Electro-osmosis linkage control method and system for salt heaving-subsidence disaster of saline soil roadbed

Abstract

The present application relates to the technical field of subgrade disaster monitoring and linkage control, and particularly relates to a method and system for electro-osmotic linkage control of salt expansion-subsidence disaster of saline soil subgrade, which divides into multiple control units, collects multiple source state data to construct a water-heat-salt-force state vector, identifies double-front parameters and calculates a salt expansion risk index, a subsidence risk index and a salt expansion-subsidence coupling transformation index for disaster state identification and grading of each control unit; solves a treatment linkage control vector through rolling optimization under constraint conditions to realize treatment execution and feedback update. The present application can identify salt crystallization front and dissolution wetting front based on multiple source state sensing results, accurately calculate a salt expansion risk index, a subsidence risk index and a salt expansion-subsidence coupling transformation index, and generate a suitable treatment linkage control vector under anti-induction constraints to realize dynamic control and closed-loop adjustment of salt expansion risk of saline soil subgrade.

Description

Technical Field

[0001] This invention relates to the field of roadbed disaster monitoring and linkage control technology, and in particular to an electroosmotic linkage control method and system for salt swelling-dissolution disasters in saline soil roadbeds. Background Technology

[0002] During long-term service, saline soil subgrades are often simultaneously subjected to moisture migration, temperature cycling, salt crystallization / dissolution, and repeated traffic loads. Existing research shows that the engineering response of saline soil is not a simple strength decay problem, but a typical "water-heat-salt-force" coupled problem: on the one hand, salt in the pore solution crystallizes and induces volume expansion under conditions such as cooling and evaporation concentration, i.e., salt swelling; on the other hand, soluble salts in the soil dissolve and are lost under wetting conditions, weakening or even eliminating the original interparticle cementation, making it prone to sudden compressive deformation and structural collapse under external loads, i.e., solution collapse. Existing research also indicates that temperature, salt content, salt type, moisture content, initial density, and overlying load all significantly affect the salt swelling behavior of saline soil, and there are obvious interactions between these factors. The identification and treatment of saline soil subgrade diseases need to be carried out from the perspective of multi-field coupling of water, heat, salt, and force, rather than relying solely on single parameters or static experience judgments.

[0003] In engineering practice, methods such as replacement, compaction, water and salt barriers, drainage systems, or chemical modifications are commonly used to prevent or repair saline soil subgrade diseases. While these measures can reduce water and salt migration or improve bearing capacity under localized conditions, their technical approaches primarily rely on static structural protection or post-treatment reinforcement, making it difficult to adjust the treatment intensity in real time based on the spatiotemporal changes in water infiltration, salt redistribution, salt crystallization expansion, and structural softening during subgrade service. Existing research also points out that while compaction and replacement are commonly used physical modification methods, they suffer from large workloads, high costs, and insufficient continuous response capability for subgrades over a wide service life. Furthermore, relying solely on isolation or reinforcement measures makes it difficult to simultaneously achieve a dynamic balance between salt swelling inhibition and solution subsidence control. In the existing technology, Chinese invention publication CN109653185B discloses a multi-data source subgrade monitoring system and its monitoring and evaluation method for saline soil areas. The core objective of this invention is "monitoring + evaluation," with a focus on multi-source data acquisition and correlation analysis, providing an evaluation basis for disease cause analysis and improvement methods. However, the above-mentioned scheme still has obvious shortcomings: First, although the invention considers multi-source meteorological and subgrade data, it does not identify the two most critical evolution interfaces in saline soil subgrades—salt crystallization front and dissolution wetting front—and therefore it is difficult to describe the formation, expansion and mutual approximation process of salt swelling zone and dissolution zone in the subgrade depth direction; Second, the invention mainly evaluates the correlation between net evaporation rate, dynamic modulus, salt content and cumulative deformation, and has not established a dedicated risk index system that can characterize salt swelling risk, dissolution risk and their coupling transformation relationship, and therefore it is difficult to depict the catastrophic transformation process of "first crystallization and cracking, then wetting and dissolution" or "first wetting and softening, then crystallization and destruction"; Third, the invention has not formed a control vector for actuators such as drainage, electroosmotic salt migration, pumping, grouting, insulation or traffic load restriction, nor has it established a feedback correction mechanism for treatment actions. In essence, it is still a monitoring and evaluation technology, rather than a closed-loop control technology integrating early warning and treatment.

[0004] In addition, Chinese invention with publication number CN103792340B discloses a monitoring and early warning system and method for saline soil based on the Internet of Things. Although this invention has the ability to remotely monitor and provide trend warnings, its monitoring objects are mainly regional saline soil properties and spatial distribution of salt content. The technical objectives are biased towards the evaluation of the current status and trend analysis of soil salinization. It does not design for traffic load, roadbed deformation, salt swelling-dissolution coupling evolution and linkage control of treatment execution mechanism under the service state of roadbed. It also lacks closed-loop adjustment logic coupled with engineering treatment actions.

[0005] Therefore, although existing technologies have certain capabilities for monitoring, early warning, or evaluating saline soil and subgrade, they still generally suffer from the following problems: First, the coupling mechanism of salt swelling and dissolution in saline soil subgrades is not sufficiently characterized, and there is a lack of risk classification methods that can reflect the dynamic transformation relationship between the two; second, there is a lack of direct linkage between early warning results and treatment actions, and a technical system for coordinated control of multiple execution mechanisms such as drainage, electroosmotic salt migration, pumping, grouting, thermal insulation and heating, and traffic load restrictions has not yet been formed; third, there is a lack of constraint mechanisms to address the reverse induction effects during treatment, such as inhibiting salt swelling but inducing dissolution or alleviating dissolution but exacerbating recrystallization salt swelling; and fourth, there is a lack of closed-loop control methods that use post-treatment feedback data to continuously adjust risk levels and control parameters. Based on this, there is an urgent need to propose an electroosmotic linkage control method and system for salt swelling-dissolution disasters in saline soil subgrades that better meets the dynamic management needs of subgrades during their service life. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and system for electroosmotic linkage control of salt swelling-dissolution disasters in saline soil roadbeds, aiming to solve at least the following technical problems existing in the prior art: Existing methods for monitoring and early warning of saline soil subgrades mostly focus on the identification of single parameters or single diseases, lacking a unified characterization and hierarchical identification method for the coupled evolution process of salt swelling and dissolution, making it difficult to accurately identify the risks of disaster occurrence, development and mutual transformation. In existing technologies, there is insufficient connection between early warning results and engineering treatment measures. A linkage control mechanism has not yet been formed for execution units such as drainage, electroosmosis salt migration, pumping, grouting, thermal insulation and regulation, and traffic load restriction, making it difficult to achieve direct driving of treatment actions by risk classification results. Existing technologies generally lack feedback correction and anti-induced constraint mechanisms in the treatment process, making it difficult to avoid the problem of suppressing one disaster but inducing another. They cannot meet the needs of dynamic regulation and closed-loop control of disaster risks in saline soil subgrades.

[0007] To address the aforementioned technical problems, this invention provides a method for electroosmotic linkage control of salt swelling-dissolution disaster in saline soil roadbeds, such as... Figure 1 As shown, the controller performs the following steps of the method: S1. Construct a multi-level water-heat-salt-force state sensing system for saline soil subgrade.

[0008] Multiple control units were divided along the longitudinal, transverse, and depth directions of the saline soil subgrade. Multi-source state data of each control unit were collected in the subgrade surface or ditch area to construct a water-thermal-salt-force state vector to characterize the state gradient information of each control unit along the depth direction.

[0009] Preferably, each control unit includes at least a shallow salt swelling sensitive zone, a middle salt migration transition zone, and a deep dissolution sensitive zone along the depth direction. The shallow salt swelling sensitive zone is used to acquire state data related to salt crystallization swelling, the middle salt migration transition zone is used to acquire state data related to water-salt migration, and the deep dissolution sensitive zone is used to acquire state data related to salt cementation, dissolution, softening, and structural settlement.

[0010] Preferably, the multi-source condition data includes infiltration data collected from the roadbed surface or ditch area. Drainage flow rate and environmental boundary condition data; the environmental boundary condition data includes moisture content. ,temperature Electrical conductivity Matrix suction pore water pressure Vertical displacement Settlement Load response and groundwater level .

[0011] The controller performs time synchronization, outlier removal, noise filtering, missing data compensation, and normalization on the collected multi-source state data to construct the first... Each control unit at time Water-heat-salt-force state vector The expression is:

[0012] in, Moisture content; For temperature; Electrical conductivity; For matrix suction; Pore ​​water pressure; This refers to the amount of swelling. This refers to the settlement amount; For load response; This refers to the infiltration rate; Groundwater level; This represents the drainage flow rate.

[0013] Preferably, the controller refreshes the water-heat-salt-force state vector according to a fixed sampling period or an event-triggered method. This is to form a continuous temporal state matrix of the roadbed service process.

[0014] S2. Identify the salt crystallization front and the dissolution wetting front, and extract the evolution characteristics of the two fronts.

[0015] like Figure 2As shown, the controller identifies the salt crystallization front in each control unit based on the water-thermal-salt-force state vector. and dissolution wetting front Among them, the salt crystallization leading edge This indicates the main interface location where pore fluid salts transition from the dissolved state to the crystalline state, which can be determined based on the conductivity gradient. Temperature gradient Matrix suction gradient and bulging rate The fusion characteristics were determined; the dissolution wetting front was determined. This indicates the location of the main interface where infiltration, wetting, salt cementation, dissolution, and softening propagate into the subgrade. It can be determined based on the moisture content gradient. rate of decrease in conductivity pore water pressure change rate and settling rate The fusion characteristics were determined.

[0016] Preferably, the control unit is at time Salt crystallization vanguard and dissolution wetting front They are represented as follows:

[0017]

[0018] in, and Forward inversion function based on weighted fusion, state estimation, or trained model; The rate of expansion; The value represents the settling rate.

[0019] Furthermore, the controller calculates the salt crystallization front based on the changes in the front position at consecutive time points. With the dissolution and wetting front Salt crystallization front migration speed and the migration speed of the dissolution and wetting front When the change Less than the preset distance threshold At that time, the corresponding control unit is marked as a double-forward coupling unit to characterize the highly sensitive region where salt swelling and solution trapping may transform into each other or be coupled and amplified.

[0020] S3. Construct salt swelling risk index, solution collapse risk index and coupling transformation index.

[0021] The controller is based on a water-thermal-salt-mechanical state vector and includes the salt crystallization front. Dissolution and wetting front Salt crystallization front migration speed and the migration speed of the dissolution and wetting front The two forward parameters are used to calculate the salt inflation risk index. Dissolution risk index and salt swelling-dissolution coupling conversion index .

[0022] In this invention, the fixed normalization scale is used to represent the numerical scale when the corresponding parameter enters a significant change range, and is not a threshold for determining a catastrophic state; when the parameter exceeds this scale, it is treated as a 1-limited amplitude. All normalization scales remain unchanged throughout the entire control cycle of this invention and are not adjusted retrospectively based on the calculation results of a single control unit.

[0023] To eliminate differences in dimensions and orders of magnitude among the monitoring parameters, all parameters involved in the weighted calculation of the risk index were subjected to amplitude-limited normalization. Therefore, in calculating the salt heave risk index... Dissolution risk index and salt swelling-dissolution coupling conversion index During the process, for non-negative parameters, a larger value indicates a higher risk. Its normalized value Calculate according to the following formula:

[0024] For parameters that may have positive or negative directions but whose risk is characterized by their magnitude of change. Its normalized value Calculate according to the following formula:

[0025] in, This is the fixed normalization scale for the corresponding parameters. After normalization, each parameter is within the range of 0 to 1. When a parameter exceeds the corresponding fixed normalization scale, its normalized value is set to 1 to avoid a single outlier parameter from excessively dominating the comprehensive risk index.

[0026] Preferably, the salt swelling risk index Based on the salt crystallization pressure potential, electrical conductivity gradient, temperature gradient, bulging rate, and freeze-thaw cycle intensity, its expression is as follows:

[0027] in, The crystallization pressure potential of normalized salts; This represents the normalized conductivity gradient. This represents the normalized temperature gradient. Normalized inflation rate; Normalized freeze-thaw cycle strength; These are the weighting coefficients.

[0028] In this invention, It directly characterizes the driving force of the transformation of the pore solution from a supersaturated state to a salt crystal precipitation state, and therefore assigns the highest weight of 0.32; It reflects the degree of salt enrichment along the depth direction. The direct deformation response after salt crystallization is reflected, and both are assigned a weight of 0.20. Reflecting changes in salt solubility and crystallization conditions, a weight of 0.18 is assigned. This is a cumulative correction parameter, assigned a weight of 0.10. The sum of the above weights is 1, and it is used consistently in all control units.

[0029] Furthermore, the salt crystallization pressure potential The preferred method is to calculate according to the following formula:

[0030] in, It is the gas constant; This represents the molar volume of the salt crystal. This refers to the pore fluid salt concentration. For temperature The salt saturation concentration below.

[0031] Normalized salt crystallization pressure potential Normalization is performed according to the following fixed scale:

[0032] Taking the direction of increasing subgrade depth as positive, the position of the candidate salt crystallization front is defined. The monitoring depths above and below are respectively and The normalized conductivity gradient Calculate according to the following formula:

[0033] The normalized temperature gradient Calculate according to the following formula:

[0034] when When the electrical conductivity of the soil layer below the candidate forward is higher than that above it, it indicates that there is a strong supply of saline pore water below the forward; when When the temperature of the soil layer above the candidate front is lower than that of the soil layer below, it creates temperature conditions conducive to the migration and crystallization of saline pore water to the lower, shallower layers. Therefore, only the conductivity gradient and temperature gradient that are positive along the depth direction are included in the salt swelling risk index; when the corresponding gradient is less than or equal to zero, its contribution to the salt crystallization direction is zero.

[0035] The influence of actual temperature and pore liquid salt concentration at the salt crystallization front on the crystallization driving force has been studied using the salt crystallization pressure potential. Characterization. The normalized conductivity gradient and normalized temperature gradient Used to characterize the directional spatial conditions between the high-salt pore water supply region and the low-temperature crystallization region.

[0036] Normalized tumescence rate and normalized freeze-thaw cycle intensity Calculate using the following formula:

[0037] in, This represents the number of effective freeze-thaw cycles or salt crystallization cycles within the corresponding statistical period.

[0038] Preferably, the solution risk index Based on the migration velocity of the dissolution wetting front, the salt cementation loss rate, the change in pore water pressure, the settling rate, and the dynamic stiffness attenuation, its expression is as follows:

[0039] in, Normalized dissolution wetting front velocity; Normalized salt cementation loss rate; This represents the normalized change in pore water pressure. Normalized settling rate; This represents the normalized dynamic stiffness attenuation. These are the weighting coefficients.

[0040] In this invention, It directly reflects the extent to which the wet and eroded zone extends into the roadbed, and is assigned a weight of 0.24. Reflecting the degree of direct loss of the salt cement structure, a weight of 0.22 is assigned. Reflecting the direct deformation response after structural softening, a weight of 0.20 is assigned. Reflecting the hydraulic disturbance during the infiltration process, a weight of 0.18 is assigned. The weight 0.16 is assigned to reflect the hysteretic change in the overall bearing capacity of the roadbed. The sum of the above weights is 1, and this weight is fixed in all control units.

[0041] Furthermore, the salt cementation loss rate The preferred method is to calculate according to the following formula:

[0042] in, For time intervals; To prevent corrections where the denominator is zero; For the control unit at time The electrical conductivity.

[0043] The above parameters are normalized according to the following fixed scale:

[0044] Preferably, the salt swelling-dissolution coupling conversion index This is a dimensionless comprehensive index used to simultaneously characterize salt swelling risk, solution collapse risk, spatial proximity of the two fronts, and the degree of dynamic transformation of catastrophic states. Its expression is:

[0045] in, Forefront of salt crystallization With the dissolution and wetting front The degree of overlap; Sensitivity to catastrophic state transitions; These are the weighting coefficients.

[0046] Furthermore, overlap The preferred method is to calculate according to the following formula:

[0047] in, To control the thickness to the target; The salt crystallization forward and dissolution wetting front All are located within the target control thickness Within, therefore overlap It is a dimensionless quantity between 0 and 1; when the frontal spacing decreases, the overlap... An increase indicates a stronger spatial coupling between the salt swelling zone and the solution sinking zone.

[0048] In this invention, the salt swelling risk index Dissolution risk index overlap Sensitivity to disaster state transition All are dimensionless quantities ranging from 0 to 1, therefore the salt swelling-dissolution coupling transformation index It is also a dimensionless quantity with a value range of 0 to 1.

[0049] In this invention, the weighting coefficient It is used consistently across all control units and under different disaster conditions, with values ​​of 0.35, 0.25, 0.20, and 0.20 respectively. Among these, the salt heave risk index... and the risk index of dissolution As two types of direct catastrophic risks, they together account for a weight of 0.60; overlap. Sensitivity to disaster state transition Spatial coupling and dynamic transformation are represented respectively, with a combined weight of 0.40. In the sulfate-chloride type saline soil subgrade targeted by this invention, the swelling and microcracks generated by salt crystallization provide channels for subsequent infiltration dissolution and structural softening; therefore, the salt swelling risk index... The weight is set to 0.35, and the collapse risk index is... The weight of is set to 0.25; the spatial overlap feature and the dynamic transformation feature are mutually independent coupled correction factors with the same level of action. The weight of both is set to 0.20, the sum of the weights is 1, and they are used in all control units.

[0050] In this invention, the salt swelling-dissolution coupling conversion index It is not a measurement of a single physical parameter, but a dimensionless comprehensive state index used for disaster state ranking and classification. Among them: Salt inflation risk index Reflects the degree of danger inherent in salt swelling; solution collapse risk index Reflects the degree of danger of the solution itself; overlap Reflecting the leading edge of salt crystallization and dissolution wetting front Spatial proximity; sensitivity to catastrophic state transition It reflects the dynamic degree of relative migration between the two forwards and the synchronous change of the two risks.

[0051] When the salt swelling risk index or risk index of dissolution When the overlap increases, it indicates an increased risk of a single catastrophic event; when the overlap... An increase indicates that the two interfaces are spatially close; when the sensitivity to catastrophic state transitions... An increase in the value indicates a faster relative migration of the front and a faster synchronous change in both risks. Therefore, the salt swelling-dissolution coupling transformation index... The larger the value, the higher the likelihood that salt swelling and solution trapping will evolve from a relatively independent state to a spatially coupled and dynamically transformed state.

[0052] The sensitivity to the transition of the disaster state Used to characterize the salt crystallization front With the dissolution and wetting front relative migration degree and And the degree of synchronous change of salt swelling risk and solution collapse risk in adjacent risk update cycles.

[0053] To eliminate the dimensional difference between the forward migration speed term and the risk index change rate term, both are dimensionless, as follows: set up The time interval between two consecutive risk updates. To control the thickness of the target, the relative migration component of the forward. Calculate according to the following formula:

[0054] Risk synchronous change component Calculate according to the following formula:

[0055] The rate of change of the risk index is calculated using the difference between adjacent risk update times, i.e.:

[0056] The sensitivity to the transition of the disaster state Calculate according to the following formula:

[0057] in, , and All are dimensionless quantities, with values ​​ranging from 0 to 1; coefficients 0.5 and 0.5 respectively indicate that the relative migration characteristics of the front and the synchronous change characteristics of dual risks have the same level of role in the identification of disaster state transition.

[0058] Preferably, the weighting coefficient , and It can be determined based on historical disaster samples, expert experience, entropy weight analysis, Bayesian updates, or machine learning training results.

[0059] S4. Based on the hysteresis grading mechanism, complete the disaster status identification and early warning output.

[0060] The controller is based on the salt swelling risk index. Dissolution risk index Salt swelling-dissolution coupling conversion index The hysteresis classification method is used to classify the disaster status of each control unit into one or more of the following: stable state, salt expansion-dominated state, solution collapse-dominated state, composite disaster state, and conflict resolution state.

[0061] Preferably, the controller sets an upward threshold and a downward threshold for each risk level. When the risk index increases, the upward threshold is used to trigger a higher-level warning. When the risk index decreases, a downward threshold lower than the upward threshold is used to cancel or reduce the warning level, so as to suppress frequent level jumps caused by short-term rainfall, snow melting, diurnal temperature differences or traffic load disturbances.

[0062] Furthermore, when the salt swelling risk index Exceeding the salt swelling dominant threshold and the collapse risk index When the value is below the collapse-dominant threshold, it is determined to be a salt swelling-dominant state; when the collapse risk index is below the threshold, it is determined to be a salt swelling-dominant state. Exceeding the dominant threshold for dissolution and the salt swelling risk index When the value is below the salt swelling dominance threshold, it is determined to be in a collapse-dominant state; when the salt swelling risk index... and the risk index of dissolution Simultaneously, a higher or the aforementioned salt swelling-dissolution coupling conversion index When the coupling threshold is exceeded, the situation is determined to be a compound disaster state; when a certain treatment action causes a rapid increase in the incremental risk of non-target disasters, the situation is determined to be a treatment conflict state. The controller outputs the identification results as risk level, risk type, and corresponding recommended treatment mode.

[0063] In this invention, the salt swelling dominance threshold and the solution collapse dominance threshold are both 0.60, the low-risk threshold for non-targeted catastrophic events is 0.30, and the coupling threshold is 0.55. Salt swelling risk index. Dissolution risk index and salt swelling-dissolution coupling conversion index All are dimensionless quantities in the range of 0 to 1. 0.60 is set as the salt expansion dominance threshold and solution collapse dominance threshold for a single disaster, used to confirm that the comprehensive risk contribution of a certain type of disaster has exceeded 60% of the full risk scale; 0.30 is set as the low-risk threshold for non-target disasters, used to confirm that another type of disaster has not yet formed a significant dominant role; 0.55 is set as the coupling threshold, which is higher than the median value of 0.50 but lower than the dominance threshold of 0.60, used to identify the coupling trend formed by the spatial approximation of two fronts and the synchronous changes of two risks before a single disaster reaches complete dominance.

[0064] S5. Construct a treatment linkage control vector for multiple actuators and perform optimization solution.

[0065] The controller generates a response and control vector based on the disaster state. The expression is:

[0066] in, This refers to the drainage flow rate; This refers to the opening degree of the drain valve; The electric field strength is the electroosmotic electric field strength. This is the electroosmotic polarity switching cycle; To pump out the flow rate; This refers to the grouting flow rate; This refers to the grouting pressure. For heat preservation or heating power; This is the traffic load limit level.

[0067] Preferably, when the disaster state is dominated by salt expansion, the treatment linkage control vector Prioritize the use of drainage, electroosmotic salt migration, pumping, and insulation or heating control to suppress shallow crystallization expansion and water-salt upward migration; when the disaster state is dominated by solution collapse, the treatment linkage control vector... Prioritize the use of tiered drainage, low-pressure grouting, and traffic load control to suppress structural collapse and softened layer expansion; when the disaster state is a complex disaster state, the treatment linkage control vector... The controller is generated in the following order: maintaining the moisture content of the control unit within a preset safe range, electroosmotic salt migration, pumping out the collected salt solution, and zonal low-pressure grouting. When the disaster state is a treatment conflict state, the controller stops the treatment actions that cause non-target disaster risk to induce an increment exceeding a preset threshold, and re-solves the treatment linkage control vector.

[0068] Furthermore, the controller employs a rolling time-domain optimization method to optimize the treatment linkage control vector under constraints such as moisture content safety window, deformation rate (bulging rate constraint and settling rate constraint), actuator output limit constraint, and anti-induced constraint. Perform rolling optimization to solve the problem.

[0069] Furthermore, the controller completes a rolling control cycle every 2 hours, with a prediction time domain of 12 hours, a prediction step size of 2 hours, and a prediction step count H=6. The current control time is denoted as... t ,by t+k Indicates the prediction of the first time domain k One predicted time, of which .

[0070] The controller will determine the risk status at the current control moment. Recorded as:

[0071] Treatment linkage control vector Each component is normalized according to its respective lower and upper allowable output limits, resulting in the normalized control vector:

[0072] For the electroosmotic polarity switching cycle For control components other than those specified, the normalized value is calculated using the following formula:

[0073] Due to the electroosmotic polarity switching cycle The shorter the length, the stronger the control effect. Its normalized value is calculated according to the following formula:

[0074] The values ​​of each normalized control component range from 0 to 1.

[0075] The controller uses risk status and treatment linkage control data from the most recent 12 completed rolling control cycles to establish a first-order affine prediction model for the risk status, namely:

[0076] in, For the natural evolution of risk; This is the risk state transition matrix; This is a response matrix for coordinated control.

[0077] Let the risk status be defined for the 12 most recent completed control cycles. and normalized control vector They are respectively to and to Then construct the matrix for:

[0078] The prediction model parameters are calculated according to the following formula:

[0079] in, This is a regularization term used to ensure the stability of matrix inversion.

[0080] For any candidate treatment and control scheme, the current measured risk status is used as the initial value, that is:

[0081] And calculate it recursively in the prediction time domain according to the following formula:

[0082] in, This means that the calculation result is limited to the range of 0 to 1.

[0083] Therefore, we can conclude that:

[0084] in, , , Predicted time The predicted values ​​of the salt swelling risk index, the predicted value of the solution collapse risk index, and the predicted value of the salt swelling-solution collapse coupling transformation index.

[0085] Control vector change The average change magnitude of each component of the normalized control vector relative to the previous prediction time. Calculate according to the following formula:

[0086] therefore, , , , and All of them are dimensionless scalars with values ​​ranging from 0 to 1.

[0087] The average objective function of the rolling optimization for:

[0088] in, To predict the length of the time domain; For the predicted time The predicted value of the salt inflation risk index; For the predicted time The predicted value of the solution risk index; For the predicted time Predicted value of salt swelling-solution trap coupling transformation index; For the predicted time The change in the treatment linkage control vector relative to the previous control cycle; For the predicted time The costs of construction disturbance or traffic disruption; These are the weighting coefficients.

[0089] In this invention, salt swelling risk, solution collapse risk, and coupling transformation risk are risk suppression targets, with a total weight of 0.80. Salt swelling risk and solution collapse risk are two types of direct catastrophic risks, each with a weight of 0.25. The coupling transformation index simultaneously characterizes the proximity of the two-front space and the dynamic transformation between the two types of catastrophes, with a weight of 0.30. The control vector change is used to limit abrupt changes in the actuator output within adjacent control cycles, and the cost of construction disturbance or traffic impact is used to limit the intensity of treatment and its engineering impact, both with a weight of 0.10. The sum of the weights of the above five items is 1, and they are used consistently across all control units, all prediction cycles, and under all three types of catastrophic conditions.

[0090] For each candidate treatment linkage control scheme, the controller uses the first control step vector of that scheme as the candidate output at the current control moment and keeps its main control parameters unchanged in the prediction time domain; the electroosmotic polarity is executed according to the switching cycle specified by the candidate scheme. The controller calculates the objective function value corresponding to each candidate scheme, and selects the scheme with the smallest objective function value from the candidate schemes that satisfy the water content safety window constraint, bulging rate constraint, settling rate constraint, actuator output limit constraint, and anti-induced constraint, and uses its first control step vector as the actual output at the current control moment. In the next rolling control cycle, state data is re-acquired, the prediction model is updated, and the solution is re-solved.

[0091] Preferably, the constraints include at least:

[0092]

[0093]

[0094]

[0095]

[0096] in, This is the current control moment; To predict the first in the time domain Prediction step number; To predict the corresponding prediction time within the time domain; and This is the boundary of the moisture content safety window; The upper limit of the permissible bulging rate; The upper limit of the allowable settlement rate; and These are the lower and upper limits of the allowable output for each component of the treatment linkage control vector, respectively. For the predicted time The predicted moisture content of the control unit; For the predicted time The predicted value of the bulging rate of the control unit; For the predicted time The predicted settling rate of the control unit; For the predicted time Predicted values ​​of the treatment linkage control vector; Predicted time The predicted incremental value of non-targeted disaster risk; This is the anti-induction threshold.

[0097] The anti-inducement constraint is used to limit any treatment action from significantly inducing another catastrophic event while suppressing one catastrophic event, thereby ensuring that the control strategy has linkage and coordination.

[0098] S6. Perform the treatment action and provide feedback for correction.

[0099] like Figure 3 As shown, the controller uses the obtained treatment linkage control vector. After each round of control execution, the controller re-acquires the water-heat-salt-force state vector after treatment and calculates the salt expansion risk index. Dissolution risk index Salt swelling-dissolution coupling conversion index The risk reduction rate and the incremental risk induced by non-targeted disasters, the classification threshold of disaster status and the linkage control vector for treatment. The solution parameters are adaptively adjusted.

[0100] Preferably, the risk reduction rate can be expressed as:

[0101]

[0102]

[0103] in, , , These are the risk reduction rates of the salt swelling risk index, the solution collapse risk index, and the salt swelling-solution collapse coupling conversion index, respectively. , , These are the risk indices corresponding to the treatment prior to the procedure. , , These are the risk indices corresponding to the treatment.

[0104] When the risk reduction rate is lower than the preset lower limit or the incremental risk of non-target disaster is higher than the preset upper limit, the controller automatically increases the control constraint level or switches the treatment mode again to form a closed-loop control process of "state perception - front-end identification - risk classification - linkage control - feedback correction".

[0105] The above-mentioned electroosmotic linkage control method relies on the electroosmotic linkage control system proposed in this invention, such as... Figure 4 and Figure 5 As shown, the system comprises a field roadbed object layer, a control and calculation layer, and an execution and feedback layer. Among them: The on-site roadbed object layer includes a shallow salt swelling sensitive zone, a middle salt migration transition zone, and a deep solution sinking sensitive zone set in the roadbed monitoring and control object. Moisture content sensors, temperature sensors, conductivity sensors, and vertical displacement monitoring units are deployed in the shallow salt swelling sensitive zone. Moisture content sensors, conductivity sensors, matrix suction sensors, and pore water pressure sensors are deployed in the middle salt migration transition zone. Pore water pressure sensors, settlement monitoring units, load response monitoring units, and necessary groundwater level monitoring units are deployed in the deep solution sinking sensitive zone.

[0106] The execution and feedback layer includes a drainage branch, an electrically controlled valve, an electroosmotic electrode, and a salt collection and pumping well as a water and salt regulation unit; a grouting pipe as a structural treatment unit; and a shallow heating zone and a traffic guidance unit as auxiliary regulation units.

[0107] The control and computing layer includes a data transmission unit that establishes a communication connection with the on-site roadbed object layer, as well as a central control platform and a front identification and status analysis module, a risk calculation and graded early warning module, and a linkage control and feedback correction module that constitute the edge controller.

[0108] The linkage control and feedback correction module outputs the treatment linkage control vector. It is used to output control commands to the execution unit and the feedback layer.

[0109] Furthermore, the controller, based on the solved treatment linkage control vector... The system involves the coordinated control of at least two of the following: drainage branch, electrically controlled valve, electroosmotic electrode, salt collection and drainage well, grouting pipe, shallow heating zone, and traffic guidance unit, to implement treatment actions.

[0110] Furthermore, the treatment linkage control vector Drainage flow rate and drain valve opening Corresponding control of drainage branches and electrically controlled valves, electroosmotic electric field strength and electroosmotic polarity switching cycle Corresponding control of electroosmotic electrodes, pumping flow rate Corresponding control of salt collection and drainage wells, grouting flow rate and grouting pressure Corresponding control of grouting pipe, insulation or heating power Corresponding control of shallow heating zones, traffic load limit levels Corresponding traffic guidance control unit.

[0111] Furthermore, the modules can be integrated into the same industrial controller, embedded edge computing terminal, or central control platform, or they can be operated collaboratively using a distributed architecture.

[0112] By employing the above technical solution, the present invention provides a method and system for electroosmotic linkage control of salt swelling-dissolution disaster in saline soil roadbeds, which has at least the following beneficial effects: 1. This invention introduces salt crystallization front, dissolution wetting front, as well as salt swelling risk index, solution collapse risk index and salt swelling-solution collapse coupling transformation index, to achieve unified identification of salt swelling-solution collapse disaster state of saline soil subgrade. Compared with the method of relying on a single monitoring parameter for judgment, it is more convenient to reflect the evolution process of the two types of disasters and their mutual transformation relationship.

[0113] 2. This invention links the risk classification results with treatment actions such as drainage, electroosmotic salt migration, pumping, grouting, thermal insulation adjustment, and traffic load restriction, and sets feedback correction and anti-induced constraints to enable the treatment process to be adjusted according to changes in the roadbed condition, thereby enhancing the connection between early warning results and engineering treatment.

[0114] 3. This invention can identify salt crystallization fronts and dissolution wetting fronts based on multi-source state perception results, accurately calculate salt swelling risk index, dissolution risk index and salt swelling-dissolution coupling transformation index, and generate appropriate treatment linkage control vectors under anti-induced constraints, so as to realize dynamic control and closed-loop adjustment of salt swelling risk of saline soil subgrade. Attached Figure Description

[0115] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the electroosmotic linkage control method in this invention; Figure 2 This is a schematic diagram of the double-forward evolution and catastrophic state identification in this invention; Figure 3 This is a schematic diagram of the treatment linkage control and feedback correction in this invention; Figure 4 This is a schematic diagram of the electroosmotic linkage control system in this invention; Figure 5 This is a schematic diagram of the spatial layout of the on-site roadbed object layer and the execution and feedback layer in this invention. Detailed Implementation

[0116] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0117] Example 1: Risk classification, early warning, and coordinated treatment in salt-induced scenarios.

[0118] This embodiment selects a section of saline soil test subgrade as the application object. The subgrade section is 60m long, 8.5m wide at the top, and 2.4m high. The fill material is sulfate-chloride type saline silty clay. Sampling and testing showed that the total salt content of the fill material was 2.31%, of which... The quality fraction was 51.8%. The quality fraction was 29.4%. The mass fraction is 10.7%, with the remainder being small amounts of carbonates and miscellaneous salts; the maximum dry density is The optimum moisture content is 12.1%, and the compaction degree is controlled at 95%. The groundwater level in this roadbed section is stable at [missing information]. There has been no effective rainfall in the last 24 hours, and the road surface evaporation is... The daytime high temperature is The lowest temperature at night was It possesses typical shallow evaporation and concentration, and nighttime cooling and salt precipitation conditions.

[0119] Six control units are arranged at equal intervals along the longitudinal direction of the roadbed section, and are respectively denoted as... Each control unit is divided into shallow salt swelling sensitive zone, middle salt migration transition zone, and deep solution sinking sensitive zone in the depth direction. Moisture content sensor, temperature sensor, conductivity sensor, matrix suction sensor, and pore water pressure sensor are installed at burial depths of 0.20m, 0.40m, 0.80m, 1.10m, and 1.50m, respectively. Vertical displacement monitoring units and settlement monitoring units are deployed at the road surface and in the 1.10-1.50m depth range, respectively. Drainage branches and electrically controlled valves are set on both sides of the road shoulder. Electroosmotic electrodes are set on both sides of the control unit, with an electrode insertion depth of 1.2m and a spacing of 2.0m. Salt collection and drainage wells are set at the roadbed shoulder. A shallow heating strip, preferably a low-power heating strip, is buried 0.15m below the road surface. A traffic guidance unit is also set up to output speed limit and load limit commands. Each sensor has a sampling period of 10 minutes. The controller uses a 12-hour rolling window to filter, normalize, and update the data. The treatment linkage control vector is re-solved every 2 hours.

[0120] After 24 hours of continuous monitoring, the controller calculated the salt swelling risk index for each of the six control units. Dissolution risk index Salt swelling-dissolution coupling conversion index The results are shown in Table 1.

[0121] Table 1 Initial Risk Calculation Results for Control Units

[0122] As can be seen from Table 1, The salt heave risk index is the highest, and the average heave rate of the road surface in the past 6 hours has reached [a certain value]. The rate exceeds the first-level salt swelling warning rate threshold set in this embodiment. Therefore, the controller is selected. This is the key control object in this embodiment, and it is calculated in detail.

[0123] The controller first reads Multi-source monitoring data at the end of the current scrolling window. The original state data of this control unit at each burial depth are shown in Table 2.

[0124] Table 2 Control Unit Initial state data

[0125] at the same time, Average road surface bulge rate over the past 6 hours for:

[0126] Average settlement rate in the depth range of 1.10–1.50 m for:

[0127] The dynamic stiffness decreased from 186 MPa to 171 MPa in the last 12 hours, corresponding to a dynamic stiffness attenuation rate. for:

[0128] In this embodiment, the controller uses the following fusion recognition model to determine the salt crystallization front. and dissolution wetting front ,Right now:

[0129]

[0130] in, For the first Candidate depth positions Salt crystallization front identification score; For the first Candidate depth positions Dissolution and wetting front identification score; candidate depth location For two adjacent monitoring depths and The center position of the interface between layers, that is:

[0131] For any of the following state parameters: water content, temperature, electrical conductivity, matrix suction, and pore water pressure Its position at candidate depth The gradient at a given point is calculated using the following formula:

[0132] , , , and They represent the first The conductivity gradient, temperature gradient, matrix suction gradient, water content gradient, and pore water pressure gradient at each candidate depth location, normalized according to the design range. and These represent the bulging rate and settlement rate after normalization according to the design range, respectively. This indicates that only the positive value is taken.

[0133] At the forefront of salt crystallization During identification, conductivity gradient Directly reflecting abrupt changes in salt enrichment along the depth direction, assigned a weight of 0.35; temperature gradient and matrix suction gradient These reflect the conditions for salt crystallization and the conditions for water-salt migration, respectively, and are each assigned a weight of 0.25. The swelling rate is a deformation response after salt crystallization and is assigned a weight of 0.15.

[0134] In the dissolution and wetting front During identification, moisture content gradient The component directly reflects the spatial abrupt change at the wetting interface and is assigned a weight of 0.35; the component representing the decrease in conductivity... and pore water pressure gradient These reflect salt dissolution and changes in hydraulic state, respectively, and are each assigned a weight of 0.25. The settling rate is the deformation response after salt cementation weakens, and is assigned a weight of 0.15.

[0135] Both identification models are weighted according to the hierarchical effect of direct interface representation quantities, process-related quantities, and deformation response quantities. The sum of the weights in each model is 1, and they are used in a fixed manner in all control units and under different disaster states.

[0136] Each gradient quantity and deformation rate is dimensionless according to a fixed normalization scale, i.e.:

[0137] In the salt crystallization front identification formula , and Take the normalized value corresponding to the gradient magnitude; in the formula for identifying the dissolution wetting front. This is used to extract the normalized component of conductivity that decreases along the depth direction, where:

[0138] To eliminate the differences in dimensions and orders of magnitude between different monitoring parameters, and considering the monitoring layer spacing, sensor design range, and the variation range of monitoring parameters under the three operating conditions of salt expansion-dominated, solution collapse-dominated, and combined disasters adopted in this embodiment, the fixed normalization scale of each parameter is set as follows:

[0139] The above values ​​are only used to convert state parameters of different dimensions into dimensionless values ​​that can be compared by weighting, and are not used as early warning thresholds for salt swelling, dissolution, or complex disasters, and remain unchanged in all embodiments and all control units.

[0140] The controller compares the candidate depth positions separately. Salt crystallization forward identification score Identification score of dissolution and wetting front And the candidate depth position corresponding to the maximum recognition score. The position is determined to be the corresponding forward position, that is:

[0141] Based on the data in Table 2, the following can be calculated: The fusion recognition results of the interfaces between the main layers are shown in Table 3. At each forward recognition time... The controller acquires status data at each monitoring depth within the same control unit, after time synchronization, outlier removal, noise filtering, and missing value compensation. The monitoring depths set in this embodiment are as follows:

[0142] The center of the interlayer interface between two adjacent monitoring depths is used as the candidate leading edge position. ,Right now:

[0143] Therefore, we can conclude that:

[0144] The data at the center burial depth of different interlayer interfaces are not directly measured by additional sensors, but are calculated based on synchronous monitoring data at adjacent monitoring depths above and below the interlayer interface. This applies to any of the following state parameters: water content, temperature, electrical conductivity, matrix suction, and pore water pressure. He is a candidate for striker. The gradient at a given point is calculated using the following formula:

[0145] in, and Forward recognition moment The effective values ​​of state parameters obtained at two adjacent monitoring depths.

[0146] The bulging rate and settling rate are calculated using the average rate of change over a continuous 6 hours prior to the front identification time, and are respectively calculated according to the following formulas:

[0147] After processing the gradient values, bulging rates, and settling rates at the center locations of each interlayer interface according to the aforementioned normalization rules, and substituting them into the salt crystallization front identification scoring formula and the dissolution wetting front identification scoring formula, the positions of each candidate front are obtained. Recognition score and .

[0148] The controller compares the identification times of the same forward separately. Positions of each candidate forward The recognition score will be used to identify the salt crystallization front. Candidate forward positions corresponding to the maximum value The location was identified as the salt crystallization front, and the wetting front was identified and scored. Candidate forward positions corresponding to the maximum value The location was identified as the dissolution and wetting front, namely:

[0149] The identification score is determined by multiple source state parameters such as conductivity, temperature, matrix suction, water content, pore water pressure and deformation response. The front position is updated only when the maximum identification score is the only maximum value, so as to avoid misjudgment of the front position due to instantaneous fluctuations of a single monitoring parameter.

[0150] The controller updates the frontier identification results every 2 hours, and constructs a frontier position time series based on 13 sets of frontier positions from the most recent 24 hours. To reduce the impact of discrete changes in frontier positions between adjacent control cycles on migration velocity calculations, the migration velocities of salt crystallization fronts and dissolution wetting fronts are determined using the linear fitting slopes of the 24-hour frontier position time series, i.e.:

[0151] Among them, the moment of forward recognition In days The average of the 13 forward identification moments. and The values ​​represent the average positions of the salt crystallization front and the dissolution wetting front over a 24-hour period, respectively; the resulting salt crystallization front migration rate... and the migration speed of the dissolution and wetting front The units are all m / d, which represent the average migration trend of the corresponding forward within a 24-hour window, rather than the instantaneous migration velocity obtained solely from the position difference between two adjacent moments.

[0152] Table 3 Control Unit Two-forward recognition results

[0153] As shown in Table 3, the highest salt crystallization front identification score was achieved in the salt crystallization front identification results. The value is 0.709, corresponding to the candidate striker position. The score was 0.30m; the next highest score was for salt crystallization front recognition. The score is 0.331, and the difference between the two is 0.378. In the results of dissolution and wetting front identification, the highest dissolution and wetting front identification score is... The value is 0.761, corresponding to the candidate forward position. The score was 0.95m; the next highest score was for the identification of the dissolution and wetting front. The score is 0.386, and the difference between the two is 0.375. Therefore, both types of forward identification scores have a unique maximum value, based on which the salt crystallization forward is determined. Dissolution and wetting front Based on the 13 sets of effective front positions formed every 2 hours within the first 24 hours, the migration rates of the salt crystallization front and the dissolution and wetting front were calculated using the linear fitting method described above, thus obtaining the salt crystallization front migration rate. for:

[0154] Migration speed of the dissolution and wetting front for:

[0155] Subsequently, the controller calculates Salt inflation risk index Considering that this section of the road is... For the master controlling salt, the crystallization pressure potential of the salt is calculated using the master controlling salt crystallization pressure potential model. ,Right now:

[0156] in, It is the gas constant; This represents the molar volume of the salt crystal. This refers to the pore fluid salt concentration. For temperature The salt saturation concentration below.

[0157] At the forefront of salt crystallization The pore fluid salt concentration was measured from a nearby pore fluid sample. Corresponding temperature Salt saturation concentration at this temperature Molar volume of salt crystals Substituting, we get:

[0158] In this embodiment, the salt swelling risk index is expressed as follows:

[0159] in, The crystallization pressure potential of normalized salts; This represents the normalized conductivity gradient. This represents the normalized temperature gradient. Normalized bulging rate; This represents the normalized freeze-thaw cycle intensity.

[0160] The weighting coefficients are:

[0161] Normalized reference values ​​are taken respectively , , , Based on the aforementioned 0.30m interlayer interface data, we can conclude that:

[0162] Due to the control unit in Table 2 The electrical conductivity at 0.20 m and 0.40 m is 18.6 mS / cm and 21.8 mS / cm, respectively, therefore the candidate forward position is... for:

[0163] conductivity gradient for:

[0164] Since this gradient is greater than zero, the normalized conductivity gradient... for:

[0165] Similarly, temperature gradient for:

[0166] Normalized temperature gradient for:

[0167] The normalized value of the freeze-thaw cycle intensity of salt crystallization over the past 24 hours is:

[0168] Therefore, the salt inflation risk index for:

[0169] The controller further calculates the collapse risk index. The calculation formula is:

[0170] in, Normalized dissolution wetting front velocity; Normalized salt cementation loss rate; This represents the normalized change in pore water pressure. Normalized settling rate; This represents the normalized dynamic stiffness attenuation.

[0171] The weighting coefficients are taken as follows:

[0172] Normalized reference value for forward speed Therefore:

[0173] In the dissolution and wetting front Nearby, conductivity at the previous moment Current conductivity ,Pick , Then the salt cementation loss rate for:

[0174] With the normalized reference value set to 0.30, we get:

[0175] Changes in pore water pressure over the past 12 hours Take the reference value The normalized value of the change in pore water pressure is:

[0176] The normalized value of the settling rate is:

[0177] The normalized reference value for dynamic stiffness attenuation is taken as 0.50, resulting in:

[0178] Substituting into the solution risk index for:

[0179] Based on this, the controller calculates the salt swelling-solution trapping coupling conversion index. Its expression is:

[0180] in, Forefront of salt crystallization With the dissolution and wetting front The degree of overlap; Sensitivity to catastrophic state transitions.

[0181] The weighting coefficients are taken as follows:

[0182] The target control thickness is:

[0183] Then the overlap between the two forwards for:

[0184] Based on the normalized calculation results of the difference in forward migration speed and the rate of risk change, this embodiment takes the sensitivity to catastrophe state transition as:

[0185] Substituting the values ​​yields the salt swelling-dissolution coupling transformation index. for:

[0186] In this embodiment, the controller adopts the following identification rule: when When the control unit is in a salt swelling-dominated state, the control unit will be determined to be in a salt swelling-dominated state; when When, it is determined to be a state dominated by dissolution; when and All are relatively high, or At that time, it was determined to be a complex disaster state. Because... satisfy Therefore, the controller determines that it is in a salt expansion-dominated state and switches to the linkage treatment and solution process for the salt expansion-dominated scenario.

[0187] In this embodiment, the moisture content safety window is set as follows: The maximum allowable bulging rate is set to... The maximum allowable settlement rate is set to The threshold for non-targeted catastrophic risk-induced incremental changes is set to... .

[0188] The controller uses the current control time Starting from the rolling optimization point, the sequence of treatment linkage control vectors in the prediction time domain is solved, and the first treatment linkage control vector at the current control moment is output. ,Right now:

[0189] in, This refers to the drainage flow rate; This refers to the opening degree of the drain valve; The electric field strength is the electroosmotic electric field strength. This is the electroosmotic polarity switching cycle; To pump out the flow rate; This refers to the grouting flow rate; This refers to the grouting pressure. For heat preservation or heating power; This is the traffic load limit level.

[0190] In this embodiment, the prediction step size is 2 hours, the number of prediction steps H=6, and the corresponding prediction time domain is 12 hours. The controller calculates the prediction time of each candidate treatment linkage control scheme within 6 prediction steps according to the risk state prediction model. , , , and And calculate the corresponding objective function based on the average objective function. Values. The candidate solutions and their solution results are shown in Table 4.

[0191] Table 4. Candidate treatment linkage control schemes and solution results

[0192] Table 4 shows the traffic load restriction levels. This indicates that the speed limit will be set to 40 km / h and the allowable axle load will be reduced by 20%.

[0193] Within the current treatment cycle, the controller calculates the change in the non-targeted disaster risk index before and after treatment, and calculates the non-targeted disaster risk induced increment according to the following formula. ,Right now:

[0194] in, This is the non-target disaster risk index at the start of the current treatment cycle; The non-target disaster risk index after implementing the current treatment plan. Non-target disaster risk index Determined based on a catastrophic risk index beyond the current treatment objective. For example, during salt expansion control treatment, the non-target catastrophic risk index... Take the risk index of dissolution During the treatment and control of solution sinkholes, the non-target catastrophic risk index Salt swelling risk index .

[0195] When non-targeted catastrophic risks induce incremental changes satisfy At that time, it was believed that the current treatment plan did not produce a significant non-targeted disaster-inducing effect; when If the controller determines that the current treatment plan may lead to an increase in the risk of non-target disasters, it will mark the corresponding treatment unit as a treatment conflict state and enter the treatment parameter adjustment process.

[0196] Non-targeted disaster risk triggering threshold Set to 0.05. Since all risk indices have been normalized and their values ​​range from 0 to 1, therefore... =0.05 indicates that when the non-target disaster risk index increases by more than 5 percentage points within a single treatment cycle, the risk change is considered to have exceeded the normal calculation fluctuation range, and the treatment strategy needs to be adjusted.

[0197] Non-targeted disaster risk triggering threshold The risk index calculation error, sensor measurement fluctuations, and the range of natural risk evolution within the control period are all taken into account. In this embodiment, the threshold for inducing non-targeted catastrophic risks is set to 0.05 to avoid frequent triggering of remediation conflict states due to minor data fluctuations, while also enabling timely identification of increases in non-targeted catastrophic risks induced by remediation measures.

[0198] In this embodiment, the non-target disaster risk indices before and after treatment are as follows:

[0199] therefore:

[0200] because:

[0201] Therefore, the current treatment plan triggers a treatment conflict status determination.

[0202] As shown in Table 4, although Scheme A has a certain inhibitory effect on shallow salt precipitation, the drainage intensity is too high, predicting an increase in the risk of non-target disasters. Option C has the potential to accelerate the softening of deep structures, so it was discarded. Option C has a higher electroosmotic intensity, a higher objective function value than Option B, and larger control fluctuations. Option B simultaneously satisfies the anti-induction constraint and the minimum objective function requirement, so the controller adopts Option B as the treatment linkage control vector in this embodiment, that is:

[0203] The controller, based on this information, coordinates and controls the drainage branch, electrically controlled valves, electroosmosis electrodes, salt collection and pumping wells, grouting pipes, shallow heating zone, and traffic guidance unit, continuously executing for 36 hours. During the treatment period, the controller rereads the sensor data and corrects the control parameters every 2 hours; due to the risk index of solution subsidence throughout the 36-hour control process... The value never exceeded 0.20, and the increase was not induced by the target catastrophic risk. Since the value did not exceed 0.05, the grouting execution unit was not triggered. Therefore, we have:

[0204] At the end of each rolling control cycle, the controller updates the risk indicators based on the measured state vector. Table 5 lists the measured or updated values ​​of the salt swelling risk index, solution collapse risk index, coupling transformation index, and deformation rate at each control moment, rather than the predicted values ​​at the non-rolling optimization process. The predicted value. After 36 hours of treatment, the controller reacquired the status data and completed the latest round of risk update, the results of which are shown in Table 5.

[0205] Table 5 Control Unit Dynamic feedback results

[0206] After 36 hours of treatment, the salt crystallization leader was re-identified by the controller. The erosion front retreated from 0.30m to 0.26m. The stability at 0.98m indicates that the shallow crystallization concentration zone has been weakened and the deep wetting and softening interface has not been raised upwards.

[0207] Correspondingly, the overlap of the two forwards after treatment The concentration of the controlling salt was reduced, thus mitigating the risk of coupling transformation. Pore fluid samples were extracted at a depth of 0.26 m after treatment, and the concentration of the controlling salt was found to have decreased to [value missing]. Maintain the main control salt saturation concentration The new salt crystallization pressure potential can be obtained using the same formula. for:

[0208] This indicates a significant decrease in the driving force for shallow salt precipitation. At this point, the calculated risk indices are:

[0209] Further calculation of the risk reduction rate:

[0210]

[0211]

[0212] Salt inflation risk index The risk index of landslides decreased by 49.8%. A decrease of 32.3%, salt swelling-dissolution coupling conversion index The moisture content decreased by 36.3%. Meanwhile, during the control period, the moisture content at each burial depth remained within the range of 10.6% to 13.4%, consistently meeting the preset moisture content safety window requirements, indicating that the treatment-linked control vector used in this embodiment... It effectively suppressed the risk of salt swelling and avoided the amplification of non-target solution collapse risk caused by excessive drainage.

[0213] Therefore, under the salt swelling-dominant working condition set in this embodiment, the method of the present invention can identify the salt crystallization front and the dissolution wetting front based on the multi-source state perception results, accurately calculate the salt swelling risk index, the dissolution risk index and the salt swelling-dissolution coupling transformation index, and generate a suitable treatment linkage control vector under the anti-induced constraint, so as to realize the dynamic control and closed-loop adjustment of the salt swelling risk of saline soil subgrade.

[0214] Example 2: Risk classification, early warning, and coordinated treatment in scenarios dominated by dissolution collapse.

[0215] This embodiment was conducted on the same saline soil test subgrade as in Example 1. The test subgrade was 60m long, 8.5m wide at the top, and 2.4m high. The type of fill material, initial compaction degree, and sensor arrangement were the same as in Example 1. The overall structure of the control system used was also the same. Figure 4 The conditions were the same. To establish a solution-dominant working condition, continuous infiltration disturbance was applied to the roadbed section before the test. This included 18 hours of surface spraying with a cumulative infiltration of 31.6 mm, and 10 hours of water replenishment in the shoulder ditches, resulting in a 0.23 m rise in the ditch water level. This was followed by 6 hours of natural settling. At the end of the treatment, the ambient temperature was maintained at [temperature missing]. Without freeze-thaw action, the groundwater level rose from the initial 2.90m to 2.62m. Due to reduced shallow evaporation, while the water content, pore water pressure, and salt dissolution and migration of the middle and lower soils increased significantly, a solution-sensitive state characterized by weakened salt cementation and structural softening gradually formed inside the roadbed.

[0216] The six control units arranged longitudinally along the roadbed section are still referred to as ～ After continuously updating the monitoring data for 24 hours, the controller calculates the salt swelling risk index for each control unit. Dissolution risk index and salt swelling-dissolution coupling conversion index The results are shown in Table 6.

[0217] Table 6 Initial Risk Calculation Results for Control Units

[0218] As shown in Table 6, the control unit The risk index of solution collapse is the highest, and the average settlement rate in the past 6 hours has reached [a certain value]. This is significantly higher than the first-level solution collapse early warning rate threshold set in this embodiment. Therefore, the controller will They were selected as key control targets, and subsequent frontline identification, risk decomposition, and coordinated handling were carried out.

[0219] Controller reads control unit The original state data at the end of the current scrolling window is shown in Table 7.

[0220] Table 7 Control Unit initial state data

[0221] at the same time, The average bulge rate of the road surface in the last 6 hours is only This indicates that the shallow crystallization uplift is relatively weak; while the average settling rate in the depth range of 1.10–1.50 m is

[0222] This indicates that the softening and compressive deformation of the middle and lower structure are quite significant. The dynamic stiffness decreased from 192 MPa to 132 MPa in the last 12 hours, corresponding to a dynamic stiffness attenuation rate of:

[0223] In this embodiment, the controller uses the same forward fusion recognition model as in Embodiment 1. Perform two-forward identification, that is:

[0224]

[0225] The fusion recognition results obtained by the controller by combining continuous 12-hour time-series data and the gradient information of the interfaces between each layer are shown in Table 8.

[0226] Table 8 Control Unit Two-forward recognition results

[0227] As shown in Table 8, the highest salt crystallization front identification score was achieved in the salt crystallization front identification results. The value is 0.284, corresponding to the candidate forward position. The score was 0.30m; the next highest score was for salt crystallization front recognition. The score is 0.221, and the difference between the two is 0.063. In the results of dissolution and wetting front identification, the highest dissolution and wetting front identification score is... The value is 0.781, corresponding to the candidate forward position. The score was 0.95m; the next highest score was for the identification of the dissolution and wetting front. The score is 0.547, and the difference between the two is 0.234. Therefore, both types of forward identification scores have a unique maximum value, based on which the salt crystallization forward is determined. Dissolution and wetting front Based on the 13 sets of effective front positions formed every 2 hours within the first 24 hours, the migration speed of the salt crystallization front and the migration speed of the dissolution and wetting front were calculated according to the linear fitting method in Example 1. The migration speed of the salt crystallization front was obtained as follows: The migration rate of the dissolution and wetting front is .

[0228] Therefore, it can be seen that under the operating conditions of this embodiment, the dissolution and wetting front... The downward propagation speed is much greater than the migration speed of the salt crystallization front. The dominant degradation mechanism inside the roadbed has changed from shallow salt precipitation to infiltration and dissolution and structural softening in the middle and lower parts.

[0229] Based on this, the controller first calculates the salt swelling risk index. At the forefront of salt crystallization The pore fluid salt concentration was measured from a nearby pore fluid sample. Corresponding temperature Salt saturation concentration at this temperature Molar volume of salt crystals The salt crystallization pressure potential can be obtained from the formula. for:

[0230] This embodiment still uses Calculate the salt inflation risk index ,in .

[0231] Normalized reference values ​​are taken respectively , , , Based on the data in Table 7 for the interface between 0.20 and 0.40 m layers, then:

[0232]

[0233]

[0234]

[0235] Since there is no freeze-thaw cycle in this embodiment, the freeze-thaw cycle strength is taken as the reference value. Substituting the values ​​into the calculation yields the salt inflation risk index. for:

[0236] Subsequently, the controller calculates the collapse risk index. Its expression is:

[0237] in:

[0238] The normalized value of the dissolution wetting front velocity is:

[0239] In the dissolution and wetting front Nearby, conductivity at the previous moment Current conductivity ;Pick , The salt cementation loss rate is:

[0240] After normalization to the reference value of 0.30, we have:

[0241] The last 12 hours have been in the dissolving and wetting front. The change in pore water pressure in the vicinity is:

[0242] Taking a reference value of 10 kPa, then:

[0243] The normalized value of the settling rate is:

[0244] The normalized value of dynamic stiffness attenuation is:

[0245] Substitute into the calculation to obtain the collapse risk index for:

[0246] Furthermore, the controller calculates the salt swelling-solution trapping coupling conversion index. The target thickness is still taken as... Then the overlap between the two forwards is:

[0247] The sensitivity to catastrophe state transitions, after normalization, is taken as follows:

[0248] according to and take The salt swelling-solution trap coupling transformation index can be obtained. for:

[0249] The identification rule used in this embodiment is: when At that time, the control unit is determined to be in a sinkhole-dominated state. Because... satisfy Therefore, the controller determines it to be in a state dominated by dissolution.

[0250] To address the aforementioned situation, this embodiment sets a moisture content safety window as follows: ; Maximum allowable rate of bulging The maximum allowable settlement rate is Non-targeted disaster risk-induced incremental threshold .

[0251] Low-pressure grouting uses a micro-expansion cement-fly ash composite grout, with the mass ratio of ordinary Portland cement, fly ash, sodium silicate, and bentonite in the cementitious material being [missing information]. The water-to-gel ratio is 0.85, and the initial viscosity is... The initial setting time is 38 minutes. Nine grouting pipes are installed in the preset grouting holes, arranged along the perimeter and central axis of the target control unit. The hole spacing is 1.5m, the hole depth is 1.40m, and the effective grouting section is 0.80–1.30m. The controller uses the current control time... As the starting point for rolling optimization, the prediction time domain length is... Internal treatment linkage control vector sequence Perform rolling optimization to solve the problem and output the first treatment linkage control vector at the current control moment. :

[0252] In this embodiment, the prediction step size is 2 hours, the number of prediction steps H=6, and the corresponding prediction time domain is 12 hours. The controller calculates the prediction time of each candidate treatment linkage control scheme within 6 prediction steps according to the risk state prediction model. , , , and And calculate the corresponding objective function based on the average objective function. Values. The candidate solutions and their solutions are shown in Table 9.

[0253] Table 9. Candidate treatment linkage control schemes and solution results for the first control step at the current control time.

[0254] Table 9 shows the traffic load restriction levels. This indicates a speed limit of 40 km / h and allows a 20% reduction in axle load; this is the traffic load restriction level. This indicates a speed limit of 30 km / h and allows for a 30% reduction in axle load. Option A employs high-flow-rate, rapid drainage, but the predicted incremental risk of non-target catastrophic events is:

[0255] There is a possibility of salt reaccumulation induced by rapid water loss in the shallow layer, so it was discarded. Although Scheme C can further suppress settlement, the grouting intensity is too high, the control variable changes significantly, and the objective function is higher than that of Scheme B. Scheme B satisfies both settlement suppression and anti-induction constraints, therefore the controller determines it as the optimal control scheme, i.e.:

[0256] The controller, based on this, controls the drainage branch, electrically controlled valves, salt collection and drainage wells, grouting pipes, and traffic guidance units, continuously for 48 hours. Throughout the control process, the opening degree of the electrically controlled valves in the drainage branch is maintained at approximately 45%, and the drainage flow rate is controlled at 1.4±0.1L / min, ensuring that the water level drop rate in the side ditch remains below 0.12m / h to avoid additional structural disturbances caused by excessively rapid unloading. Grouting employs a three-stage pulse injection method, with each stage lasting 8 minutes and an interval of 20 minutes, resulting in a total grouting volume of 9.6L per hole.

[0257] The controller completes a rolling control cycle every 2 hours. After each rolling control cycle, the state vector is reacquired, and incremental update values ​​are generated based on the non-target catastrophic risk induced after execution. And the measured changes in settlement rate affect the first treatment linkage control vector currently being implemented. Fine-tuning was performed. Table 10 lists the updated heave rate, sedimentation rate, salt heave risk index, dissolution risk index, coupling transformation index, and state determination results at each control time during the 48-hour treatment period, rather than the predicted times during the rolling optimization process. The predicted value.

[0258] Table 10 Control Unit Dynamic feedback results

[0259] As shown in Table 10, in the first 12 hours of the initial control phase, the salt swelling risk index rose slightly from 0.241 to 0.248, and the incremental risk of non-targeted disasters was 0.007, which was less than the preset threshold. This indicates that although shallow water and salt redistribution occurred during the combined process of graded drainage and low-pressure grouting, it did not lead to a significant amplification of the risk of salt swelling. As the treatment continued, the controller gradually reduced the settlement rate to [a lower threshold]. and increase the risk index of solution collapse It dropped to 0.297, returning to below the steady-state threshold.

[0260] After 48 hours of treatment, the controller re-identified the dual-leader system, and the results showed that the salt crystallization leader... Stable at And the dissolution and wetting front It moved down from the initial 0.95m to The corresponding overlap between the two forwards becomes:

[0261] Meanwhile, the migration rate of the dissolution and wetting front Reduced to This indicates that the spread of wet dissolution in the middle and lower parts has been significantly slowed down.

[0262] At 48 hours, the controller re-extracted the dissolution wetting front. The updated solution risk index is calculated based on nearby parameters. The normalized value of the solution wetting front velocity is now:

[0263] The conductivity at that location at the previous moment was The current conductivity is The salt cementation loss rate is:

[0264] After normalization, it becomes:

[0265] The change in pore water pressure in the last 12 hours has decreased to ,Right now The normalized value of the settling velocity is:

[0266] At this point, the dynamic stiffness recovers to 159 MPa, and the attenuation relative to the initial reference of 192 MPa is:

[0267] Therefore:

[0268] Substituting into the solution risk index for:

[0269] The controller simultaneously recalculates the salt swelling risk index. After treatment, at the salt crystallization front... The pore fluid salt concentration measured nearby was:

[0270] The corresponding temperature is:

[0271] And still take:

[0272] The salt crystallization pressure potential is:

[0273] Therefore:

[0274] After treatment, the conductivity gradient and temperature gradient at the 0.20–0.40 m interface decreased to [values ​​to be inserted here]. The normalized value of the bulging rate is Therefore, the salt inflation risk index for:

[0275] Further substitution of the updated , , and normalized disaster state transition sensitivity The salt swelling-solution trap coupling transformation index was obtained. for:

[0276] Finally, the controller calculated the risk reduction rates before and after treatment as follows:

[0277]

[0278]

[0279] In summary, the salt inflation risk index Decreased by 29.0%, the risk index of landslide collapse. A decrease of 59.1%, salt swelling-dissolution coupling conversion index The decrease was 41.1%. Throughout the entire control process, the incremental risk of non-target disasters was consistently below the preset threshold, indicating that the graded drainage-low-pressure grouting-weak pumping-traffic load restriction linkage control strategy adopted in this embodiment can effectively suppress the dominant risk of solution subsidence while avoiding the amplification of new shallow salt expansion caused by excessive drainage or excessive grouting disturbance.

[0280] Therefore, under the continuous infiltration condition set in this embodiment, the method of the present invention can accurately calculate the subsidence risk index and coupling transformation index based on the results of multi-source state perception and dissolution wetting front identification, and solve the appropriate treatment linkage control vector under the anti-induced constraint, so as to realize the dynamic control and closed-loop adjustment of the dominant risk of subsidence of saline soil subgrade.

[0281] Example 3: Linkage control in a combined salt swelling-dissolution scenario.

[0282] This embodiment still selects the same saline soil test subgrade as in Embodiments 1 and 2. The subgrade section is 60m long, 8.5m wide at the top, and 2.4m high. The fill material is sulfate-chloride type saline silty clay with an initial total salt content of 2.31% and a compaction degree of 95%. To form a salt swelling-dissolution composite transformation condition, three types of disturbances—evaporation concentration, intermittent infiltration, and heavy traffic—were applied sequentially to the subgrade section within 42 hours before the start of the test. The first 18 hours were the strong evaporation stage, and the surface evaporation was converted to... The daytime high temperature is The lowest temperature at night was The following 10 hours were a period of intermittent rainfall and ditch replenishment, with a cumulative infiltration of 18.4 mm and a ditch water level rise of 0.16 m. The final 14 hours were a period of heavy traffic disturbance, with an average axle load of 72 kN, equivalent to a traffic flow of 386 vehicles / hour. After the above-mentioned combined disturbance, the shallow soil still maintained a strong tendency for salt precipitation and uplift, while the middle to deep soil layers showed signs of wetting and downward migration, weakening of salt cementation, and localized compression deformation, forming a complex catastrophic state in which the salt crystallization front and the dissolution and wetting front approached each other.

[0283] The six control units arranged longitudinally along the roadbed section are still referred to as After the controller updates the status data of each control unit over the past 24 hours, it calculates the salt swelling risk index for each unit. Dissolution risk index and the salt swelling-dissolution coupling transformation index The results are shown in Table 11.

[0284] Table 11 Initial Risk Calculation Results for Control Units

[0285] As shown in Table 11, the control unit Salt inflation risk index With the risk index of dissolution All are at a high level, and the salt swelling-dissolution coupling transformation index achieve This is significantly higher than the composite disaster judgment threshold of 0.55 set in this embodiment, therefore the controller will... Selected as a key control target.

[0286] Controller Read The original state data at the end of the current scrolling window is shown in Table 12.

[0287] Table 12 Control Unit initial state data

[0288] at the same time, The average bulge rate of the road surface over the past 6 hours is The average settlement rate in the depth range of 0.80–1.20 m is It is evident that differential deformation characteristics, characterized by both shallow bulging and mid-layer compression, have emerged in this control unit. The dynamic stiffness decreased from 183 MPa to 142 MPa in the last 12 hours; therefore, the dynamic stiffness attenuation is:

[0289] In this embodiment, the controller still adopts the same dual-forward fusion recognition model as in the previous embodiments, that is:

[0290]

[0291] Based on the data of each layer shown in Table 12 and the continuous monitoring sequence of the previous 12 hours, the double-forward identification results obtained by the controller are shown in Table 13.

[0292] Table 13 Control Unit Two-forward recognition results

[0293] As shown in Table 13, the highest salt crystallization front identification score was achieved in the salt crystallization front identification results. The value is 0.782, corresponding to the candidate forward position. The score was 0.30m; the next highest score was for salt crystallization front recognition. The score is 0.648, and the difference between the two is 0.134. In the results of dissolution and wetting front identification, the highest dissolution and wetting front identification score is... The value is 0.815, corresponding to the candidate forward position. The score was 0.60m; the next highest score was for the identification of the dissolution and wetting front. The score is 0.601, and the difference between the two is 0.214. Therefore, both types of forward identification scores have a unique maximum value, based on which the salt crystallization forward is determined. Dissolution and wetting front .because The distance is less than the 0.40m threshold for determining the double-forward coupling set in this embodiment, therefore the controller will... This is labeled as a dual-forward coupling unit. Based on the 13 sets of effective forward positions formed every 2 hours within the first 24 hours, the migration speed of the salt crystallization forward and the migration speed of the dissolution and wetting forward are calculated according to the linear fitting method in Example 1, thus obtaining the salt crystallization forward migration speed. Migration speed of the dissolution and wetting front The different advancing directions of the two types of forwards and the continuously decreasing distance between them indicate that the control unit is in a clearly sensitive stage of complex transformation.

[0294] The controller first calculates the salt swelling risk index. At the forefront of salt crystallization The pore fluid salt concentration was measured from a nearby pore fluid sample. Corresponding temperature The salt saturation concentration is taken at this temperature. Molar volume of salt crystals The salt crystallization pressure potential is:

[0295] This embodiment still uses the formula. Calculate the salt inflation risk index, where

[0296] Based on the data in Table 12 for the interface between 0.20 and 0.40 m layers, then:

[0297]

[0298]

[0299]

[0300] This embodiment does not consider freeze-thaw cycles, therefore the freeze-thaw cycle strength is taken as the standard. Substituting into the equation yields the salt inflation risk index. for:

[0301] Subsequently, the controller calculates the collapse risk index. Its expression is:

[0302] in:

[0303] The normalized value of the dissolution wetting front velocity is:

[0304] In the dissolution and wetting front Nearby, conductivity at the previous moment Current conductivity ,Pick , The salt cementation loss rate is:

[0305] After normalization, we have:

[0306] The last 12 hours have been in the dissolving and wetting front. Changes in pore water pressure in the vicinity Therefore, after normalization The normalized value of the settling velocity is:

[0307] The normalized value of dynamic stiffness attenuation is:

[0308] Substituting into the solution risk index for:

[0309] Furthermore, the controller calculates the salt swelling-solution trapping coupling conversion index. The target thickness is still taken as... Then the overlap between the two forwards is:

[0310] In this embodiment, the sensitivity to catastrophic state transition is normalized based on the evolution of the leading edge position and the rate of change of the risk index over the previous 12 hours. According to the formula and take The salt swelling-solution trap coupling transformation index can be obtained. for:

[0311] In this embodiment, when At that time, the controller determines the target control unit to be in a complex disaster state. Because... The above conditions are met, therefore it is determined to be a salt swelling-dissolution collapse complex disaster state.

[0312] To address the aforementioned situation, this embodiment sets the moisture content safety window as follows: The maximum allowable bulging rate is set to The maximum allowable settling rate is set to The threshold for non-targeted catastrophic risk-induced incremental changes is set to... .

[0313] Simultaneously, to reconstruct the softened and weak layer within a depth range of 0.65–1.10 m, low-pressure zoned grouting was performed using an ultrafine cement-fly ash-sodium silicate composite grout, wherein the mass ratio of ultrafine cement, fly ash, sodium silicate, and bentonite was [missing information]. The water-cement ratio is 0.90, and the slurry density is... The initial viscosity is The initial setting time is 42 minutes. Grouting pipes are installed in the pre-set grouting holes; a total of 11 grouting pipes are installed, with a hole spacing of 1.3m and a hole depth of 1.30m. The total grouting volume per hole is controlled as follows: .

[0314] The controller uses the current control time As the starting point for rolling optimization, the prediction time domain length is... Internal treatment linkage control vector sequence Perform rolling optimization to solve the problem and output the first treatment linkage control vector at the current control moment. :

[0315] In this embodiment, the prediction step size is 2 hours, the number of prediction steps H=6, and the corresponding prediction time domain is 12 hours. The controller calculates the prediction time of each candidate treatment linkage control scheme within 6 prediction steps according to the risk state prediction model. , , , and And calculate the corresponding objective function based on the average objective function. Values. The candidate solutions and their solutions are shown in Table 14.

[0316] Table 14 Candidate treatment linkage control schemes and solution results

[0317] Table 14 shows the traffic load restriction levels. This indicates a speed limit of 40 km / h and allows a 20% reduction in axle load; this is the traffic load restriction level. This indicates a speed limit of 30 km / h and an allowable axle load reduction of 30%. Table 14 shows that Scheme A has an excessively high tiered drainage intensity and an insufficient grouting volume. While it can reduce the wetting level of the middle layer in a short time, it predicts an increase in the risk of non-target catastrophic events. Option C carries the risk of inducing shallow recrystallization and localized additional bulging, and is therefore discarded. Option C has excessively high electroosmotic and pumping intensities, resulting in large fluctuations in the control variables and a higher objective function value than Option B. Option B simultaneously satisfies both the anti-induced constraint and the minimum objective function requirement, therefore the controller determines it as the optimal control option, i.e.:

[0318] The controller, in conjunction with this, controls the drainage branch, electrically controlled valves, electroosmotic electrodes, salt collection and pumping wells, grouting pipes, shallow heating zones, and traffic guidance units, continuously operating for 60 hours. During the first 12 hours, the controller prioritizes using the drainage branch, electrically controlled valve 14, and shallow heating zones to regulate the water content within a depth range of 0.40–0.80 m. The shallow heating zone is preferably a low-power electric heating tape to reduce the dual risks of continued upward movement of the dissolution and wetting front and rapid recrystallization of shallow salts. During the subsequent 24 hours, the controller maintains the electroosmotic electric field strength. The electroosmotic polarity was switched every 6 hours to migrate the enriched salt in the middle layer to the salt collection wells on the shoulder. Simultaneously, low-pressure zonal grouting was performed on the softened zone from 0.65 to 1.10 meters. Over the next 24 hours, the controller gradually reduced the grouting frequency while maintaining weak pumping and load control to achieve convergence of risk redistribution within the roadbed. Throughout the entire control process, the maximum measured increment of non-target catastrophic risk induced by each control cycle was 0.024, consistently below the preset threshold of 0.04, and no treatment conflict state was triggered. Table 15 lists the updated heave rate, settlement rate, salt heave risk index, dissolution risk index, coupling transformation index, and state determination results at each control time during the 60-hour treatment period, rather than the predicted times during the non-rolling optimization process. The predicted value.

[0319] Table 15 Control Unit Dynamic feedback results

[0320] As shown in Table 15, after 24 hours of implementing the linkage control, Salt swelling-dissolution coupling transformation index The value has decreased from 0.683 to 0.523, indicating a significant mitigation of the complex catastrophic state; at the end of 60 hours, the uplift rate and settlement rate have decreased to [missing values]. and All of them have met the control constraints set in this embodiment.

[0321] After 60 hours of treatment, the controller performed dual-leader identification again, and the results showed that the salt crystallization leader... Back to Dissolution and wetting front Move down to At this point, the overlap between the two forwards becomes:

[0322] Compared with 0.800 before treatment, the spatial coupling degree of the double front is significantly reduced, indicating that the shallow salt precipitation interface and the middle dissolution and softening interface have been effectively separated.

[0323] After treatment, at the forefront of salt crystallization The pore fluid concentration was measured again from a nearby sample and found to have decreased to [value missing]. Corresponding temperature The salt crystallization pressure potential is recalculated using the same formula:

[0324] but:

[0325] After treatment, the conductivity gradient and temperature gradient at the 0.20–0.40 m interlayer interface decreased to:

[0326] The normalized value of the bulging rate is:

[0327] Therefore, the updated salt inflation risk index for:

[0328] In the dissolution and wetting front Nearby, the migration speed of the dissolved and moistened front after treatment Descending to Therefore, the velocity of the dissolution and wetting front is:

[0329] The conductivity at this location at the previous moment was:

[0330] The current conductivity is:

[0331] The salt cementation loss rate is:

[0332] After normalization, it becomes:

[0333] The change in pore water pressure in the last 12 hours has decreased to Normalized value The normalized value of the settling velocity is:

[0334] At this point, the dynamic stiffness recovers to 173 MPa, and the attenuation rate relative to the initial reference of 183 MPa is:

[0335] Therefore, after normalization, it becomes:

[0336] Substituting into the solution risk index for:

[0337] Furthermore, based on the updated , , and the sensitivity to catastrophe state transition after normalization. The salt swelling-solution trap coupling transformation index can be obtained. for:

[0338] Finally, the controller calculated the risk reduction rates before and after treatment as follows:

[0339]

[0340]

[0341] In summary, the salt inflation risk index The risk index of landslides decreased by 53.6%. A decrease of 61.7%, salt swelling-dissolution coupling conversion index The water content decreased by 58.3%. Meanwhile, during the control period, the water content in the depth range of 0.40 to 0.80 m remained between 12.2% and 16.3%, meeting the preset water content safety window requirements. This indicates that the linkage control strategy adopted in this embodiment effectively avoided the conflict between "excessive drainage to suppress dissolution leading to aggravated shallow salt precipitation" and "excessive salt migration and drainage to weaken salt swelling leading to further softening of the middle layer structure".

[0342] Therefore, under the salt swelling-dissolution complex transformation condition set in this embodiment, the method of the present invention can accurately determine the complex disaster state based on the double-forward identification results and the multi-risk index calculation results, and generate a treatment linkage control vector that takes into account drainage, electroosmotic salt migration, pumping, zoned low-pressure grouting, shallow heating adjustment and traffic load restriction under the anti-induced constraint, so as to realize the dynamic control and closed-loop adjustment of the complex disaster risk of saline soil subgrade.

[0343] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0344] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0345] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for electroosmotic linkage control of salt swelling-dissolution disaster in saline soil roadbed, characterized in that, The method includes the following steps: Multiple control units were divided along the longitudinal, transverse and depth directions of the saline soil subgrade, and multi-source state data of each control unit were collected in the subgrade surface or ditch area to construct a water-thermal-salt-force state vector to characterize the state gradient information of each control unit along the depth direction. Based on the water-heat-salt-force state vector, the dual-front parameters, including the salt crystallization front, the dissolution and wetting front, and their migration velocities, are identified in each control unit. Based on the water-heat-salt-force state vector and the double-front parameters, the salt swelling risk index, solution sinking risk index and salt swelling-solution sinking coupling transformation index are calculated respectively to identify and classify the disaster state of each control unit. Based on the disaster state, a treatment linkage control vector is generated, and a rolling optimization solution is performed under constraints. The treatment action is implemented based on the treatment linkage control vector obtained by the solution, and the risk reduction rate and the non-target disaster risk-induced increment are calculated based on the updated water-heat-salt-force state vector after treatment. This enables adaptive correction of the disaster state classification threshold and the solution parameters of the treatment linkage control vector.

2. The electroosmotic linkage control method according to claim 1, characterized in that, The multi-source condition data includes infiltration data collected from the roadbed surface or ditch area. Drainage flow and environmental boundary condition data; No. Each control unit at time Water-heat-salt-force state vector The expression is: ； in, Moisture content; For temperature; Electrical conductivity; For matrix suction; Pore ​​water pressure; This refers to the amount of swelling. This refers to the settlement amount; For load response; This refers to the groundwater level.

3. The electroosmotic linkage control method according to claim 1, characterized in that, The control unit includes at least the following along the depth direction: The shallow salt swelling sensitive area is used to obtain state data related to salt crystallization swelling; the middle salt migration transition area is used to obtain state data related to water-salt migration; and the deep dissolution sensitive area is used to obtain state data related to salt cementation, dissolution, softening, and structural settlement.

4. The electroosmotic linkage control method according to claim 1, characterized in that, The salt crystallization forward Used to indicate the main interface locations where salts in pore fluid transition from a dissolved state to a crystalline state; The dissolution and wetting front Used to indicate the location of the main interfaces where infiltration, wetting, salt cementation, dissolution, and softening propagate into the subgrade; The control unit at time Salt crystallization vanguard and dissolution wetting front They are represented as follows: ； ； in, and Forward inversion function based on weighted fusion, state estimation, or trained model; The conductivity gradient; For temperature gradient; The matrix suction gradient; The rate of expansion; For moisture content gradient; The rate of decrease in conductivity; The rate of change of pore water pressure; Settlement rate; when Less than the preset distance threshold At that time, the corresponding control unit is marked as a double-forward coupling unit to characterize the highly sensitive region where salt swelling and solution trapping may transform into each other or be coupled and amplified.

5. The electroosmotic linkage control method according to claim 1, characterized in that, The salt swelling risk index The calculation formula is: ； in, The crystallization pressure potential of normalized salts; This represents the normalized conductivity gradient. This represents the normalized temperature gradient. Normalized bulging rate; Normalized freeze-thaw cycle strength; These are the weighting coefficients; The collapse risk index The calculation formula is: ； in, Normalized dissolution wetting front velocity; Normalized salt cementation loss rate; This represents the normalized change in pore water pressure. Normalized settling rate; This represents the normalized dynamic stiffness attenuation. These are the weighting coefficients; The salt swelling-dissolution coupling conversion index The expression used to characterize the spatial coupling degree and dynamic transformation trend of salt swelling and solution collapse is: ； in, Forefront of salt crystallization With the dissolution and wetting front The degree of overlap; Sensitivity to catastrophic state transitions; These are the weighting coefficients.

6. The electroosmotic linkage control method according to claim 1, characterized in that, The hysteresis classification method is used to classify the disaster status of each control unit into one or more of the following: stable state, salt expansion-dominated state, solution collapse-dominated state, complex disaster state, and conflict resolution state. When the salt swelling risk index Exceeding the salt swelling dominant threshold and the collapse risk index When the value is below the threshold for dominant dissolution, it is determined to be in a state dominated by salt swelling. When the collapse risk index Exceeding the dominant threshold for dissolution and the salt swelling risk index When the temperature is below the salt swelling threshold, it is determined to be in a collapse-dominated state. When the salt swelling risk index and the risk index of dissolution Simultaneously, a higher or the aforementioned salt swelling-dissolution coupling conversion index When the coupling threshold is exceeded, it is determined to be a compound disaster state; When a certain action causes a rapid increase in the incremental risk of non-target disasters, it is determined to be a state of action conflict.

7. The electroosmotic linkage control method according to claim 1, characterized in that, The treatment linkage control vector The expression is: ； in, This refers to the drainage flow rate; This refers to the opening degree of the drain valve; The electric field strength is the electroosmotic electric field strength. This is the electroosmotic polarity switching cycle; To pump out the flow rate; This refers to the grouting flow rate; This refers to the grouting pressure. For heat preservation or heating power; Traffic load limit level; When the disaster state is dominated by salt expansion, the treatment linkage control vector Prioritize the use of drainage, electroosmotic salt migration, pumping, and heat preservation or heating to control the amount of water and salt. When the disaster state is a solution-dominant state, the treatment linkage control vector Prioritize the use of graded drainage, low-pressure grouting, and traffic load control to suppress structural collapse and the expansion of the softening layer; When the disaster state is a complex disaster state, the treatment linkage control vector It is generated in the following order: maintaining the moisture content of the control unit within a preset safe range, electroosmotic salt migration, pumping out the collected salt solution, and zonal low-pressure grouting; When the disaster state is a conflict state, the controller stops any actions that cause non-target disaster risk to increase beyond a preset threshold, and re-solves the treatment linkage control vector. .

8. The electroosmotic linkage control method according to claim 1, characterized in that, The rolling optimization solution under the constraints includes: A rolling time-domain optimization method is used to solve the treatment linkage control vector under constraints of moisture content safety window, bulging rate, settling rate, actuator output limit, and anti-induced constraints. The constraints include at least the following: ； ； ； ； ； in, This is the current control moment; To predict the first in the time domain Prediction step number; To predict the corresponding prediction time within the time domain; and This is the boundary of the moisture content safety window; The upper limit of the permissible bulging rate; The upper limit of the allowable settlement rate; and These are the lower and upper limits of the allowable output for each component of the treatment linkage control vector, respectively. For the predicted time The predicted moisture content of the control unit; For the predicted time The predicted value of the bulging rate of the control unit; For the predicted time The predicted settling rate of the control unit; For the predicted time Predicted values ​​of the treatment linkage control vector; Predicted time The predicted incremental value of non-targeted disaster risk; The anti-induction threshold; The objective function for the rolling optimization is: ； in, To predict the length of the time domain; For the predicted time The predicted value of the salt inflation risk index; For the predicted time The predicted value of the solution risk index; For the predicted time Predicted value of salt swelling-solution trap coupling transformation index; For the predicted time The change in the treatment linkage control vector relative to the previous control cycle; For the predicted time The costs of construction disturbance or traffic disruption; These are the weighting coefficients.

9. The electroosmotic linkage control method according to claim 1, characterized in that, The formula for calculating the risk reduction rate is as follows: ； ； ； in, , , These are the risk reduction rates of the salt swelling risk index, the solution collapse risk index, and the salt swelling-solution collapse coupling conversion index, respectively. , , These are the risk indices corresponding to the treatment prior to the procedure. , , These are the corresponding risk indices after treatment; When the risk reduction rate is lower than the preset lower limit or the incremental risk of non-target disaster is higher than the preset upper limit, the control constraint level is increased or the treatment mode is switched back to form a closed-loop control process of "state perception - front-end identification - risk classification - linkage control - feedback correction".

10. A system for implementing the electroosmotic linkage control method according to any one of claims 1-9, characterized in that, It includes the on-site roadbed object layer, the control and calculation layer, and the execution and feedback layer; The on-site roadbed object layer includes a shallow salt swelling sensitive zone, a middle salt migration transition zone, and a deep solution sinking sensitive zone set in the roadbed monitoring and control object. Moisture content sensors, temperature sensors, conductivity sensors, and vertical displacement monitoring units are deployed in the shallow salt swelling sensitive zone. Moisture content sensors, conductivity sensors, matrix suction sensors, and pore water pressure sensors are deployed in the middle salt migration transition zone. Pore water pressure sensors, settlement monitoring units, load response monitoring units, and groundwater level monitoring units are deployed in the deep solution sinking sensitive zone. The execution and feedback layer includes a drainage branch, an electrically controlled valve, an electroosmotic electrode, and a salt collection and pumping well as a water and salt regulation unit; and a grouting pipe as a structural treatment unit. In addition, a shallow heating zone and a traffic guidance unit serve as auxiliary adjustment units; The control and computing layer includes a data transmission unit that establishes a communication connection with the on-site roadbed object layer, as well as a central control platform and a front identification and status analysis module, a risk calculation and graded early warning module, and a linkage control and feedback correction module that constitute the edge controller.

Images