0 MPa when it is 0.8, 4.6 MPa when it is 1.0, 4.2 MPa when it is 1.2, and 3.5 MPa when it is 1.4. The design strength requirement is 4.0 MPa, and W0=1.0 is determined as the target water-cement ratio.

[0031] The upper and lower limits of the operating parameters are determined as follows: (1) The spraying pressure p1 ranges from 0.2 to 0.8 MPa. The lower limit p1 is... 1min =0.2MPa is determined by the minimum stable operating pressure of the variable frequency pump in the spraying equipment; the upper limit value p 1max =0.8MPa is determined by the rated working pressure of the equipment and the erosion resistance of the aggregate layer surface. If the pressure is higher than this value, the aggregate layer surface will be washed away.

[0032] (2) The range of the travel speed p2 is 1 to 4 m / min. The lower limit p2 is... 2min =1m / min is determined by construction efficiency requirements; upper limit value p 2max =4m / min is determined by the requirement for uniformity of spraying coverage. If the value is higher than this, the slurry sprayed in a single spray cannot evenly cover the surface of the aggregate layer.

[0033] (3) The online water-cement ratio p3 ranges from 0.4 to 1.6. The lower limit p... 3min=0.4 is determined by the pumpability requirements of the slurry. Below this value, the slurry viscosity is too high and cannot pass through the pipelines and nozzles of the spraying equipment; the upper limit value p 3max =1.6 is determined by the curing strength requirement. When it is higher than this value, the cement particles settle faster, the amount of effective cement that actually enters the pores decreases, and the strength decreases after curing.

[0034] The preset deviation threshold ε is determined in the following way: Using natural stone chips from the same source as those used on the construction site, multiple sets of specimens were prepared according to the construction mix proportion. For each set of specimens, deviations of the equivalent water-cement ratio relative to the target water-cement ratio W0 were set to 0, ±0.01, ±0.02, ±0.03, ±0.04, and ±0.05, respectively. After standard curing, the 28-day unconfined compressive strength of each set of specimens was measured, establishing the correlation between equivalent water-cement ratio deviation and strength.

[0035] The test results show that when |ΔW|≤0.03, the strength of each specimen meets the design strength requirements; when |ΔW|>0.03, the strength of some specimens begins to fall below the design strength. Therefore, the absolute value of the maximum equivalent water-cement ratio deviation that can meet the design strength requirements is taken as the preset deviation threshold, and ε=0.03 is selected.

[0036] Based on the aforementioned gradation data and mud content data, a priori porosity distribution model is established using a stochastic packing algorithm. The specific steps are as follows: (1) Based on the gradation data obtained from the sieve analysis, determine the proportion of particles in each particle size range. Let m be the mass proportion of the j-th particle size range, and generate the corresponding number of spherical particles in the three-dimensional virtual space according to this proportion. The diameter of each particle is randomly selected within its corresponding particle size range to simulate the actual size distribution characteristics of natural stone chips.

[0037] (2) A random drop algorithm is used to drop each particle one by one into a three-dimensional virtual space with dimensions of 1m×1m×1m. During the drop process, the particles fall from random positions above the space and are deposited at the bottom of the space or on the surface of already accumulated particles under simulated gravity. The final accumulation position of the particles is determined by their falling trajectory and the contact and collision with already accumulated particles. By controlling the amount of overlap between particles, the geometrically reasonable accumulation structure is ensured.

[0038] (3) Calculate the volumetric porosity for each virtual packing structure. The porosity η is calculated using the following formula: η = 1 - (total particle volume / virtual packing space volume). Simultaneously, calculate the permeability tensor K by simulating fluid flow in the virtual pore channels. The calculation formula is as follows: Where Q is the volumetric flow rate (m³ / s), μ is the fluid dynamic viscosity (Pa·s), L is the flow direction length (m), A is the cross-sectional area (m²), and ΔP is the applied pressure difference (Pa).

[0039] (4) Repeat the above calculations for 300 virtual packing structures to obtain 300 sets of porosity data and corresponding permeability data. Statistical analysis of the 300 generated virtual packing structures revealed that when the porosity varied within the range of 20% to 36%, the permeability K remained stable at the order of 10. -8 ~10 -10 Within the range of m², under this gradation condition, the porosity of natural stone chips in a randomly loosely packed state exhibits an approximately normal distribution, with a mean of 28%, a standard deviation of 4%, and a 95% confidence interval of 20%–36%.

[0040] The subgrade is constructed in layers, with each layer controlled to a thickness of 30-40 cm. A 22t vibratory roller is used for compaction, with a rolling speed controlled at 2-4 km / h and 6-8 passes, until a compaction degree of 95% is achieved. After compaction, the compaction degree is tested using the sand cone method; areas that fail to meet the standard must be re-compacted. After the subgrade treatment is completed, a small amount of water is sprayed on the surface to retain moisture and prevent drying and cracking.

[0041] like Figure 2 As shown, a natural stone slag layer 2 is laid on the treated subgrade 3. The thickness of the natural stone slag layer 2 is 20cm~25cm, preferably 22cm. A bulldozer is used for rough leveling, followed by a grader for fine leveling to ensure the surface flatness meets the requirements. A vibratory roller is used for compaction, achieving a compaction degree of over 93%.

[0042] A natural stone aggregate layer is laid on top of the natural stone crushed stone subbase 2. After the aggregate layer is penetrated and cured by cement slurry, a cement slurry cured layer 1 is formed. The thickness of the natural stone aggregate layer is 18cm~20cm, preferably 20cm, to ensure uniform thickness. A bulldozer or grader is used for laying. During the laying process, the thickness is checked using a steel rod insertion method, and the thickness deviation is controlled within ±2cm. After laying, a 12t vibratory roller is used for preliminary compaction: one pass of static compaction and one pass of weak vibration, to achieve a preliminary compaction of the aggregate while maintaining an initial porosity of 20%~35%. The roller speed is controlled at 2~3km / h during preliminary compaction. Field testing shows that the initial porosity of the aggregate layer in this embodiment is approximately 26%~32%, with an average of 29%, which meets the construction requirements.

[0043] The specific method for arranging the electrode array in the aggregate layer is as follows: (1) Depth of arrangement: The electrode array is set at three depths: the upper, middle and lower parts of the aggregate layer. The upper electrode array is set about 5 cm below the top of the aggregate layer; the middle electrode array is set about 1 / 2 of the aggregate layer thickness and about 10 cm from the top; the lower electrode array is set about 5 cm above the bottom of the aggregate layer and about 15 cm from the top.

[0044] (2) Arrangement direction: At least 4 rows of electrodes are arranged along the construction width direction at each depth position. Each row of electrodes contains at least 8 pairs of electrodes. Each pair of electrodes is used alternately as excitation electrode and measurement electrode. Each pair of electrodes is distributed at equal intervals along the construction width direction, and the distance between adjacent pairs of electrodes is 15cm.

[0045] (3) Electrode structure: Each electrode pair consists of two electrodes, and each electrode pair works alternately as an excitation electrode and a measurement electrode. The excitation electrode applies an AC signal of a preset frequency of 50Hz to the aggregate layer, and the measurement electrode collects the response signal.

[0046] Before spraying the cement slurry, the initial resistivity of each channel is collected using an electrode array to establish an initial electrical reference model for the aggregate layer. Specifically, the control unit sequentially applies a 50Hz AC signal to each electrode pair, measures the potential difference between adjacent electrode pairs, and calculates the initial resistivity value of each channel according to Ohm's law. For the four-electrode method, the apparent resistivity ρ = K·(ΔU / I), where K is the electrode arrangement coefficient (determined by the electrode spacing and geometry; in this embodiment, the calibrated value is 0.15m), ΔU is the potential difference between the measuring electrodes, and I is the excitation current intensity. A total of 168 channels (3 (depth layer) × 8 (width position) × 7 (adjacent electrode pair combinations) of initial resistivity data are collected.

[0047] The aggregate layer is spatially divided into discrete grids, as follows: Along the direction of travel (length direction): Divide 100m into 100 intervals, each interval is 1m long, and numbered m=1,2,…,100; Along the width direction: Divide 12m into 8 intervals, each interval being 1.5m wide, corresponding to the positions of 8 pairs of electrodes, numbered n=1,2,…,8; Along the depth direction: divided into 3 layers, corresponding to the upper, middle and lower rows of electrodes, numbered s=1,2,3; The total number of spatial grids is M = 100 × 8 × 3 = 2400, and each grid is numbered according to the three-dimensional index (m, n, s).

[0048] After the cement slurry spraying begins, the control unit collects the resistivity response signal of each channel at the current moment through the electrode array, with a control step size (Δt = 5 seconds in this embodiment) as the cycle. Detection shows that as the slurry penetrates further, the real-time resistivity of each channel dynamically changes within the range of 400~1000 Ω·m.

[0049] The control unit uses the porosity a priori distribution model as a priori constraint and the real-time resistivity spatial distribution at the current moment as the observation input. It performs Bayesian inversion point by point on the discrete spatial grid to obtain the porosity distribution data φᵢ and the equivalent water-cement ratio distribution data Wᵢ for each spatial grid.

[0050] During the spraying process, the inversion results showed that: in the early stage of spraying, the Wᵢ value of some grids was lower than the target value (such as 0.6~0.8), indicating that the slurry had not fully penetrated in these areas; in the middle stage of spraying, the Wᵢ of each grid gradually approached 1.0; in the end of spraying, the Wᵢ of most grids was close to 1.0, and the deviation was less than ε.

[0051] Based on the equivalent water-cement ratio distribution data Wᵢ of each spatial grid and the target water-cement ratio W0=1.0, the equivalent water-cement ratio deviation data of each spatial grid is calculated: ΔWᵢ=Wᵢ−W0=Wᵢ−1.0. The initial value of ΔWᵢ is distributed in the range of -0.4~0.6, and gradually converges to within ±0.03 as spraying progresses.

[0052] The control unit uses the equivalent water-cement ratio deviation data ΔWᵢ of each spatial grid as input to establish the following optimization objective function: P1: Spraying pressure (MPa), with the constraint 0.2≤p1≤0.8; P2: Travel speed (m / min), with the constraint 1≤p2≤4; P3: Online water-cement ratio, with the constraint 0.4≤p3≤1.6.

[0053] The control unit uses the equivalent water-cement ratio deviation data ΔWᵢ of each spatial grid as input to establish an optimization objective function: minimizing the sum of squares of the equivalent water-cement ratio deviations of all 2400 spatial grids. Simultaneously, constraints are set for three decision variables: spraying pressure p1 is controlled between 0.2 MPa and 0.8 MPa, travel speed p2 is controlled between 1 m / min and 4 m / min, and online water-cement ratio p3 is controlled between 0.4 and 1.6.

[0054] Under the above constraints, the control unit in the prediction time domain T pWithin 60 seconds, using the equivalent water-cement ratio deviation ΔWᵢ(t) at the current time t as the initial state, and decision variables p1, p2, and p3 as control inputs, solve for the optimal control sequence that minimizes the objective function. After solving, only the first control variable in the optimal control sequence (i.e., the optimal parameter combination from time t to t+Δt) is output and executed. At the next time t+Δt, the optimization problem is resolved with the updated ΔWᵢ, achieving "rolling" optimization.

[0055] In this embodiment, the computation time for a single optimization solution is approximately 0.1~0.2 seconds, far less than the control step size of 5 seconds, fully meeting the real-time requirements. The optimization solver outputs the optimal operating parameter set {p} for each spatial grid. 1i p 2i p 3i}, where p 1i p is the optimal spray pressure value for the i-th grid. 2i p is the optimal travel speed value for the i-th grid. 3i Let be the optimal online water-cement ratio for the i-th grid.

[0056] Prepare cement slurry according to the target water-cement ratio W0=1.0. Add cement and water to the mixing tank at a mass ratio of 1:1, and mix for 3-5 minutes using a high-speed mixer at a speed of 1000-1500 rpm. After filtering the prepared cement slurry through a sieve, it is loaded into the slurry storage tank of the spraying equipment.

[0057] The spraying equipment was configured according to the initial optimal operating parameter set: spraying pressure p1 = 0.5 MPa, travel speed p2 = 2.5 m / min, and online water-cement ratio p3 = 1.0. The spraying equipment was started to spray cement slurry onto the surface of the aggregate layer.

[0058] The spraying equipment is equipped with multiple sets of nozzles, with a spraying width of 12m and a nozzle height of 30~50cm from the surface of the aggregate layer, ensuring uniform spraying coverage without any missed spots or excessive overlap.

[0059] During spraying, the control unit automatically executes the following closed-loop control cycle every 5 seconds: (1) Obtain the multi-channel resistivity response signal of each channel at the current moment through the electrode array, and update the resistivity spatial distribution data.

[0060] (2) Using the current resistivity spatial distribution as the observation input and the porosity prior distribution model as the prior constraint, Bayesian inversion is performed point by point on the discrete spatial grid to obtain the porosity distribution data and equivalent water-cement ratio distribution data of each spatial grid at the current time, and then the equivalent water-cement ratio deviation data ΔWᵢ of each spatial grid is calculated.

[0061] (3) Using the equivalent water-cement ratio deviation data ΔWᵢ of each spatial grid as input, establish an optimization objective function, resolve the optimal control sequence in the prediction time domain, and obtain the updated optimal operating parameter set {p} for each spatial grid. 1i p 2i p 3i}

[0062] (4) The spray pressure value p in the updated optimal operating parameter set 1i and travel speed value p 2i The system outputs data in real time to the variable frequency pump and the walking motor to adjust the spraying status.

[0063] (5) Check the online water-cement ratio p in the updated optimal operating parameter set. 3i Check if the water-cement ratio matches the currently prepared cement slurry. If it does, continue spraying; if it does not match, and the water-cement ratio needs to be adjusted in some areas, stop spraying, prepare the cement slurry again according to the new water-cement ratio, and then continue spraying according to the updated parameter set.

[0064] (6) Repeat steps (1) to (5) until the equivalent water-cement ratio deviation data ΔWᵢ of all spatial grids satisfies |ΔWᵢ|<0.03. Then the control unit determines that the spraying is complete and automatically stops spraying.

[0065] like Figure 2 and Figure 3 As shown, after the cement slurry 8 is sprayed onto the surface of the aggregate layer, it naturally penetrates downwards along the penetration direction 9 under the action of gravity, forming a cement slurry penetration area 5 between the aggregate particles, and finally forming a cement bonded interlocking structure 6.

[0066] After spraying, cover the surface of the solidified layer with geotextile and maintain moisture. During the curing period, keep the surface moist by regularly sprinkling water according to the temperature. The curing period is 28 days.

[0067] After the curing period, the quality of the cured layer was tested using core sampling. Nine core sampling points were evenly arranged along the length of the test section, with one core sampling point every 10m. A core sample with a diameter of 100mm was drilled from each core sampling point.

[0068] Core sample visual inspection Each core sample was intact and dense, with no obvious cracks or looseness, and the aggregate and cement paste were well bonded.

[0069] Unconfined compressive strength test Unconfined compressive strength tests were conducted on each core sample after 28 days. The test results are as follows: the core sample strength values ​​were 4.3 MPa, 4.6 MPa, 4.2 MPa, 4.8 MPa, 4.4 MPa, 4.5 MPa, 4.1 MPa, 4.7 MPa, and 4.5 MPa, with a mean of 4.5 MPa, a standard deviation of 0.4 MPa, and a coefficient of variation of 8.9%. All core sample strengths met the design strength requirement of 4.0 MPa, and the strength dispersion was small, indicating good curing uniformity.

[0070] To verify the technical effectiveness of this invention, a control example was set up in an adjacent section of the test section. The control example was 100m long and used the same natural stone chips, cement, design thickness, and design strength as the test section. The control group used a traditional fixed-parameter spraying method: a fixed water-cement ratio of 1.0, a fixed spraying pressure of 0.5MPa, and a fixed travel speed of 2.5m / min, without real-time monitoring or parameter adjustments. After the curing period, the control group also underwent core sampling and testing.

[0071] The control group test results showed that the strengths of the nine core samples were 3.2 MPa, 4.5 MPa, 3.8 MPa, 5.1 MPa, 3.5 MPa, 4.8 MPa, 3.3 MPa, 4.9 MPa, and 3.6 MPa, with a mean of 4.1 MPa, a standard deviation of 0.9 MPa, and a coefficient of variation of 22.0%. Although the mean strength met the design strength requirements, the strength of some core samples was lower than that of others, indicating that the slurry penetration was significantly uneven and that there were fingering and clogging phenomena.

[0072] Comparative data show that the present invention reduces the standard deviation of strength from 0.9 MPa to 0.4 MPa, a reduction of approximately 55%; and reduces the coefficient of variation from 22.0% to 8.9%, significantly improving curing uniformity. Simultaneously, the total slurry consumption in the present invention group was 1120 L, while the control group was 1350 L, resulting in a slurry consumption saving of approximately 17%. This indicates that the present invention, while ensuring curing quality, avoids excessive waste of slurry in the finger-flow area.

[0073] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A construction control method for in-situ permeation and curing of road base courses, characterized in that, Includes the following steps: S100: Obtain the control benchmark dataset before construction. The control benchmark dataset includes the gradation data of natural stone chips, mud content data, target water-cement ratio, upper and lower limits of operating parameters, and preset deviation thresholds. Establish a priori porosity distribution model. S200: A subgrade layer is laid on the treated soil subgrade, and an aggregate layer is laid on the subgrade layer; an electrode array is arranged in the aggregate layer along the construction width direction, and a multi-channel resistivity response signal is obtained through the electrode array; the spatial distribution of resistivity of the aggregate layer is determined according to the resistivity response signal. S300: Combining the resistivity spatial distribution with the porosity a priori distribution model, obtain the porosity distribution data and equivalent water-cement ratio distribution data in each spatial grid of the aggregate layer, and calculate the equivalent water-cement ratio deviation data based on the equivalent water-cement ratio distribution data and the target water-cement ratio; S400: Establish an optimization objective function based on the equivalent water-cement ratio deviation data; use the optimization objective function and operating parameters to perform rolling optimization to obtain the optimal set of operating parameters; S5 00: Prepare cement slurry according to the optimal operating parameter set and spray it on the surface of the aggregate layer. Repeat steps S300 and S400 during the spraying process to obtain an updated operating parameter set and continue spraying until the equivalent water-cement ratio deviation data of each spatial grid is within the preset deviation threshold. S600: Standard maintenance shall be carried out on the road surface after the cement slurry spraying is completed.

2. The construction control method for in-situ permeation and curing road base course according to claim 1, characterized in that, In step S100, the a priori porosity distribution model is established in the following manner: The gradation data and mud content data of the natural stone chips used in the current construction are obtained. A random stacking algorithm is used to generate a random stacking structure of multiple natural stone chip particles in a three-dimensional virtual space. The random stacking algorithm determines the number ratio of particles of each size according to the gradation data and simulates the natural stacking process of particles through random placement and gravity settling. Calculate the porosity and permeability tensors for each virtual packing structure; The prior distribution model is determined by the statistical distribution of the porosity and permeability tensors of all virtual structures.

3. The construction control method for in-situ permeation and curing of road base courses according to claim 1 or 2, characterized in that, In step S200, the thickness of the subbase layer is 20cm~25cm; the thickness of the aggregate layer is 18cm~20cm; the aggregate layer is initially compacted after paving to maintain an initial porosity of 20%~35%.

4. The construction control method for in-situ permeation and curing of road base courses according to claim 1 or 2, characterized in that, In step S200, the electrode array consists of multiple electrodes, which are respectively located in the upper, middle and lower parts of the aggregate layer. The upper electrode array is located 3cm to 5cm below the surface of the aggregate layer, the middle electrode array is located at 1 / 2 of the thickness of the aggregate layer, and the lower electrode array is located 3cm to 5cm above the bottom of the aggregate layer. At least four rows of electrodes are arranged along the construction width direction at each depth location. Each row of electrodes contains at least eight pairs of electrodes. Each pair of electrodes alternates as an excitation electrode and a measurement electrode. Each pair of electrodes is distributed at equal intervals along the construction width direction, and the distance between adjacent pairs of electrodes is 10cm to 20cm.

5. A construction control method for in-situ permeation and curing of road base courses according to claim 1 or 2, characterized in that, In step S300, the acquisition of the porosity distribution data and the equivalent water-cement ratio distribution data specifically involves: Using the porosity a priori distribution model as a priori constraint and the resistivity spatial distribution as the observation input, Bayesian inversion is used to calculate point by point on the discrete spatial grid to obtain the porosity distribution data φᵢ and the equivalent water-cement ratio distribution data Wᵢ for each spatial grid, where i is the spatial grid number, i=1,2,…,M, M is the total number of spatial grids, M≥8; The equivalent water-cement ratio deviation data is the difference between the equivalent water-cement ratio and the target water-cement ratio for each spatial grid: Where W0 is the target water-cement ratio.

6. A construction control method for in-situ permeation and curing of road base courses according to claim 1 or 2, characterized in that, In step S400, the optimization objective function is the sum of squares of the equivalent water-cement ratio deviations of each spatial grid: Where J is the objective function value and ΔWᵢ is the equivalent water-cement ratio deviation of the i-th spatial grid; The operating parameters include spraying pressure p1, travel speed p2, and online water-cement ratio p3; The rolling optimization solution is as follows: the optimization objective function J and the running parameters p1, p2, p3 are used as decision variables, and the upper and lower limits of the running parameters are used as constraints. Solve for the optimal control sequence that minimizes J in the prediction time domain, and output the optimal parameter combination data for each spatial grid according to the spatial grid index {p 1i p 2i p 3i }, where p 1i p is the optimal spray pressure value for the i-th spatial grid. 2i p is the optimal travel speed value for the i-th spatial grid. 3i Let be the optimal online water-cement ratio for the i-th spatial grid.

7. The construction control method for in-situ permeation and curing road base course according to claim 6, characterized in that, Among the upper and lower limits of the operating parameters, the upper and lower limits of the spraying pressure p1 are 0.2 to 0.8 MPa, the upper and lower limits of the traveling speed p2 are 1 to 4 m / min, and the upper and lower limits of the online water-cement ratio p3 are 0.4 to 1.

6.

8. A construction control method for in-situ permeation and curing of road base courses according to claim 1 or 2, characterized in that, In step S500, repeating steps S300 and S400 to obtain an updated set of operating parameters and continuing spraying specifically involves: The multi-channel resistivity response signal at the current moment is continuously acquired through the electrode array. Based on the current resistivity response signal, step S300 is executed to obtain the equivalent water-cement ratio deviation data of each spatial grid. Then, step S400 is executed to obtain the updated optimal operating parameter set of each spatial grid. The spraying pressure and travel speed values ​​from the updated optimal operating parameter set are output in real time to adjust the spraying status; if the online water-cement ratio value in the updated optimal operating parameter set is inconsistent with the water-cement ratio of the currently prepared cement slurry, spraying is stopped and the cement slurry is re-prepared, and then spraying continues according to the updated parameter set.

9. A road base course, characterized in that, It is prepared using the construction control method for in-situ penetration and curing of road base layer as described in any one of claims 1-8.

10. A road base course according to claim 9, characterized in that, The road base layer, from bottom to top, includes: an earthen roadbed, a subbase laid on the earthen roadbed, and an aggregate layer on the subbase. The aggregate layer is cured by cement slurry penetration to form a cured layer.

Images