Intelligent monitoring and control method for water-stable base crack

Abstract

The application provides a water-stable base crack intelligent monitoring control method, belongs to the field of road engineering construction monitoring and control, and is used for solving the problem that each control element is independent in the water-stable base crack control method in the related art, and global optimization and dynamic adjustment cannot be realized. The method unifies the moisture field, the stress field, the damage field and the construction control field into an optimization framework by constructing a scalar value functional representing the deviation of the system from the crack-free state, solves the functional minimum value to obtain the optimal control field and generates the construction instruction. At the same time, the functional parameters are updated online based on the measured data of the intelligent aggregate to realize adaptive calibration, and the contribution of each field quantity to the crack risk is calculated through the variational derivative to realize the explainability analysis. The method realizes the global optimal control of the crack risk, the adaptive evolution of the model and the explainability of the decision-making process, and significantly improves the crack prediction accuracy and the control effect.

Description

Technical Field

[0001] This application relates to the field of monitoring and control during road construction, and in particular to an intelligent monitoring and control method for cracks in water-stabilized base courses. Background Technology

[0002] As the main load-bearing layer of the road structure, the control of shrinkage cracks in the water-stabilized base course is crucial to ensuring the long-term service performance of the road. With the increasing requirements for settlement control and crack resistance in high-fill road sections, traditional crack control methods are no longer sufficient to meet engineering needs.

[0003] In existing technologies, crack control in water-stabilized base courses mainly employs a combination of mix design optimization and segmented curing. This involves adding recycled aggregates and adjusting cement dosage to reduce material shrinkage, and using a segmented curing process of "compaction followed by film covering + moisture retention + watering" to suppress moisture evaporation. Some intelligent solutions attempt to introduce sensor monitoring and simple feedback control, but the control logic remains based on preset rules.

[0004] However, existing technologies have the following drawbacks: First, mix design optimization and curing processes are independent of each other, lacking a unified theoretical framework for overall optimization; second, control decisions rely on preset rules and cannot be dynamically adjusted according to real-time environmental changes and construction quality fluctuations; third, the coupling relationships between various physical fields (moisture field, stress field, damage field) are fragmented, resulting in limited accuracy in crack prediction. These drawbacks make it difficult for existing methods to achieve precise control of crack risk. Summary of the Invention

[0005] This application provides an intelligent monitoring and control method for cracks in water-stabilized base courses. It can achieve global optimal control of crack risk by deeply coupling mix optimization, construction control and environmental response through a unified functional variational framework.

[0006] In a first aspect, this application provides an intelligent monitoring and control method for cracks in water-stabilized base courses. The method includes the following steps: obtaining preset parameters; constructing a system state field based on the preset parameters; constructing a scalar functional, wherein the scalar functional characterizes the degree to which the water-stabilized base course system deviates from the ideal crack-free state, and the scalar functional is a function of the system state field including external environmental driving terms; obtaining an optimal control field that minimizes the risk of cracking in the system by solving for the minimum value of the scalar functional; and generating and outputting control commands for controlling construction equipment based on the optimal control field.

[0007] By adopting the above technical solution, the previously scattered issues of mix proportion optimization, compaction control, and watering maintenance are unified into a scalar functional framework, and the globally optimal control strategy is solved using variational methods. This solution overcomes the limitations of traditional step-by-step decision-making, upgrading crack control from rule-driven to optimization-driven, and enabling dynamic adjustment of control parameters based on real-time conditions to minimize crack risk.

[0008] Furthermore, the system state field is a spatiotemporal field, including a moisture content field, a displacement field, a damage field, and a construction control field; the construction control field includes a compaction control field and a water spraying control field, the compaction control field being used to characterize the action of the compaction equipment, and the water spraying control field being used to characterize the action of the water spraying equipment.

[0009] By adopting the above technical solutions, the control actions of construction equipment are expressed continuously and in a field-like manner, enabling compaction and water spraying control to participate in the solution as direct variables of functional optimization, thus achieving a deep integration of control decision-making and physical field evolution.

[0010] Furthermore, the scalar functional includes a moisture-mechanical coupling potential, a damage evolution potential, a construction control potential, and an information entropy potential; the moisture-mechanical coupling potential, the damage evolution potential, the construction control potential, and the information entropy potential together constitute the scalar functional.

[0011] By adopting the above technical solution, the multi-physics coupling relationship is decomposed into four potential energy terms with clear physical meaning, so that the functional construction has a clear physical basis, while ensuring the synergistic effect between various physical processes.

[0012] Furthermore, the moisture-mechanical coupling potential is an energy integral expression that includes the stiffness tensor, strain tensor, matrix suction potential, and moisture gradient penalty term, which vary with moisture content and damage.

[0013] By adopting the above technical solutions, the deterioration effect of moisture content changes on material stiffness, the shrinkage deformation driven by matrix suction, and the influence of non-uniform moisture distribution on stress concentration were accurately characterized, providing an accurate mechanical basis for the prediction of drying shrinkage cracks.

[0014] Furthermore, the damage evolution potential is a phase-field damage integral expression that includes fracture energy, characteristic length scale, damage field gradient, and strain over-limit penalty term.

[0015] By adopting the above technical solution, the discrete crack problem is transformed into a continuous field problem, which can simulate the entire process of crack nucleation and propagation, avoid the pre-defined assumptions about crack paths in traditional methods, and improve the accuracy of crack prediction.

[0016] Furthermore, the construction control potential is an integral expression that includes the deviation terms between the compaction control field and the reference control field, the deviation terms between the water spraying control field and the reference control field, the coupling term between compaction energy and moisture content, and the coupling term between water spraying efficiency and evaporation rate.

[0017] By adopting the above technical solutions, construction costs, control deviations, and the matching degree between construction actions and the physical field are incorporated into the optimization objectives, so that the optimal control field obtained by the solution can not only effectively control cracks, but also meet the requirements of engineering economy and feasibility.

[0018] Furthermore, the information entropy potential is an integral expression that includes the mixed entropy term of the water content distribution and the damage entropy term.

[0019] By adopting the above technical solution and introducing the concept of thermodynamic entropy, the system evolution follows natural laws, while the constraints of the first few potential energies guide it to the desired state, thus improving the physical completeness of the functional framework.

[0020] Furthermore, the step of obtaining the optimal control field that minimizes the risk of cracking in the system by solving for the minimum value of the scalar functional specifically includes: iteratively updating the system state field according to the generalized gradient flow equation, wherein the generalized gradient flow equation is the product of the negative dynamic coefficient tensor and the variational derivative of the scalar functional with respect to the system state field, where the partial derivative of the system state field with respect to time is equal to the product of the negative dynamic coefficient tensor and the variational derivative of the scalar functional with respect to the system state field; and extracting the optimal control field from the system state field during the iterative update process.

[0021] By adopting the above technical solution, the functional minimization problem is transformed into a gradient flow evolution problem, and the optimal control field is obtained through iterative solution. This method has good numerical stability and convergence, and can meet the computational efficiency requirements of real-time control.

[0022] Furthermore, the method also includes: calculating the variational derivative of the scalar functional with respect to any field quantity in the system state field; and determining the contribution of the field quantity to the crack risk based on the ratio of the product of the variational derivative and the corresponding field quantity to the scalar functional.

[0023] By adopting the above technical solutions, interpretable analysis of crack risk is achieved, which can quantitatively identify the dominant factors that lead to increased crack risk, providing engineers with decision-making basis and problem tracing capabilities.

[0024] Furthermore, the method also includes: acquiring measured strain data and measured damage data from intelligent aggregates embedded in the water-stabilized base course; constructing an observation functional, wherein the observation functional is a combination of the deviation terms between the predicted strain data and the measured strain data and the deviation terms between the predicted damage data and the measured damage data; and updating the undetermined parameters in the scalar functional online with the goal of minimizing the observation functional.

[0025] By adopting the above technical solution, a digital twin closed-loop calibration mechanism was established. The functional parameters were continuously optimized using measured data, enabling the model to adapt to material variations and environmental changes, thus ensuring the control accuracy for long-term use.

[0026] In summary, this application has at least the following beneficial effects: A method for intelligent monitoring and control of cracks in water-stable base courses based on functional variation is provided, which unifies multi-physics coupling, construction control and environmental response into a single optimization framework, and achieves global optimal control of crack risk. By using variational derivatives and contribution calculations, interpretable analysis of crack risk was achieved, overcoming the limitations of traditional black-box models. The online calibration mechanism based on measured data of intelligent aggregates enables the control model to have adaptive learning capabilities, ensuring long-term service performance.

[0027] It should be understood that the description in the Summary Section is not intended to limit the key or essential features of the embodiments of this application, nor is it intended to restrict the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0028] The above and other features, advantages, and aspects of the embodiments of this application will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein: Figure 1 A schematic diagram of an exemplary operating environment in which embodiments of this application can be implemented is shown.

[0029] Figure 2 A flowchart of an intelligent monitoring and control method for cracks in a water-stabilized base course, as described in an embodiment of this application, is shown. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0031] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0032] This application provides an intelligent monitoring and control method for cracks in water-stabilized base courses. It deeply integrates multi-physics coupling, construction control and environmental response through a unified functional variational framework, achieving global optimal control of crack risk. It also has interpretable analysis and adaptive learning capabilities, significantly improving crack prediction accuracy and control effect.

[0033] Figure 1 A schematic diagram of an exemplary operating environment in which embodiments of this application can be implemented is shown.

[0034] Reference Figure 1 The operating environment includes a perception layer, a computing layer, an execution layer, and a communication network that supports data interaction between the layers, together forming a closed-loop control system that integrates data acquisition, intelligent decision-making, and automatic execution.

[0035] The sensing layer consists of a raw material testing unit, an intelligent aggregate monitoring network, an environmental monitoring station, and construction process monitoring equipment. The raw material testing unit, located at the feed inlet of the mixing plant, includes a recycled aggregate crushing value tester, a water absorption rate tester, and a cement grade testing device. The recycled aggregate crushing value tester measures the crushing index of the recycled aggregate, with a measurement range of 0% to 30%. The system automatically issues a warning when the measured value exceeds 20%. The water absorption rate tester measures the saturated surface-dry water absorption rate of the recycled aggregate, with a measurement accuracy of ±0.1% and a typical measurement range of 2% to 8%. The cement grade testing device determines the actual strength grade of the cement arriving on site. All of these devices automatically initiate testing upon arrival of each batch of materials, and the test data is uploaded to the computing layer in real time via the communication network.

[0036] The intelligent aggregate monitoring network is implemented by embedding intelligent aggregates with built-in MEMS sensors in key cross-sections of the base layer. The intelligent aggregates have the same shape as ordinary aggregates, with a diameter of 20 mm to 40 mm, preferably 30 mm. The embedding depth is divided into three layers: shallow, medium, and deep. The shallow layer is 50 mm to 80 mm below the base layer surface, preferably 60 mm; the medium layer is 150 mm to 250 mm, preferably 200 mm; and the deep layer is 300 mm to 400 mm, preferably 350 mm. The horizontal spacing between adjacent intelligent aggregates is 5 meters to 15 meters, increasing to 3 meters to 5 meters at cross-section changes. The temperature sensor built into the intelligent aggregate has a range of [missing information]. The range is ℃ to 80℃, with an accuracy of ±0.3℃; the humidity sensor range is 0% to 100% relative saturation, with an accuracy of ±2%; the strain sensor range is... to Accuracy is ±5 All sensors collect data at a frequency of 1 to 60 times per minute, with a setting of 30 times per minute during critical construction periods.

[0037] The environmental monitoring station is deployed in an open area at the edge of the construction site, no more than 200 meters from the work surface, preferably 50 to 100 meters. The environmental monitoring station integrates temperature and humidity sensors, an anemometer, and a rain gauge. The temperature and humidity sensors are housed in a Stevenson screen, with a temperature range of [missing information]. The temperature range is ℃ to 60℃, with an accuracy of ±0.2℃; the relative humidity range is 0% to 100%, with an accuracy of ±2%; the anemometer is installed at a height of 2 meters above the ground, with a range of 0 m / s to 30 m / s and a starting wind speed of 0.5 m / s; the rain gauge is a tipping bucket type with a resolution of 0.2 mm. The environmental monitoring station collects data at intervals of 1 minute to 10 minutes, increasing the frequency to once per minute during rainfall or periods of high evaporation.

[0038] The construction process monitoring equipment includes a continuous compaction monitoring system installed on the compaction equipment. This system consists of an accelerometer sensor mounted on the roller shaft of the vibratory roller and a processing unit. The accelerometer sensor has a sampling frequency of 200 Hz to 1000 Hz, preferably 500 Hz. The processing unit calculates the compaction value in real time during the compaction process. and the coefficient of variation of compaction uniformity compaction degree Expressed as a percentage, its calculation range is 80% to 105%; coefficient of variation It reflects the dispersion of compaction degree, with a value ranging from 0% to 15%. Monitoring data is uploaded via communication network once per second.

[0039] The computing layer is deployed in the central control room and includes high-performance edge computing servers and a quantum computing cloud service interface. The edge computing servers are equipped with four NVIDIA Tesla V100 or equivalent GPU accelerator cards, each with 32GB of VRAM, for a total server memory of 256GB and storage capacity of 20TB. The edge computing servers perform variational solutions for scalar functionals, gradient flow iterative updates, and online calibration calculations for observation functionals. A single computation cycle takes 30 to 300 seconds, preferably 60 seconds, ensuring minute-level real-time decision-making capabilities. The quantum computing cloud service interface connects to a third-party quantum computing cloud platform via a dedicated encrypted channel. During the model training phase, it is used to solve high-dimensional functional parameter optimization problems. Each training task requires 1 to 24 hours of quantum computing resource access, with 15 to 25 parameters to be determined, preferably 20.

[0040] The execution layer includes intelligent compaction equipment and an intelligent mulching and watering system. The intelligent compaction equipment primarily consists of a 20-ton to 26-ton single-drum vibratory roller, preferably 22 tons. The controller installed on the equipment receives compaction control commands from the optimal control field and automatically adjusts the number of compaction passes, vibration amplitude, and travel speed. The number of compaction passes is adjustable from 4 to 10 passes, set to 6 passes under normal construction conditions, and automatically increased to 8 passes in areas with insufficient compaction. The vibration amplitude is divided into three levels: low, medium, and high. Low amplitude is 0.8 mm to 1.2 mm, medium amplitude is 1.4 mm to 1.8 mm, and high amplitude is 2.0 mm to 2.4 mm, preferably dynamically switched according to the material type. The travel speed is adjustable from 2 km / h to 6 km / h, preferably 4 km / h.

[0041] The intelligent mulching and watering system includes an automatic mulching machine and an intelligent water sprinkler truck. The automatic mulching machine is installed behind the compaction equipment, with a mulching width of 3 to 5 meters, matching the width of the base layer. The mulching machine automatically unfolds a polyethylene plastic film according to the mulching timing command; the film thickness is 0.08 mm to 0.15 mm, preferably 0.12 mm. The intelligent water sprinkler truck is equipped with a water tank of 10 to 15 cubic meters, preferably 12 cubic meters. A set of solenoid valve nozzles is installed at both the front and rear ends of the truck body. Each set of nozzles contains 6 to 12 nozzles, with a single nozzle flow rate of 2 liters / minute to 8 liters / minute. The water sprinkler truck dynamically adjusts the sprinkling interval and single-sprinkler volume according to the sprinkling control field. The sprinkling interval adjustment range is 20 minutes to 180 minutes, set to 30 minutes in hot, dry weather, and extended to 120 minutes in rainy weather. The single-sprinkler volume adjustment range is 0.3 liters / square meter to 1.5 liters / square meter, preferably 0.8 liters / square meter.

[0042] The communication network employs Industrial Internet of Things (IIoT) technology, utilizing LoRa or 5G networks to achieve low-latency bidirectional data transmission between sensing layer devices and computing layer servers. LoRa networks are suitable for scenarios with long transmission distances and small data volumes, operating in the 470 MHz to 510 MHz frequency band with transmission rates of 0.3 kbps to 50 kbps and communication distances of 3 km to 8 km. 5G networks are suitable for scenarios with short transmission distances and large data volumes, operating in the 3.3 GHz to 3.6 GHz or 4.8 GHz to 5.0 GHz frequency bands with peak transmission rates exceeding 1 Gbps and end-to-end latency below 20 milliseconds. Sensing layer devices automatically select the network to access based on data volume. Raw material testing data and environmental monitoring data are transmitted via LoRa, while high-frequency data from intelligent aggregates and compaction monitoring data are transmitted via 5G. Control commands generated by the computing layer are distributed to execution layer devices in real time via the 5G network.

[0043] The aforementioned layers are tightly coupled through a communication network. The real-time data collected by the perception layer drives the computation layer to solve for the optimal control field, and the execution layer operates automatically according to instructions, thereby ensuring the engineering feasibility of the intelligent monitoring and control method for cracks in water-stabilized base courses.

[0044] Figure 2 A flowchart of an intelligent monitoring and control method for cracks in a water-stabilized base course, as described in an embodiment of this application, is shown.

[0045] Reference Figure 2 The method specifically includes the following steps: S1: Get preset parameters.

[0046] The method in this step specifically includes: obtaining preset parameters. These preset parameters include raw material parameters, environmental parameters, engineering target parameters, and construction process data. The raw material parameters are obtained through a raw material testing unit and include the range of recycled aggregate content, recycled aggregate water absorption rate, cement grade, and cement dosage range. Specifically, the range of recycled aggregate content is expressed as follows: This indicates the ratio of recycled aggregate to the total mass of coarse aggregate, with an upper limit of 50%; the water absorption rate of recycled aggregate. Expressed as a percentage, the typical value ranges from 2% to 8% as measured by a water absorption tester; cement grade The strength grade of cement is expressed in megapascals (MPa); the cement dosage range is expressed as follows: , representing the proportion of cement to the total dry mass, i.e., 3% to 6%. These raw material parameters will be used to construct the material constitutive model and the material coefficient tensor in the functional.

[0047] Environmental parameters are collected in real time through environmental monitoring stations, including the current ambient temperature. (Unit: °C), Relative Humidity (expressed as a percentage) and wind speed (Unit: m / s). Environmental monitoring stations collect data at a frequency of minutes, increasing to once per minute during critical construction periods. These environmental parameters are used to calculate the surface evaporation rate. This serves as the input for the external environment driving terms in the functional.

[0048] The target parameters for the project are input by the construction party according to the design requirements, including the design strength. (Unit: MPa, usually refers to 7-day unconfined compressive strength), maximum allowable shrinkage rate (Unit: microstrain) and material cost ceiling (Unit: Yuan / m) 3 These parameters will serve as constraints in the functional optimization process, ensuring that the optimal control field obtained meets the basic engineering requirements.

[0049] Construction process data is acquired in real time through construction process monitoring equipment, including compaction degree. and its coefficient of variation Compaction degree Expressed as a percentage, its calculation range is 80% to 105%, coefficient of variation Reflecting the dispersion of compaction degree, the values ​​range from 0% to 15%. These data are collected in real time by a continuous compaction monitoring system installed on the compaction equipment and uploaded once per second. The construction process data will be used to construct the initial damage distribution in the system state field and to correct the strength constraints and control parameters in the functional during construction disturbance sensing.

[0050] After all preset parameters are acquired, they are aggregated to the computing layer through the communication network and normalized according to the preset data format to eliminate the influence of dimensions, providing a unified input for the subsequent construction of the system state field and scalar functional.

[0051] S2: Construct the system state field based on the preset parameters.

[0052] In this step, a system state field is constructed based on the preset parameters. The system state field is a spatiotemporal field, defined as follows: ,in For spatial coordinates, Using time as the coordinate system, this state field comprehensively describes the physical and control states of the water-stabilized base course at any location and time during construction and curing. The system state field includes the moisture content field. Displacement field Damage field Temperature field and construction control site The construction control field includes a compaction control field. and sprinkler control field The compaction control field characterizes the action of the compaction equipment, and the water spraying control field characterizes the action of the water spraying equipment. Temperature field Used to describe the temperature distribution within the base layer, its initial value can be measured by the temperature sensor built into the smart aggregate, or assumed to be related to the ambient temperature. Consistent.

[0053] Moisture field The initial distribution is based on the initial moisture content in the raw material parameters. The porosity after compaction Construction. At the start of construction. Assuming the moisture content is uniformly distributed along the depth direction, i.e. ,in Determined by the optimum moisture content in the mix design, typically ranging from 4% to 8%. Void ratio after compaction. The dry density and particle size distribution measured by the raw material testing unit are calculated using conventional soil mechanics formulas, with typical values ​​ranging from 0.3 to 0.6. In subsequent steps, the moisture field will dynamically evolve according to the moisture migration equation and evaporation boundary conditions, and its evolution will be controlled by the moisture-mechanical coupling potential in the subsequently constructed scalar functional.

[0054] Displacement field Used to describe the deformation of base materials under changes in moisture and load. Initial moment. Assuming the base layer is in a state of no deformation, that is The boundary conditions of the displacement field are set according to the geometric constraints of the base layer. The bottom and lateral boundaries of the base layer are usually set to have zero normal displacement, while the surface is a free boundary. The displacement field will be used to calculate the strain tensor. This is then used for calculating the mechanical potential energy in subsequent functionals.

[0055] Damage field It is a scalar field with a range of values. Where 0 represents completely undamaged material, and 1 represents complete material cracking and failure. The initial value of the damage field is set to zero, i.e. This indicates that there are no cracks in the base layer initially after construction. In subsequent steps, the damage field will dynamically evolve according to the strain magnitude and energy criterion, and its evolution law is controlled by the damage evolution potential in the scalar functional. The introduction of the damage field enables this method to continuously describe the initiation and propagation process of cracks, avoiding the shortcomings of traditional methods that require pre-setting crack paths.

[0056] Construction control field It is a two-dimensional spatiotemporal field because it only acts on the surface of the base layer and does not change along the depth direction. The construction control field is composed of the compaction control field. and sprinkler control field It consists of two parts. Compaction control field It is a multidimensional control field, which actually contains three sub-components: the number of rolling passes control component. Vibration amplitude control component and driving speed control component These control components are all continuous value fields, corresponding to the operating parameters of the compaction equipment. Sprinkler control field. It also contains two sub-components: the sprinkler interval control component. and single spray volume control During the initialization phase, the construction control field is based on a reference control field pre-set by engineering experience. and Settings, i.e. , The reference control field can be preset according to the project type and climate conditions. For example, the typical number of compaction passes is 6, the vibration amplitude is medium, the driving speed is 4 km / h, the watering interval is 60 minutes, and the single watering volume is 0.8 L / m. 2 .

[0057] The aforementioned fields are spatially discretized using a finite element mesh. The mesh size is set according to accuracy requirements, typically ranging from 0.1 meters to 0.5 meters, and refined to 0.05 meters in crack-sensitive areas. In the time dimension, an adaptive time step is used, initially set to 1 minute, automatically shortening to 10 seconds during drastic environmental changes or control command switching. The constructed system state field will serve as the basic input for constructing the scalar functional in subsequent steps. By minimizing the functional, each field is driven to evolve towards the minimum crack risk state, and the optimal control field is extracted from it.

[0058] S3: Construct a scalar functional, which is used to characterize the degree to which the water-stabilized base system deviates from the ideal state of no cracks, and the scalar functional is a function of the system state field that includes external environmental driving terms.

[0059] In this step, a scalar-valued functional is constructed. ,in The system state field constructed for step S2, Represents the spatial gradient of the system's state field. The external environment is the driving factor. This functional is a scalar operator whose input is the system state field over the entire spatiotemporal domain, and whose output is a real number. The magnitude of this real number quantitatively reflects the generalized energy difference between the current state of the water-stabilized base system and the ideal crack-free state. The smaller the functional value, the closer the system is to the ideal state and the lower the risk of cracking.

[0060] The scalar functional includes the water-mechanical coupling potential. Damage evolution potential Construction control potential and information entropy potential The moisture-mechanical coupling potential, damage evolution potential, construction control potential, and information entropy potential together constitute the scalar functional. Right now , in This is the boundary integral term.

[0061] The water-mechanical coupling potential The energy integral expression, which includes the stiffness tensor, strain tensor, matrix suction potential, and moisture gradient penalty term varying with water content and damage, has the following specific form: ; in, Let be the integration space domain, that is, the geometric region of the entire water-stabilized base course. The first term... For elastic strain energy density, To vary with moisture content and damage The changing fourth-order stiffness tensor, in its degenerate form, is expressed as: ,in For the material at its optimum moisture content The initial stiffness tensor under undamaged conditions is determined by laboratory tests based on the cement grade and aggregate gradation parameters in the raw materials. It is a moisture content sensitivity coefficient, with a typical value range of 50 to 200, preferably 100; For the strain tensor, derived from the displacement field Calculated. Second term. It is the matrix attractive potential energy. For matrix suction, the van Genuchten model is used to describe it: ,in These are the parameters of the soil-water characteristic curve. Residual moisture content These parameters, representing the saturated moisture content, are the water absorption rate of the recycled aggregate as determined by the raw material testing unit. and cement dosage It was obtained through conversion using empirical formulas. The third term... As a penalty for moisture gradient, This is a penalty coefficient, typically ranging from 0.01 to 0.1, used to suppress drastic spatial changes in moisture content and prevent stress concentration caused by sudden changes in local moisture levels.

[0062] The damage evolution potential The phase-field damage integral expression, which includes fracture energy, characteristic length scale, damage field gradient, and strain over-limit penalty, has the following specific form: ; in, The fracture energy of the material, expressed in J / m. 2 The cement grade and recycled aggregate content in the raw material parameters are determined by fracture tests, with a typical value range of 50 J / m. 2 Up to 200 J / m 2 ; The characteristic length scale is related to the maximum aggregate size, typically ranging from 2 to 5 times the maximum aggregate size, and from 10 mm to 50 mm; the first term This is the classic form of the phase-field damage model, describing the diffusion of damage and energy dissipation, where... The item represents the energy consumed by local damage. The first term represents the nonlocal energy dissipation of the damage gradient, and together they control the width and evolution rate of the damage band. The second term is the strain over-limit penalty term. For equivalent strain, it is usually taken as the first invariant of the strain tensor or the von Mises equivalent strain; The critical strain is determined by the tensile strength of the material. and elastic modulus Estimated as tensile strength of materials and elastic modulus The cement grade and aggregate gradation in the raw material parameters are determined through indoor tests, with typical values ​​ranging from 1 MPa to 3 MPa and from 10 GPa to 30 GPa, respectively. For indicator functions, when The value is 1 if the condition is met, and 0 otherwise. This is the penalty weighting coefficient, with a typical value range of [value range missing]. to It is used to accelerate damage evolution in the strain over-limit region.

[0063] The construction control potential The integral expression, which includes the deviation terms between the compaction control field and the reference control field, the deviation terms between the sprinkler control field and the reference control field, the coupling term between compaction energy and moisture content, and the coupling term between sprinkler efficiency and evaporation rate, has the following specific form: ; in, and For reference control, preset values ​​are made based on engineering experience, such as 6 rolling passes and 60-minute watering interval. and To control the deviation weighting coefficient, the typical value range is 0.1 to 1.0, which is used to balance the importance of control accuracy and other potential energy terms. This is a coupling term between compaction energy and moisture content, indicating that compaction should be performed at the optimum moisture content. The term increases when the compaction control field deviates from the optimum moisture content, thus guiding the matching of the compaction control field and the moisture content field during the optimization process; in the coupling term middle, Specifically refers to the number of compaction passes control component in the compaction control field. This is the weighting coefficient for the coupling term, typically ranging from 0.5 to 2.0. This is a coupling term between watering efficiency and evaporation rate, where... The surface evaporation rate is calculated from environmental parameters: , in The surface saturated water vapor pressure, This is the actual water vapor pressure. The density of water (take 1000 kg / m³) 3 ), The latent heat of vaporization is taken as 2450 kJ / kg. For wind speed, For ambient temperature, Relative humidity; For the watering interval, in the coupling term middle, Specifically refers to the single-spray volume control component in a sprinkler control field. Specifically refers to the watering interval control component in a sprinkler control field. This is the water spraying efficiency weighting coefficient, typically ranging from 0.5 to 2.0. This item indicates that the amount of water sprayed should match the amount of evaporation; too much or too little water will cause this item to increase.

[0064] The information entropy potential The integral expression for the mixing entropy term and the damage entropy term, which include the water content distribution, is as follows: ; in, The reference temperature is the daily average of the ambient temperature, and the unit is Kelvin. The first term is the mixing entropy of the moisture content distribution, which is based on the form of binary mixing entropy in thermodynamics. It describes the degree of disorder in the moisture distribution. This term is smaller when the moisture content is uniformly distributed and larger when the moisture content is unevenly distributed. The first term is the saturated moisture content, determined by the porosity in the raw material parameters. The second term is the damage entropy. exist and The value is 0 when the time is right and positive when the value is in the middle, which describes the degree of disorder of the damaged system; The damage entropy weighting coefficient typically ranges from 0.01 to 0.1. The introduction of the information entropy potential causes the system evolution to follow the second law of thermodynamics, tending to increase the entropy value. However, it is also constrained by other potential energy terms, eventually reaching a balance between entropy increase and energy decrease.

[0065] The scalar functional also includes external environment driving terms. The external environment driving term is a boundary integral expression, which includes surface evaporation driving and temperature driving, and its specific form is as follows: ; in, The boundary of the upper surface of the substrate; the first term is the evaporation-driven term. This is the evaporation potential, consistent with the evaporation rate term in the Penman equation, multiplied by the surface water content. This indicates the actual amount of evaporation that occurs. The first term is the evaporation coupling coefficient; the second term is the temperature-driven term. The surface temperature of the substrate (i.e., the temperature field) (value on the surface) For ambient temperature, the square term of the temperature difference characterizes the thermal stress driving force. The temperature coupling coefficient is denoted by . This boundary term incorporates the driving effect of the environment on the base layer into the functional in the form of energy, enabling environmental factors to directly influence the system evolution.

[0066] Completed scalar functional It fully encompasses the internal physical processes of the system (moisture-mechanical coupling, damage evolution), control objectives (construction control potential), thermodynamic laws (information entropy potential), and external environmental drivers, providing a unified objective function for solving the optimal control field using the variational method in subsequent steps.

[0067] S4: By solving for the minimum value of the scalar functional, the optimal control field that minimizes the risk of cracks in the system is obtained.

[0068] The specific steps of this method include: iteratively updating the system state field according to the generalized gradient flow equation, wherein the generalized gradient flow equation is the product of the negative dynamic coefficient tensor and the variational derivative of the scalar functional with respect to the system state field, where the partial derivative of the system state field with respect to time is equal to the product of the negative dynamic coefficient tensor and the variational derivative of the scalar functional with respect to the system state field; and extracting the optimal control field from the system state field during the iterative update process.

[0069] The generalized gradient flow equation is the core mathematical model in this step, and its physical significance lies in driving the system's state field. Along scalar functionals The direction of fastest descent evolves until a local minimum of the functional is reached. This equation is specifically expressed as: ; in, For system state field time The partial derivatives represent the rates of change of each field quantity; scalar functional For the system state field The variational derivative, also known as the first-order variation, is the rate of change of the functional value when the system's state field undergoes a small change. It plays the role of "driving force" in the gradient flow equation. The kinetic coefficient tensor is a positive definite symmetric tensor used to regulate the evolution rate of various field quantities, ensuring the thermodynamic consistency of the system. The component form is ,in The evolution coefficient of the water cut field, typically ranging from [value range missing]. to m 2 / s; The evolution coefficient of the displacement field has a typical range of values. to m 2 / (N·s); This is the evolution coefficient of the damage field, with a typical value range of [value range missing]. to s -1 ; The evolution coefficient of the temperature field typically ranges from 100 to 100. to m 2 / s; The evolution coefficient of the compaction control field is typically taken from the range of values. to (Control unit) 2 / s; The evolution coefficient of the sprinkler control field, with a typical value range of [value missing]. to (Control unit) 2 / s. These coefficients are calibrated through numerical experiments or adaptively adjusted according to convergence speed requirements.

[0070] In actual calculations, the first step is to... Calculate scalar functionals Variational derivatives with respect to each component of the system's state field. Taking the water content field as an example. For example, its variational derivative is derived from the functional pair constructed in step S3. Taking the variational form, the specific expression is: ; in, Let be the partial derivative of the stiffness tensor with respect to water content, from The explicit form of the derivative is obtained; This is the partial derivative of the matrix suction with respect to the water content, obtained by differentiating using the van Genuchten model; The Laplace operator for water content is calculated by second-order difference on a spatially discrete grid; and The construction control potential and boundary driving terms are respectively paired. The variational derivative of the expression can be obtained directly by differentiating its integral expression. Other field quantities, such as displacement field... Damage field Temperature field Compaction control field Sprinkler control area The variational derivatives are calculated using a similar method. Each variational derivative is composed of the superposition of contributions from multiple physical processes, reflecting the deep coupling between various field quantities.

[0071] After obtaining the variational derivatives of all field quantities, the update quantities of each field quantity are calculated according to the generalized gradient flow equation. When using the explicit Euler time integration method, the update formula is: ; in, For the first The system state field at each time step For the updated system state field, Let be the time step. The choice of time step must satisfy numerical stability conditions, and is usually determined by the Courant-Friedrich-Lévy conditions. Right now , in This represents the spatial grid size. In actual calculations, the initial time step is set to 1 minute and is adaptively adjusted based on convergence. When the functional value decreases too quickly, the step size is automatically reduced; when the change is gradual, the step size is appropriately increased.

[0072] To improve computational efficiency and stability, implicit or semi-implicit time integration methods can be employed. For example, an implicit scheme can be used for diffusion-dominated terms, while an explicit scheme can be used for reaction terms, with the field quantities updated alternately. For large-scale three-dimensional problems, parallel computing techniques can be used, dividing the spatial domain into multiple sub-regions and distributing them to different computational cores for parallel solution.

[0073] The iterative update process continues until the convergence condition is met. The convergence criterion typically includes two aspects: first, the change in the system state field is less than a preset threshold; second, the change in the system state field is less than a preset threshold. Right now , in Values to Second, the relative decrease in the functional value is less than the preset threshold. Right now , in Values to Also set the maximum number of iterations. The typical value is 1000 to 10000 steps to prevent infinite loops.

[0074] After the system state field converges, from the converged system state field Extracting the compaction control field and sprinkler control field As the optimal control field. The specific extraction method is as follows: , , in The point represents the convergence time. These two control fields are spatially distributed continuous fields, representing the spatial distributions of compaction parameters and water spraying parameters that minimize the risk of cracking in the system, respectively. For the compaction control fields... Its three sub-components correspond to the expected values ​​of the number of compaction passes, amplitude, and velocity, respectively; for the sprinkler control field This includes two sub-components corresponding to the expected values ​​of the watering interval and the water volume per watering cycle, respectively. These control fields will be converted into specific construction equipment control instructions in step S5.

[0075] S5: Generate and output control commands for controlling construction equipment based on the optimal control field.

[0076] In this step, control commands for controlling construction equipment are generated and output based on the optimal control field. The optimal control field obtained in step S4 is a spatially distributed continuous field, including the compaction control field. and sprinkler control field Among them, the compaction control field Includes control of the number of rolling passes Vibration amplitude control component and driving speed control component Sprinkler control field Includes watering interval control component and single spray volume control These control components are continuous numerical fields, with the following value ranges: (component of number of compaction passes) The corresponding number of compaction passes is 4 to 10; vibration amplitude component The unit is millimeters, corresponding to three ranges: low (0.8 mm to 1.2 mm), medium (1.4 mm to 1.8 mm), and high (2.0 mm to 2.4 mm); driving speed component. Units are kilometers per hour; watering intervals are expressed in parts. The unit is minutes; the amount of water sprayed per application is [amount missing]. The unit is liters per square meter.

[0077] When analyzing the control parameters of the compaction equipment from the optimal control field, the continuous control components must first be converted into discrete commands executable by the construction equipment. For the number of compaction passes, the control component for the number of compaction passes is... Instructions to obtain an integer number of iterations by rounding to the nearest integer. And it is limited to the range allowed by the equipment. For vibration amplitude, the control component is based on the vibration amplitude. The value maps it to discrete amplitude levels: when When, the command is a low amplitude setting; when When, the instruction is medium amplitude; when At that time, the command is set to high gear. Regarding driving speed, the driving speed control component is... Directly used as a speed command And, depending on the equipment capacity, the speed is limited to the range of 2 km / h to 6 km / h.

[0078] When analyzing the control parameters of the sprinkler system from the optimal control field, for the sprinkler interval, the sprinkler interval control component is... Directly used as a watering interval instruction The duration of water spraying should be limited to between 20 and 180 minutes, depending on the equipment's capacity. For each spraying operation, the water volume should be controlled in portions. Directly used as a water spraying volume instruction The flow rate is limited to 0.3 liters / m² to 1.5 liters / m² depending on the equipment capacity. Considering the uniformity requirements of the water spraying operation, spatial smoothing can be used for water control fields with drastic spatial variations. For example, weighted averaging of command values ​​for adjacent grid points can prevent water trucks from frequently adjusting parameters over short distances.

[0079] When generating specific control commands, the actual operating mode and coverage area of ​​the construction equipment also need to be considered. The operating width of intelligent compaction equipment is typically 2 to 3 meters. When moving longitudinally along the road, its control parameters correspond to the spatial position of the current compaction strip. Therefore, it is necessary to sample the spatially distributed control field along the compaction path to generate a sequence of time-varying commands that varies with position. Specifically, based on the real-time GPS positioning of the compaction equipment, the current location is queried. The corresponding control parameters are used to generate real-time commands and send them to the equipment controller. For water trucks, the watering parameters are also dynamically adjusted according to their location and direction of travel to achieve on-demand watering.

[0080] Various control commands generated are transmitted to the construction equipment via the communication network. The command format uses standardized data protocols, such as the MQTT-based message format. Each command includes a device identifier, command type, timestamp, location coordinates, and parameter values. Upon receiving the commands, the intelligent compaction equipment and the intelligent mulching and watering system are parsed and executed by the equipment controller. Simultaneously, the construction equipment feeds back its execution status and real-time location to the computing layer for control optimization in the next cycle.

[0081] S6: Obtain measured strain data and measured damage data from the intelligent aggregates embedded in the water-stabilized base course; construct an observation functional, which is a combination of the deviation terms between the predicted strain data and the measured strain data, and the deviation terms between the predicted damage data and the measured damage data; update the undetermined parameters in the scalar functional online with the goal of minimizing the observation functional.

[0082] This method can be continuously executed during construction and maintenance, forming a digital twin closed-loop calibration mechanism. This enables the scalar functional model to adapt to material variations and environmental changes, ensuring long-term control accuracy. Step S6 can be executed in parallel with the iterative update of step S4, or after each control cycle, to update the parameters of the scalar functional constructed in step S3 for use in subsequent control cycles.

[0083] Specifically, the measured strain and damage data of the smart aggregates embedded in the water-stabilized base course are acquired through a smart aggregate monitoring network. The spatial coordinates of the smart aggregates are recorded during installation. ,in Number the smart aggregates. The total number of units installed is typically 20 to 50. Each smart aggregate is sampled at a specific frequency. Real-time data collection during critical construction periods Set to 30 times / minute, then reduce to 1 time / minute during normal maintenance. Let the current time be... The obtained measured data includes strain tensor components. ,in Representing the six independent components of the strain tensor, and the damage scalar. The damage value is indirectly calculated using the acoustic emission sensor or electrical impedance sensor built into the smart aggregate.

[0084] Calculate predicted strain and predicted damage data at the same location and time based on the current scalar functional model. The specific calculation method is as follows: [The calculation is performed using the current scalar functional model and the current system state field.] Based on this, the evolution process described in steps S2 to S4 is solved to obtain the location of the smart aggregate. Predicted strain tensor at the location and predicted damage value These predictions reflect the evolution of the system state field to time t under the current model parameters. The physical state that should exist at that time.

[0085] Constructing observation functionals The observation functional is a combination of the deviation terms between predicted strain data and measured strain data, and the deviation terms between predicted damage data and measured damage data. The specific mathematical form of the observation functional is: ; in, This is the parameter vector to be updated, containing parameters in the scalar functional that need to be calibrated using measured data, such as the stiffness tensor. fracture energy Feature length scale Moisture gradient penalty coefficient Control potential weight coefficient Information entropy potential weighting coefficient and dynamic coefficient tensor The components, etc. The number of these parameters is usually 15 to 25, preferably 20. A sliding time window can be used to select the set of time points for calibration, such as all acquisition times within the most recent 24 hours. Let Frobenius norm be the tensor, and let square root be the sum of squares of the six components of the strain tensor. and The weighting coefficients for the strain deviation and damage deviation terms are set based on the confidence level of the measurement data, with typical values ​​ranging from [value range missing]. , to This is used to balance the contribution of the two types of data to parameter updates.

[0086] To minimize the observation functional To achieve this, the undetermined parameters in the scalar-valued functional are updated online. This is because the observed functional... It's about the parameter vector. The nonlinear function has a moderate number of parameters and can be solved using Bayesian optimization or gradient descent methods.

[0087] When using the gradient descent method, it is necessary to calculate the gradient of the observed functional with respect to the parameters. .because Indirectly dependent on scalar functional models The gradient calculation requires the adjoint method to reduce computational complexity. The core of the adjoint method is to solve an adjoint equation that propagates backward from the original problem, thereby efficiently obtaining gradient information. Specifically, for each time point... Introducing accompanying variables Solve the adjoint equation: ; in, Let be the variational derivative matrix of the scalar functional with respect to the system state field. To observe the derivative of the functional loss with respect to the system's state field, it can be directly calculated from the deviation term at the location of the smart aggregate. The adjoint equation starts from the current time... Integrating in reverse to the initial time step yields the adjoint variable field. Then, the observation functional with respect to parameters is calculated. gradient: ; After obtaining the gradient, update the parameters using an adaptive gradient descent algorithm such as Adam or RMSprop: , in The learning rate typically ranges from [value missing]. to .

[0088] When using the Bayesian optimization method, the observed functional Treating it as a Gaussian process prior, the next sampling point is selected through a sampling function (such as the desired improvement in EI or the confidence upper limit UCB), gradually approaching the global optimum. Bayesian optimization is suitable for cases with a small number of parameters (usually less than 20) and a non-convex objective function. Its advantage lies in its ability to handle noise and find the global optimum. The specific steps are: first, initialize a set of parameter samples. Calculate the corresponding observed functional value; construct a Gaussian process surrogate model; select the next most promising parameter point based on the acquisition function. Calculate the observed functional value at that point and update the surrogate model; repeat until convergence.

[0089] Regardless of the optimization method used, after updating the parameters, it is necessary to check whether the updated parameters are within a reasonable physical range. For example, fracture energy. The values ​​should be kept positive, and the weighting coefficients should be kept within a certain order of magnitude. If the parameters exceed the preset boundaries, they should be truncated or reinitialized.

[0090] Updated scalar functional parameters The data is saved and used for solving subsequent steps. In the next control cycle, steps S2 to S5 will recalculate the optimal control field based on the updated scalar functional, achieving adaptive evolution of the model. Through this continuous online calibration mechanism, the scalar functional model can continuously adapt to batch variations in material properties, environmental changes, and fluctuations in construction quality, ensuring the stability and reliability of crack control effects during long-term service.

[0091] S7: Calculate the variational derivative of the scalar functional with respect to any field quantity in the system state field; determine the contribution of the field quantity to the crack risk based on the ratio of the product of the variational derivative and the corresponding field quantity to the scalar functional.

[0092] This method can be executed after obtaining the optimal control field or during crack risk warning. Its core lies in using variational principles to achieve interpretable analysis of crack risk, making the originally black-box optimization process transparent and understandable, and providing engineers with a scientific basis for decision-making. Step S7 can be directly executed based on the variational derivative results calculated during the iterative update process of step S4, without the need for repeated calculations.

[0093] In steps S2 to S5, this method obtains the optimal control field by solving for the minimum of the scalar functional, but only knows "what to do" without understanding "why to do it." Step S7, by introducing contribution analysis, reveals the key factors leading to crack risk, enabling engineers to understand the physical reasons behind the control commands and to intervene manually when necessary.

[0094] Specifically, firstly, the variational derivatives of the scalar functionals under the current system state field with respect to any one of the following fields: moisture content field, displacement field, damage field, temperature field, compaction control field, and sprinkler control field. These variational derivatives have already been calculated as driving forces in the generalized gradient flow equations of step S4, so the calculation results of step S4 can be directly used. For any field quantity... Its variational derivative It has already been calculated and stored during the iterative update process at each time step. The physical meaning of the variational derivative is: in the field quantity Adding a unit perturbation at a certain spatial location will cause the scalar functional of the entire system to change. To what extent does it change? Therefore, the larger the absolute value of the variational derivative, the more sensitive the field quantity at that location is to the overall crack risk of the system.

[0095] The contribution of the field quantity to the crack risk is calculated based on the ratio of the product of the variational derivative and the corresponding field quantity to the scalar functional. The defining formula is: ; in, For scalar functionals with respect to field quantities The variational derivative, This represents the value of the field at the current moment. The scalar functional value at the current moment. Product Dimensions and Similarly, its physical meaning can be understood as the "contribution power" of the field quantity to the functional value in the current state. Divided by The dimensionless contribution is then obtained, which facilitates comparison between different field quantities. Taking the absolute value is to focus on the magnitude of the contribution rather than its direction.

[0096] For water content field Contribution This reflects the contribution of localized moisture content anomalies to the overall crack risk. When a certain area... When the moisture content is significantly higher than normal, it indicates that the moisture content in the area deviates too much from the optimal value, which is the dominant factor leading to an increased risk of cracking. It is necessary to strengthen water spraying maintenance or adjust the timing of compaction.

[0097] For displacement field Contribution This reflects the contribution of localized deformation concentration to crack risk. Regions with a high contribution from the displacement field typically correspond to potential stress concentration areas, and the damage evolution in these areas requires close monitoring.

[0098] For damage field Contribution This directly reflects the impact of existing damage on system stability. When a certain area... When the damage increases rapidly, it indicates that the damage is expanding at an accelerated pace and may soon form a macroscopic crack, requiring immediate reinforcement measures.

[0099] For temperature field Contribution This reflects the contribution of abnormal temperature to the risk of cracking, such as increased drying shrinkage due to high temperature or cold shrinkage stress due to low temperature. When this value is too high, temperature control measures should be taken.

[0100] For compaction control field Contribution This reflects the impact of current compaction parameter settings on crack risk. For example, if... If the value of a certain area is high and strongly correlated with the number of compaction passes, it indicates that the number of compaction passes in that area is insufficient or excessive, and the compaction process needs to be adjusted.

[0101] For sprinkler control field Contribution This reflects the impact of current sprinkler settings on the risk of cracking. If If the correlation with the watering interval control component is strong, it indicates that the watering frequency needs to be adjusted; if the correlation with the single watering amount control component is strong, it indicates that the single watering amount needs to be adjusted.

[0102] In practical applications, the contribution of multiple field quantities can be calculated and ranked, and the top three factors with the largest contributions can be identified as the dominant factors in crack risk. For example, the calculation results for a certain region might be: , , , This indicates that abnormal moisture content is the primary issue, damage development is a secondary issue, and the influence of control parameters is relatively small.

[0103] The contribution levels are output as heatmaps or numerical lists to identify the dominant factors in crack risk, providing engineers with a basis for decision-making. Heatmaps use color intensity to represent contribution levels and are overlaid on the base geometry model, allowing engineers to visually see the sources of risk at various locations across the entire work surface. Numerical lists provide a breakdown of the contribution level for each smart aggregate location or each grid node, facilitating quantitative analysis and recording.

[0104] For example, the system can display on the control interface: "The risk of cracks is high in the section from K12+300 to K12+500. The main contributing factor is abnormal moisture content (0.52), which suggests increasing the watering frequency; the secondary factor is cumulative damage (0.28), which suggests local pressure replenishment." Such interpretable output allows engineers not only to know what measures need to be taken, but also to understand why these measures are necessary, thereby increasing their trust in the automated control system and enabling more precise human intervention when necessary.

[0105] Through contribution analysis of step S7, this method achieves a leap from "black box optimization" to "transparent cognition," enabling the intelligent monitoring and control method for cracks in water-stabilized base courses to not only possess excellent control performance but also have interpretability that engineers can understand. This is an important innovative feature that distinguishes it from traditional intelligent control methods.

[0106] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0107] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for intelligent monitoring and control of cracks in water-stabilized base courses, characterized in that, Includes the following steps: Get preset parameters; The system state field is constructed based on the preset parameters; A scalar value functional is constructed to characterize the degree to which the water-stabilized base system deviates from the ideal state of no cracks, and the scalar value functional is a function of the system state field that includes external environmental driving terms. By solving for the minimum value of the scalar functional, the optimal control field that minimizes the risk of cracks in the system is obtained. Based on the optimal control field, control commands for controlling construction equipment are generated and output.

2. The method according to claim 1, characterized in that, The system state field is a spatiotemporal field, including a water content field, a displacement field, a damage field, and a construction control field; The construction control field includes a compaction control field and a water spraying control field. The compaction control field is used to characterize the action of the compaction equipment, and the water spraying control field is used to characterize the action of the water spraying equipment.

3. The method according to claim 2, characterized in that, The scalar functionals include the moisture-mechanical coupling potential, the damage evolution potential, the construction control potential, and the information entropy potential. The moisture-mechanical coupling potential, damage evolution potential, construction control potential, and information entropy potential together constitute the scalar functional.

4. The method according to claim 3, characterized in that, The moisture-mechanical coupling potential is an energy integral expression that includes the stiffness tensor, strain tensor, matrix suction potential, and moisture gradient penalty term, which vary with moisture content and damage.

5. The method according to claim 3, characterized in that, The damage evolution potential is a phase-field damage integral expression that includes fracture energy, characteristic length scale, damage field gradient, and strain over-limit penalty.

6. The method according to claim 3, characterized in that, The construction control potential is an integral expression that includes the deviation terms between the compaction control field and the reference control field, the deviation terms between the water spraying control field and the reference control field, the coupling term between compaction energy and moisture content, and the coupling term between water spraying efficiency and evaporation rate.

7. The method according to claim 3, characterized in that, The information entropy potential is an integral expression that includes the mixed entropy term of the water content distribution and the damage entropy term.

8. The method according to any one of claims 1 to 7, characterized in that, The process of obtaining the optimal control field that minimizes the system crack risk by solving for the minimum value of the scalar functional specifically includes: The system state field is iteratively updated according to the generalized gradient flow equation, which states that the partial derivative of the system state field with respect to time is equal to the product of the negative dynamic coefficient tensor and the variational derivative of the scalar functional with respect to the system state field. The optimal control field is extracted from the system state field during the iterative update process.

9. The method according to claim 8, characterized in that, Also includes: Calculate the variational derivative of the scalar functional with respect to any field quantity in the system state field; The contribution of the field quantity to the crack risk is determined by the ratio of the product of the variational derivative and the corresponding field quantity to the scalar functional.

10. The method according to any one of claims 1 to 7, characterized in that, Also includes: Acquire measured strain and damage data from intelligent aggregates embedded in water-stabilized base courses; Construct an observation functional, which is a combination of the deviation terms between predicted strain data and measured strain data and the deviation terms between predicted damage data and measured damage data; With the goal of minimizing the observed functional, the undetermined parameters in the scalar functional are updated online.

Images