A foundation pit drainage optimization method and system based on water level monitoring

Abstract

This invention relates to the field of foundation pit engineering construction technology, specifically to a foundation pit drainage optimization method and system based on water level monitoring. It includes: dividing dewatering zones according to geological conditions and construction plans, and arranging dewatering wells and observation wells; determining differentiated target drawdown depths based on excavation depth and support status; establishing a zoned hydraulic coupling model through pulse pumping tests and determining the coupling kernel function parameters; predicting future water level changes based on the coupling kernel function and historical outflow through convolution operations; calculating the optimal outflow sequence using a rolling time-domain optimization method, with the minimum cumulative outflow as the objective and satisfying the target drawdown depth as the constraint; controlling the pump operating power and implementing coordinated compensation; and periodically updating the coupling kernel function parameters to adapt to system changes. This invention improves the accuracy of dewatering control, reduces energy consumption, and ensures construction safety.

Description

Technical Field

[0001] This invention relates to the field of foundation pit engineering construction technology, specifically to a foundation pit drainage optimization method and system based on water level monitoring. Background Technology

[0002] Dewatering is a crucial aspect of deep foundation pit engineering, directly impacting construction safety and the stability of the surrounding environment. With increasing urban underground space development and the growing complexity of foundation pit projects, higher demands are being placed on the refined and intelligent control of dewatering systems.

[0003] Existing foundation pit dewatering technologies have the following shortcomings: First, dewatering control methods are crude, often employing independent control strategies for each zone and setting fixed pumping volumes based on experience. This makes it difficult to adapt to dynamic changes in excavation depth and support conditions during construction, leading to large water level fluctuations, posing both safety hazards and causing excessive dewatering. Second, there is a lack of quantitative description of the hydraulic coupling relationships between zones. Groundwater systems are interconnected through aquifers; pumping in one zone affects the water levels of adjacent zones, but existing methods cannot predict or compensate for this mutual influence, resulting in control conflicts. Third, optimization decision-making capabilities are insufficient. Traditional methods determine pumping schemes based on static hydrogeological parameters, failing to fully utilize real-time monitoring data and unable to dynamically adjust based on water level feedback, resulting in significant energy waste. Fourth, model adaptability is poor. Groundwater system characteristics change with long-term pumping and surrounding construction activities, causing the prediction accuracy of fixed-parameter models to gradually decrease, making it difficult to guarantee control effectiveness throughout the entire construction cycle. Summary of the Invention

[0004] This invention provides a method and system for optimizing foundation pit drainage based on water level monitoring, which solves the problems of insufficient accuracy in precipitation control and excessive energy consumption in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] This invention provides a method for optimizing foundation pit drainage based on water level monitoring, comprising:

[0007] S100: The foundation pit area is divided into multiple dewatering zones. Dewatering wells are set around the perimeter of each dewatering zone, and observation wells are set inside each dewatering zone.

[0008] S200: Obtain the current excavation depth and support completion status of the precipitation zone, and determine the target drawdown depth of the precipitation zone based on the current excavation depth and support completion status;

[0009] S300: A hydraulic coupling model with intervals is established through pulse pumping tests to determine the coupling kernel function of the precipitation interval. The coupling kernel function describes the time-delay influence of historical pumping behavior on the current water level.

[0010] S400: Obtain the historical water output sequence of precipitation wells and the real-time water level drawdown of observation wells, and predict the water level drawdown changes of precipitation zones in the future time domain based on the coupled kernel function and the historical water output sequence;

[0011] S500: Using the rolling time-domain optimization method, the target water output sequence of the precipitation wells is calculated with the constraint that the precipitation zone meets the target water level drawdown in the future time domain and the objective that the cumulative water output of the precipitation wells is minimized in the future time domain.

[0012] S600: Controls the operating power of the pumps in the corresponding dewatering wells according to the target water output sequence of the dewatering wells.

[0013] As a preferred embodiment of the present invention, dividing the foundation pit area into multiple dewatering zones includes:

[0014] Based on the geological survey data of the foundation pit area, obtain the soil layer types and permeability coefficients at different locations;

[0015] Obtain the planned excavation depth and excavation sequence for the area based on the construction plan for the foundation pit;

[0016] Areas with the same soil type, permeability coefficient differences less than a preset threshold, and planned excavation depths are divided into the same precipitation zone.

[0017] As a preferred embodiment of the present invention, determining the target drawdown depth of the dewatering zone based on the current excavation depth and the completion status of the support includes:

[0018] For dewatering zones where support is incomplete, the target water level drawdown is set as the sum of the current excavation depth and the first safety margin;

[0019] For dewatering zones where support has been completed, the target water level drawdown is set as the sum of the current excavation depth and the second safety margin;

[0020] The first safety margin is greater than the second safety margin.

[0021] As a preferred embodiment of the present invention, the step of establishing a segmented hydraulic coupling model through pulse pumping tests and determining the coupling kernel function for each precipitation interval includes:

[0022] Start the rainwater wells corresponding to the first precipitation zone to pump water for a preset time at a set flow rate, while keep the rainwater wells corresponding to other precipitation zones stopped;

[0023] Monitor the water level changes of the observation wells in the second precipitation zone during and after pumping, and obtain the response curve of water level drawdown over time;

[0024] The response curve is fitted with an exponential decay function to determine the coupling kernel function parameters of the first precipitation zone to the second precipitation zone. The coupling kernel function parameters include the coupling strength coefficient and the time-delay decay coefficient.

[0025] As a preferred embodiment of the present invention, the prediction of water level drawdown changes in future time domains for precipitation zones based on the coupled kernel function and historical outflow sequence includes:

[0026] Obtain the water output sequence of all precipitation wells within the historical time window;

[0027] For any precipitation zone, perform temporal convolution operations on the historical water output sequences of all precipitation wells with the corresponding coupled kernel functions;

[0028] The predicted water level drawdown for the current time zone is obtained by summing up all the results of the temporal convolution operations.

[0029] Assuming the precipitation wells continue operating at their current output, the predicted drawdown depth of the precipitation zone at various future times is calculated.

[0030] As a preferred embodiment of the present invention, the target water output sequence for calculating precipitation wells includes:

[0031] The prediction time domain length and the control time domain length are set, wherein the prediction time domain length is greater than the control time domain length;

[0032] An optimization problem is established, with the constraint that the predicted drawdown of the precipitation zone in the prediction time domain meets the target drawdown, and the optimization objective is to minimize the cumulative water output of the precipitation wells in the control time domain.

[0033] Solving the optimization problem yields the target water output sequence of the precipitation wells within the control time domain;

[0034] The target water output is executed in the first control cycle of the target water output sequence, and the optimization problem is solved again in the next control cycle.

[0035] As a preferred embodiment of the present invention, it further includes:

[0036] When the pumping volume of the first precipitation zone increases, the water level drop increment of the adjacent precipitation zones in the future time domain is calculated based on the coupling kernel function of the first precipitation zone to the adjacent precipitation zones.

[0037] During the time lag period of the incremental drawdown, the pumping rate of the corresponding wells in the adjacent precipitation zones is reduced synchronously to compensate for the incremental drawdown experienced by the adjacent precipitation zones.

[0038] As a preferred embodiment of the present invention, it further includes:

[0039] The deviation between the predicted drawdown in precipitation zones and the measured drawdown measured by observation wells is calculated periodically.

[0040] When the absolute value of the deviation exceeds a preset threshold or the deviation accumulates unidirectionally over multiple consecutive periods, it is determined that the parameters of the coupling kernel function have changed.

[0041] The parameters of the coupling kernel function were re-identified using the historical outflow sequence and measured drawdown data within a sliding time window, and the least squares method was employed.

[0042] The updated coupled kernel function parameters are used to re-predict the drawdown of precipitation zones and recalculate the target water output sequence of precipitation wells.

[0043] This invention also proposes a foundation pit drainage optimization system based on water level monitoring, comprising:

[0044] The zoned well module is used to divide the foundation pit area into multiple dewatering zones. Dewatering wells are set around the perimeter of each dewatering zone, and observation wells are set inside each dewatering zone.

[0045] The target determination module is used to obtain the current excavation depth and support completion status of the precipitation zone, and determine the target water level drawdown of the precipitation zone based on the current excavation depth and support completion status.

[0046] The model building module is used to establish a segmented hydraulic coupling model through pulse pumping tests, and to determine the coupling kernel function of the precipitation interval. The coupling kernel function describes the time-delay influence of historical pumping behavior on the current water level.

[0047] The water level prediction module is used to obtain the historical water output sequence of precipitation wells and the real-time water level drawdown of observation wells, and predict the water level drawdown changes of precipitation zones in the future time domain based on the coupled kernel function and the historical water output sequence.

[0048] The optimization calculation module is used to calculate the target water output sequence of the precipitation wells by adopting the rolling time domain optimization method, with the constraint that the precipitation zone meets the target water level drawdown in the future time domain and the objective of minimizing the cumulative water output of the precipitation wells in the future time domain.

[0049] The execution control module is used to control the operating power of the water pumps in the corresponding dewatering wells according to the target water output sequence of the dewatering wells.

[0050] The beneficial effects of this invention are:

[0051] 1. This invention establishes a segmented hydraulic coupling model through pulse pumping tests, uses coupling kernel functions to quantitatively describe the time-delayed impact of historical pumping behavior on the current water level, and achieves accurate prediction of water levels in multiple zones based on convolution operations. This method is the first to introduce time-delay system theory into foundation pit dewatering control, overcoming the shortcomings of existing technologies in quantifying segmented hydraulic coupling, and significantly improving prediction accuracy and control reliability.

[0052] 2. This invention employs a rolling time-domain optimization method, combining a coupled prediction model and differentiated objective constraints to calculate the optimal pumping scheme in real time, and eliminates inter-regional interference through a collaborative compensation mechanism. This closed-loop optimization strategy significantly reduces cumulative water output and system energy consumption while ensuring safe precipitation, thus minimizing adverse impacts on the surrounding environment.

[0053] 3. This invention establishes an adaptive update mechanism for model parameters, dynamically adjusting the parameters of the coupling kernel function through deviation monitoring and online identification techniques. This mechanism endows the system with self-learning capabilities, maintaining the long-term effectiveness of the prediction model throughout the construction cycle and avoiding performance degradation caused by time-varying systems in fixed-parameter models. Attached Figure Description

[0054] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0055] Figure 1 This is a flowchart illustrating an optimized method for foundation pit drainage based on water level monitoring, according to the present invention.

[0056] Figure 2 This is a schematic diagram of the structure of a foundation pit drainage optimization system based on water level monitoring according to the present invention. Detailed Implementation

[0057] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0058] Example 1: As Figure 1 As shown, the present invention provides a method for optimizing foundation pit drainage based on water level monitoring, comprising:

[0059] S100: The foundation pit area is divided into multiple dewatering zones. Dewatering wells are set around the perimeter of each dewatering zone, and observation wells are set inside each dewatering zone.

[0060] Furthermore, the division of the foundation pit area into multiple dewatering zones includes:

[0061] Based on the geological survey data of the foundation pit area, obtain the soil layer types and permeability coefficients at different locations;

[0062] Obtain the planned excavation depth and excavation sequence for the area based on the construction plan for the foundation pit;

[0063] Areas with the same soil type, permeability coefficient differences less than a preset threshold, and planned excavation depths are divided into the same precipitation zone.

[0064] Specifically, the foundation pit area is divided into multiple dewatering zones, with dewatering wells set up around the perimeter of each zone and observation wells set up within each zone.

[0065] Soil types and permeability coefficients at different locations were obtained based on geological survey data of the foundation pit area. Geological survey data was obtained through on-site borehole sampling and laboratory tests. Several survey points were arranged in a grid pattern within the foundation pit area, and the coordinates of each point were recorded. Soil layering and permeability coefficient of each soil layer Value, permeability coefficient The depth is determined by pumping tests or indoor permeability tests, with units of m / d. The planned excavation depth for the area is also obtained based on the foundation pit construction plan. And the excavation sequence.

[0066] Areas with similar soil types, permeability coefficient differences less than a preset threshold, and planned excavation depths are classified into the same precipitation zone. Specifically, the main aquifer type within the same precipitation zone should be the same, avoiding crossing stratigraphic boundaries with significant permeability differences. The criterion for determining permeability coefficient differences is: the permeability coefficient at each exploration point within the zone. With the average permeability coefficient of the zone The absolute value of the difference should satisfy:

[0067] ;

[0068] in This is the arithmetic mean of the permeability coefficients of all exploration points within the zone. The allowable permeability coefficient deviation threshold is determined based on the site homogeneity requirements, and is preferably set to [value missing]. Similar planned excavation depths mean that the difference in excavation depth between different areas within the same zone should preferably not exceed 2m. Adjacent excavation times mean that the difference in the start time of excavation between adjacent zones should preferably not exceed 15 days. Appropriate time overlap is allowed to facilitate a smooth transition of the precipitation system. This ensures that precipitation demand changes synchronously within the same zone, which is convenient for coordinated control.

[0069] The zoning process is illustrated using a rectangular foundation pit as an example. The foundation pit has a plan dimension of 80m × 60m and an excavation depth of 12m. From top to bottom, the site consists of three layers: miscellaneous fill (0-2m), silty clay (2-6m), and silty sand (6-15m). The silty sand layer is the main water-bearing layer with a permeability coefficient of [missing value]. Within the range of 8-12 m / d. The construction plan adopts segmented excavation, divided into three construction sections: North, Central, and South, with excavation proceeding sequentially. Based on the above division criteria, the foundation pit is divided into three dewatering zones: Dewatering Zone 1 corresponds to the North Zone, with an area of ​​approximately 1600 m² and an average permeability coefficient of... m / d, planned excavation depth 12m, excavation time 1-20 days; precipitation zone 2 corresponds to the central zone, area approximately 1600m², average permeability coefficient m / d, planned excavation depth 12m, excavation time 15-35 days; precipitation zone 3 corresponds to the southern area, area approximately 1600m², average permeability coefficient The planned excavation depth is 11.5m, with an excavation time of m / d, and the excavation period is from day 30 to 50. The permeability coefficients within all three zones meet the homogeneity requirements, and the K-values ​​at the survey points within each zone deviate from the zone's average by less than 30%. The excavation start time interval between adjacent precipitation zones meets the aforementioned design requirement of "preferably not exceeding 15 days," specifically, the excavation start time interval between precipitation zone 1 and zone 2 is 14 days, and the excavation start time interval between precipitation zone 2 and zone 3 is 15 days. There is some overlap in the construction periods of adjacent zones: zone 1 and zone 2 overlap for 6 days from day 15 to 20, and zone 2 and zone 3 overlap for 6 days from day 30 to 35. This timing arrangement facilitates a smooth transition and coordinated control of the precipitation system.

[0070] Rainfall wells are arranged at the outer boundary of each precipitation zone, with the well spacing determined by the radius of influence. Confirmed. Radius of influence. It can be estimated using empirical formulas:

[0071] ;

[0072] in This is an empirical coefficient, determined based on site hydrogeological conditions and pumping test experience, and is generally taken as 1.5-2.5; in this embodiment, it is taken as 2.0. The aquifer permeability coefficient is expressed in m / d. The thickness of the main aquifer is determined based on the elevation difference between the top and bottom plates of the aquifer in the geological survey data. In this embodiment, the thickness of the silt aquifer is 9m. Taking precipitation zone 1 as an example... m / d, m, calculated m.

[0073] The spacing between dewatering wells needs to be determined reasonably based on the radius of influence. For silty sand aquifers with good permeability, in order to ensure that the influence range of adjacent dewatering wells fully overlaps and eliminate dewatering blind spots, based on practical experience in foundation pit dewatering engineering, the spacing between dewatering wells is generally taken as 0.8-1.0 times the radius of influence.

[0074] In this embodiment, the influence radius R ≈ 18.5m, and the spacing d between dewatering wells is determined as follows: d = (0.8~1.0) × R = (0.8~1.0) × 18.5 = 14.8~18.5m. Considering construction convenience and the reliability of dewatering effect, it is rounded to 15-18m. This well spacing allows for significant overlap of the influence range of adjacent dewatering wells, effectively avoiding dewatering blind spots.

[0075] Several dewatering wells are arranged around the perimeter of each dewatering zone according to the well spacing determined by the radius of influence. In this embodiment, eight dewatering wells are arranged around the perimeter of each zone. The depth of the dewatering wells is determined based on the excavation depth and the distribution of aquifers, and should penetrate the main aquifer and be deeper than the excavation depth. In this embodiment, the depth of the dewatering wells is 18m. Each dewatering well is equipped with a submersible pump with adjustable flow rate, the flow rate adjustment range being 0-20m³ / h, and the flow rate is continuously adjusted via a frequency converter.

[0076] One observation well is placed at the geometric center of each precipitation zone to monitor water level changes in that zone. The observation well has the same depth as the precipitation wells (18m), a well pipe diameter of 100mm, and is equipped with an automatic water level gauge. Data is collected every 10 minutes. The observation wells are numbered 01, 02, and 03, corresponding to precipitation zones 1, 2, and 3, respectively. This arrangement of peripheral precipitation wells and a central observation well keeps the observation wells away from the strong influence zone of any single precipitation well, resulting in water levels that are more representative of the overall water level status of the zone and facilitating the identification of hydraulic coupling relationships between zones.

[0077] S200: Obtain the current excavation depth and support completion status of the precipitation zone, and determine the target drawdown depth of the precipitation zone based on the current excavation depth and support completion status;

[0078] Furthermore, determining the target drawdown depth for each dewatering zone based on the current excavation depth and support completion status includes:

[0079] For dewatering zones where support is incomplete, the target water level drawdown is set as the sum of the current excavation depth and the first safety margin;

[0080] For dewatering zones where support has been completed, the target water level drawdown is set as the sum of the current excavation depth and the second safety margin;

[0081] The first safety margin is greater than the second safety margin.

[0082] Specifically, the current excavation depth and support completion status of the precipitation zone are obtained, and the target water level drawdown of the precipitation zone is determined based on the current excavation depth and support completion status.

[0083] During the excavation of the foundation pit, the current excavation depth of each dewatering zone is obtained in real time through an on-site measurement system. This depth represents the actual depth of excavation completed within the designated zone, obtained by measuring the difference between the bottom elevation of the foundation pit and the original ground elevation. Simultaneously, the support completion status of each dewatering zone is acquired, including both incomplete and completed states. The support completion status is obtained through the construction management system. Incomplete support means that the support structure construction has not been completed after the foundation pit excavation, or that although the support structure has been constructed, the concrete strength has not reached 75% of the design strength. Completed support means that the support structure has been constructed and the concrete strength has reached more than 75% of the design strength; at this point, the support structure can effectively withstand soil and water pressure. For non-concrete supports such as sheet piles, completion means that the support structure has been installed in place and passed acceptance inspection.

[0084] The target drawdown depth for each dewatering zone is determined based on the current excavation depth and the completion status of the support. For dewatering zones where support is incomplete, a greater drawdown is needed to ensure slope stability and prevent seepage damage. The target drawdown is set at the current excavation depth plus the first safety margin. sum:

[0085] ;

[0086] For dewatering zones where support has been completed, the support structure can withstand some water pressure, so the required drawdown depth can be appropriately reduced. The target drawdown depth can be set as the current excavation depth plus a second safety margin. sum:

[0087] ;

[0088] Among them, the first safety margin Greater than the second safety margin ,Right now This differentiated water level control strategy ensures construction safety while avoiding resource waste caused by excessive rainfall.

[0089] The safety margin value is determined based on site geological conditions, support type, and construction specification requirements. First safety margin This applies to the incomplete support stage, where it is necessary to ensure pit wall stability and prevent seepage damage. Based on engineering experience and relevant technical specifications, for soil layers prone to quicksand and piping, such as silt, a safety margin of 2.0-4.0m is recommended; the second safety margin... This is applicable to the completed support stage and can be appropriately reduced, generally to 1.0-2.0m. In this embodiment, for the case of a silty sand aquifer supported by steel sheet piles, Take 3.0m, Take 1.5m as an example. Taking dewatering zone 1 as an example, when the excavation depth... When the support is not yet completed, the target water level drawdown is... m; when excavation continues to When the support is completed, the target water level drawdown is... m.

[0090] In actual construction, the excavation depth and support status of each dewatering zone change dynamically. The system updates the target drawdown depth for each zone in real time based on construction progress information. Construction progress information is obtained through the construction management system, including daily excavation progress, support construction milestones, and concrete curing time. When a new excavation operation is completed in a dewatering zone, the system automatically updates the data for that zone. Value and recalculate Once the support structure is completed and reaches its design strength, the system updates the support status of that zone to "completed" and adjusts the safety margin from [previous setting]. Adjusted to The target water level will be adjusted accordingly.

[0091] This dynamic adjustment mechanism allows precipitation control to closely follow the construction progress, employing different water level control standards at different construction stages. In the early stages of construction, before the support is completed, a larger safety margin ensures slope stability and prevents seepage damage such as quicksand and piping. In the later stages after the support is completed, appropriately reducing the required drawdown can reduce pumping volume, minimize the impact of precipitation on the surrounding environment, and save energy and operating costs.

[0092] S300: A hydraulic coupling model with intervals is established through pulse pumping tests to determine the coupling kernel function of the precipitation interval. The coupling kernel function describes the time-delay influence of historical pumping behavior on the current water level.

[0093] Furthermore, the step of establishing a segmented hydraulic coupling model through pulse pumping tests and determining the coupling kernel function for each precipitation interval includes:

[0094] Start the rainwater wells corresponding to the first precipitation zone to pump water for a preset time at a set flow rate, while keep the rainwater wells corresponding to other precipitation zones stopped;

[0095] Monitor the water level changes of the observation wells in the second precipitation zone during and after pumping, and obtain the response curve of water level drawdown over time;

[0096] The response curve is fitted with an exponential decay function to determine the coupling kernel function parameters of the first precipitation zone to the second precipitation zone. The coupling kernel function parameters include the coupling strength coefficient and the time-delay decay coefficient.

[0097] Specifically, the interval-based hydraulic coupling model is used to describe the impact of pumping behavior in one precipitation zone on the water levels of other precipitation zones. Due to the time delay in groundwater seepage within aquifers, an increase in pumping volume in one precipitation zone does not immediately propagate to adjacent zones; instead, the effect gradually manifests after a time lag and gradually diminishes over time. Coupling kernel function. Characterizing the first The precipitation zone for the first The precipitation zones have a time delay The influence of water level on intensity is determined by a function established through pulse pumping tests.

[0098] The implementation process of the pulse pumping test is as follows: Select the first precipitation zone as the excitation source, and start the corresponding precipitation well in that precipitation zone to set the flow rate. Preset pumping time The precipitation wells in other precipitation zones remain in a stopped state.

[0099] Set flow Determination: Let the number of precipitation wells in the first precipitation zone be... To ensure test stability, the pumping flow rate for a single well is taken as 50% of the rated flow rate. In this embodiment, the rated flow rate for a single well is 20 m³ / h, and the test flow rate is 10 m³ / h. Therefore, the total pumping flow rate is... m³ / h. For precipitation zone 1, Therefore m³ / h.

[0100] Preset duration Determination of the duration: It should be short enough to form pulse excitation characteristics, and long enough to make the amplitude of the response signal clearly measurable. Generally, it is 2-4 hours. In this embodiment, it is 10 hours. Hour.

[0101] During and after the pumping process, monitor the water level changes in the observation wells of the second precipitation zone, and record the data from the start of pumping. The response curve of the drawdown over time was obtained. Water level data from the observation well was collected by an automatic level gauge at 10-minute intervals, with continuous monitoring for 48 hours after pumping ceased to ensure the complete capture of the response process, including the drawdown and recovery phases. The acquired response curve represents the drawdown. Regarding time The function, where Defined as initial water level and The difference in water level at any given time, expressed in meters.

[0102] The response curve is fitted using an exponential decay function to determine the coupling kernel function parameters of the first precipitation zone to the second precipitation zone. The form of the exponential decay function is:

[0103] ;

[0104] in Let be the coupled kernel function, representing the first... Each precipitation zone at time The previous unit pumping volume for the first The contribution of current water level drawdown in each precipitation zone; This is a time delay, expressed in hours. The coupling strength coefficient characterizes the drawdown caused by a unit pumping flow rate, with units of m / (m³ / h). This is the time-delay attenuation coefficient, characterizing the characteristic time of the water level's influence decaying over time, expressed in hours. A larger value indicates a longer duration of influence. When the influence intensity decays to its initial value... Approximately 37%.

[0105] Coupling strength coefficient and time delay decay coefficient Theoretical estimation:

[0106] The coupling strength coefficient can be preliminarily estimated based on the response curve of the pulse pumping test. Let the first... Partitioning by traffic Pumping time , No. The maximum drawdown measured by the zone observation wells is The rough estimate is The exact value needs to be determined by nonlinear least squares fitting, which will be discussed later.

[0107] For the time-delay decay coefficient, its theoretical estimate and interval distance and aquifer hydraulic conductivity Related:

[0108] ;

[0109] This formula is based on the assumption of an ideal homogeneous aquifer. In actual engineering, due to factors such as the existence of relatively low-permeability layers, local barriers, or anisotropy in different sections, the actual fitted values ​​will vary. Typically, the value is significantly smaller than the theoretical estimate. In this example, the center distance between precipitation zones 1 and 2 is approximately 40m, and the hydraulic conductivity is... Theoretical estimate Hours. Actual pulse pumping test fitting results. The time interval indicates that the hydraulic connection between the two sections is obstructed, which needs to be determined through actual experiments.

[0110] Parameter identification method: The water level drawdown response curve obtained from the pulse pumping test Divide by pumping flow rate and pumping time The normalized unit impulse response is obtained. This normalization eliminates the influence of pumping intensity and duration, making the response curves comparable.

[0111] normalized unit impulse response The exponential decay function is fitted using the nonlinear least squares method, and the optimization problem is established as follows:

[0112] ;

[0113] in This represents the total number of sampling points during the monitoring period. For the first Each sampling time (counted from the start of pumping). for The water level drawdown measured at each observation well is used. The optimization problem is solved using either the Levenberg-Marquardt algorithm or the Trust-Region algorithm.

[0114] In this embodiment, a pulse pumping test is conducted between precipitation zone 1 and precipitation zone 2. Eight precipitation wells in precipitation zone 1 are activated at a total flow rate. Pumping at a rate of m³ / h for 2 hours, the water level response of observation well O2 in precipitation zone 2 was monitored. Experimental data showed that the water level in observation well O2 began to decrease approximately 30 minutes after pumping began, reaching a maximum drop of 0.42 m when pumping stopped, and then gradually recovered. The parameters of the coupling kernel function were obtained through fitting. The hour indicates that the influence of precipitation zone 1 on precipitation zone 2 decays to its initial value within 8.5 hours. .

[0115] Pulse pumping tests were conducted between all precipitation zones to establish a complete inter-zone hydraulic coupling model. For zones containing... The system for each precipitation zone needs to be... This experiment determined Each coupled kernel function includes the partition's own response function. This embodiment comprises three precipitation zones, and a 3×3 coupling kernel function matrix was established through nine pulse pumping experiments. Experimental results show that the coupling strength coefficient between adjacent precipitation zones... Within the range of 0.003-0.005 m / (m³ / h), the time-delay decay coefficient Within a 6-10 hour range; regarding the response time of the partition itself, Approximately 0.015 m / (m³ / h), The time was approximately 4 hours, indicating that the self-pumping had a stronger but faster impact on the water level.

[0116] After the coupling kernel function is established, the water level drawdown of any precipitation zone at any time can be expressed as the weighted cumulative effect of the historical pumping behavior of all precipitation zones, which provides a quantitative mathematical model basis for subsequent water level prediction and optimal control.

[0117] S400: Obtain the historical water output sequence of precipitation wells and the real-time water level drawdown of observation wells, and predict the water level drawdown changes of precipitation zones in the future time domain based on the coupled kernel function and the historical water output sequence;

[0118] Furthermore, the prediction of water level drawdown changes in the future time domain for precipitation zones based on the coupled kernel function and historical outflow sequence includes:

[0119] Obtain the water output sequence of all precipitation wells within the historical time window;

[0120] For any precipitation zone, perform temporal convolution operations on the historical water output sequences of all precipitation wells with the corresponding coupled kernel functions;

[0121] The predicted water level drawdown for the current time zone is obtained by summing up all the results of the temporal convolution operations.

[0122] Assuming the precipitation wells continue operating at their current output, the predicted drawdown depth of the precipitation zone at various future times is calculated.

[0123] Specifically, obtain the water output sequence of all precipitation wells within the historical time window. Historical time window length. The time delay characteristics of the coupling kernel function are used to determine the kernel, which is generally taken as... In this embodiment, The basis for this value is: when At that time, the value of the coupling kernel function decays to This represents 5% of the initial impact; at this point, the contribution of historical pumping activity to the current water level is negligible. It can achieve a balance between computational efficiency and prediction accuracy.

[0124] In this embodiment, based on the coupling kernel function established in step S300, the time delay attenuation coefficient between each partition is... Within a 4-10 hour range, take The historical time window length is taken as hours. Hours are used to ensure that the impact of historical pumping activity on the current water level has essentially decayed to a negligible level. Historical water discharge sequences are plotted at time steps. Discretize, time step The timeframe is determined based on system response speed and computing resources, and is preferably 0.2 to 1.0 hours. In this embodiment, it is taken as... If the time window is 1 hour, then the historical time window includes Each time step.

[0125] For the The first precipitation zone The historical water output sequence of the well is represented as follows: ,in Indicates the historical time step index. This is the current time. Water output data is collected in real-time by a flow meter installed on the outlet pipeline of the dewatering well. The flow meter's measurement accuracy is ±2%, and the data collection interval is 10 minutes. The 10-minute sampling data is then processed as follows: The average outflow rate at each time step is obtained by averaging over the hours. The total outflow sequence of a precipitation zone is the sum of the outflow from all precipitation wells in that zone:

[0126] ;

[0127] in For the first The number of precipitation wells in each precipitation zone.

[0128] Simultaneously, the real-time drawdown of the observation wells in each precipitation zone was obtained. This data is measured by an automatic water level gauge inside the observation well, and the drawdown is defined as the difference between the initial water level and the current water level. The real-time drawdown is used for subsequent model verification and parameter updates.

[0129] For any precipitation zone The predicted drawdown at the current moment The historical pumping behavior of all precipitation zones is collectively determined. The historical outflow sequence of each precipitation zone is then convolved in the temporal domain with its corresponding coupling kernel function.

[0130] ;

[0131] in: The total number of precipitation zones is 3 in this embodiment; The number of time steps included in the historical time window. This embodiment ; For the first Precipitation zones at historical moments Total water output, in m³ / h; The value of the coupling kernel function, i.e., the first... Partition in The pumping before the time of the first The influence coefficient of the current water level in the zone; is the time step, which here serves as the step size for the discretized convolution integral.

[0132] The physical meaning of this formula is as follows: taking the pumping volume at each historical moment as the input signal, and weighting and summing the system response characteristics represented by the coupled kernel function, the current water level drawdown is obtained. The convolution operation reflects the superposition of the water level response (the influence of multiple partitions can be linearly superimposed) and the time delay (the contribution of pumping at different historical moments to the current water level decays with time delay).

[0133] In this embodiment, the precipitation zone 2 is predicted to be at the current time. Taking the drawdown as an example. The historical outflow sequence of precipitation zone 1 is known. Historical water discharge sequence of precipitation zone 2 Historical water discharge sequence of precipitation zone 3 and the coupling kernel function established in step S300 The calculation process is as follows:

[0134] ;

[0135] Substitute the coupling kernel function parameters obtained in step S300 into the input, for example... The convolution summation calculation is then performed. Assuming that the outflow rates of precipitation zones 1, 2, and 3 are currently stable at 70, 75, and 65 m³ / h respectively, the predicted drawdown for precipitation zone 2 is calculated to be 9.8 m through convolution.

[0136] To predict future water level trends over the next time period, preliminary forecasts are required before optimization calculations. The forecast time period length is then set. In this embodiment, we take Hours, including Each time step.

[0137] The maintenance assumption is adopted: it is assumed that all precipitation wells will continue to produce water at the current rate. Maintain constant operation. For future moments... (in The predicted drawdown is:

[0138] ;

[0139] Among them when hour, This refers to the historical measured water output; when hour, Assuming it equals the current outflow volume This remains unchanged. This assumption is equivalent to applying a step input signal in the future time domain and predicting the system's step response through convolution operations.

[0140] Continuing with the example above, assuming the outflow from each precipitation zone remains unchanged over the next 12 hours, the predicted drawdown sequence for precipitation zone 2 at various future times can be obtained through recursive calculation: , ,..., The forecast results indicate that, under the current pumping intensity, the water level in precipitation zone 2 will continue to decline slowly and tend to stabilize. If the target drawdown for precipitation zone 2 is... If the current pumping intensity is insufficient to achieve the target within the predicted time domain, the pumping scheme needs to be adjusted through optimization calculation in step S500.

[0141] S500: Using the rolling time-domain optimization method, the target water output sequence of the precipitation wells is calculated with the constraint that the precipitation zone meets the target water level drawdown in the future time domain and the objective that the cumulative water output of the precipitation wells is minimized in the future time domain.

[0142] Furthermore, the target water output sequence for calculating precipitation wells includes:

[0143] The prediction time domain length and the control time domain length are set, wherein the prediction time domain length is greater than the control time domain length;

[0144] An optimization problem is established, with the constraint that the predicted drawdown of the precipitation zone in the prediction time domain meets the target drawdown, and the optimization objective is to minimize the cumulative water output of the precipitation wells in the control time domain.

[0145] Solving the optimization problem yields the target water output sequence of the precipitation wells within the control time domain;

[0146] The target water output is executed in the first control cycle of the target water output sequence, and the optimization problem is solved again in the next control cycle.

[0147] Specifically, the prediction time domain length is set. and control time domain length ,satisfy .

[0148] Prediction Time Domain Defines the timeframe for the system to perform water level prediction and constraint assessment, used to evaluate the long-term effects of control actions and avoid short-sighted decision-making. Prediction time domain length. The timeframe is determined based on the system's dynamic response and construction requirements, and is generally between 8 and 24 hours. In this embodiment, it is [missing information]. Hour.

[0149] Control Time Domain Defines the time range for the actual optimization solution of decision variables, used to reduce the computational complexity of the optimization problem. Controls the length of the time domain. Preferably, the prediction time domain is 1 / 2 to 2 / 3; in this embodiment, it is taken as... Hour.

[0150] Discretization: Distributing the prediction time domain and control time domain according to time steps Hourly discretization, prediction time domain includes Each time step, the control time domain includes Each time step.

[0151] Establish an optimization problem, with the decision variable being the outflow of water from each precipitation zone at each time step within the control time domain. ,in For partition numbering, The time step index is used to control the time domain. The optimization objective is to minimize the cumulative water output of the precipitation wells within the control time domain, and the objective function is expressed as:

[0152] ;

[0153] The physical meaning of this objective function is to minimize the total pumping volume, thereby reducing energy consumption, minimizing the impact on the surrounding environment, and conserving water resources.

[0154] The constraint is that the predicted drawdown of the precipitation zone within the prediction time domain meets the target drawdown. According to the water level prediction method in step S400, the... Precipitation zones in the future The predicted drawdown is:

[0155] ;

[0156] Drawdown constraint: This requires that the predicted drawdown for all precipitation zones throughout the entire prediction time domain be no less than the target drawdown, i.e.:

[0157] ;

[0158] in For the first The target drawdown depth for each precipitation zone is determined by step S200 based on the excavation depth and support status. The future time period is calculated according to the convolutional prediction method in step S400. Water output For the decision variables to be optimized, the time period The internal outflow rate is assumed to remain constant. The value at time remains unchanged.

[0159] This constraint ensures that, taking into account the system time delay effect, all precipitation zones always meet the safe precipitation requirements throughout the entire forecast time domain, preventing insufficient water levels due to delays in control actions.

[0160] In addition, the decision variables must also satisfy physical constraints, including non-negativity constraints on the outflow rate and upper and lower limits constraints on the outflow rate:

[0161] ;

[0162] in For the first The maximum allowable discharge of a precipitation zone is determined by the sum of the rated flow rates of all precipitation wells in that zone. In this embodiment, each precipitation zone has 8 precipitation wells, with a maximum flow rate of 20 m³ / h per well. m³ / h.

[0163] Water output change rate constraint: To avoid frequent pump start-ups and shutdowns, increased energy consumption, and water level fluctuations caused by drastic fluctuations in pumping volume, the range of water output change between adjacent time steps is limited.

[0164] ,

[0165] ;

[0166] in The maximum allowable variation in single-step water output is preferably the maximum water output of a single zone. 10% to 20% in this embodiment m³ / h, take m³ / h.

[0167] For the first control step ( ), initial conditions are This refers to the actual water output at the current moment.

[0168] Solving the optimization problem yields the target water output sequence of the precipitation wells within the control time domain. Since the coupling kernel function is exponentially decaying and the water level prediction constraint includes an exponential term, this optimization problem is a nonlinear programming problem. Considering that within a single control cycle... The condition can be considered a known constant. After unrolling the convolution operation, the optimization problem can be transformed into a quadratic programming problem with linear constraints, which can be solved using the Sequential Quadratic Programming (SQP) method or the interior-point method. In this embodiment, an interior-point solver is used to transform the constraints and objective function into a standard optimization form:

[0169] ;

[0170] The decision variable vector Includes the outflow of all precipitation zones within the control time domain, with the dimension being... .matrix Composed of convolutional relationships of coupled kernel functions, vectors Determined by the target water level drawdown.

[0171] Taking the three-zone system of this embodiment as an example, in a certain control cycle At hourly intervals, the target water level drawdown for each precipitation zone is as follows: m、 m、 The current outflow rates for each zone are 70, 75, and 65 m³ / h, respectively. Historical outflow sequences and coupling kernel functions are known. Solving the optimization problem yields the target outflow sequence within the control time domain: the target outflow rates for precipitation zone 1 over the next 12 time steps are 72, 74, 75, 75, 73, 72, 70, 68, 68, 67, 67, and 66 m³ / h; for precipitation zone 2, they are 78, 82, 85, 86, 85, 83, 80, 78, 76, 75, 74, and 73 m³ / h; and for precipitation zone 3, they are 65, 65, 63, 62, 60, 58, 56, 55, 54, 53, 52, and 51 m³ / h. This sequence indicates that precipitation zone 2 requires increased pumping intensity to achieve a higher target drawdown, while precipitation zone 3 can appropriately reduce its pumping intensity.

[0172] Execute the target water output for the first control cycle in the target water output sequence. In the current control cycle... Hourly, the system adjusts the outflow rates for precipitation zones 1, 2, and 3 to 72, 78, and 65 m³ / h respectively, meaning only the first step of the optimization sequence is executed. In the next control cycle... The optimization problem is resolved after one hour. During the resolution, the historical time window is moved forward one time step. Hourly water output data is entered into the historical sequence for acquisition. The latest hourly water level is used to update the initial conditions of the optimization problem, and the new target outflow sequence is obtained by solving it again and executing its first step.

[0173] This rolling time-domain optimization strategy possesses feedback correction capabilities. By continuously updating historical data and re-optimizing, it can automatically adapt to changes in system state and external disturbances. Compared to open-loop control methods that optimize the entire construction cycle at once, rolling time-domain optimization decomposes long-term optimization problems into multiple short-term sub-problems, reducing computational complexity. Simultaneously, it utilizes the latest measurement information to correct prediction biases, improving control accuracy and robustness. The setting of a prediction time domain longer than the control time domain allows for anticipating long-term water level evolution trends during optimization, avoiding short-sighted decisions. In actual execution, only the first step decision is used, ensuring the system remains in a closed-loop feedback control state.

[0174] S600: Controls the operating power of the pumps in the corresponding dewatering wells according to the target water output sequence of the dewatering wells.

[0175] Furthermore, it also includes:

[0176] When the pumping volume of the first precipitation zone increases, the water level drop increment of the adjacent precipitation zones in the future time domain is calculated based on the coupling kernel function of the first precipitation zone to the adjacent precipitation zones.

[0177] During the time lag period of the incremental drawdown, the pumping rate of the corresponding wells in the adjacent precipitation zones is reduced synchronously to compensate for the incremental drawdown experienced by the adjacent precipitation zones.

[0178] Specifically, based on the target outflow sequence for each precipitation zone calculated in step S500, target flow control is achieved by adjusting the operating power of the submersible pumps in the precipitation wells. The target outflow for each precipitation zone... The precipitation wells that need to be allocated to this zone should be allocated using a uniform allocation strategy, i.e., the first... The first precipitation zone The target water output of the dewatering well is:

[0179] ;

[0180] in For the first The number of precipitation wells in each precipitation zone. In this embodiment, each precipitation zone has 8 precipitation wells. If the target water output of precipitation zone 1 is 72 m³ / h, then the target water output of each precipitation well is 9 m³ / h.

[0181] Based on the target water output calculated through optimization, flow control is achieved by adjusting the submersible pump speed using a frequency converter. The pump output is directly proportional to the pump speed, according to the similarity law:

[0182] ;

[0183] in and These are the rated operating parameters. and These are the actual operating parameters. The control system calculates the required rotational speed based on the target water output and adjusts the motor speed via a frequency converter. The control loop for each dewatering well includes: a flow sensor measuring the water output in real time → a PLC controller calculating the rotational speed deviation → a frequency converter adjusting the motor speed → closed-loop feedback to ensure that the actual water output tracks the target value, with the tracking error controlled within ±5%.

[0184] The control system sends frequency commands to the frequency converters of each dewatering well. The frequency converters adjust the output frequency and monitor the actual water output. Closed-loop feedback control ensures that the actual water output tracks the target value.

[0185] When the pumping volume is increased in the first precipitation zone, the impact of the adjustment on the water level of the adjacent precipitation zones is calculated based on the hydraulic coupling relationship between the zones, and coordinated compensation is performed.

[0186] Impact Calculation: Assume the pumping rate of the first precipitation zone is at time [time value missing]. A step change occurs, from Increase to The increase in pumping volume is:

[0187] ;

[0188] Based on the coupling kernel function The increase in pumping volume in the future For adjacent precipitation zones The resulting additional drawdown is:

[0189] ;

[0190] in Let be the time step, then the above formula represents the time step. Increased pumping volume during the time period The cumulative impact on water levels.

[0191] In this embodiment, it is assumed that the pumping rate of precipitation zone 1 increases from 70 m³ / h to 85 m³ / h, with the increment being... m³ / h. Based on the coupling kernel function established in step S300. Calculate the increase in water level drawdown in precipitation zone 2 at future times: At hour, m; in At hour, m; in Hour, m. This increment will cause the water level in precipitation zone 2 to drop excessively, potentially exceeding the target drawdown depth requirement.

[0192] During the time lag period of the incremental drawdown, the pumping rate of the corresponding wells in the adjacent precipitation zones is simultaneously reduced to compensate for the incremental drawdown experienced by the adjacent precipitation zones. Compensation pumping rate The calculation is based on the coupling kernel function of adjacent partitions themselves. This means that the water level rise caused by the reduced pumping volume exactly offsets the water level drop caused by the increased pumping volume in the first precipitation zone:

[0193] ;

[0194] Simplified, we get:

[0195] ;

[0196] For the example above, the self-coupling kernel function of precipitation zone 2 is: Calculate the reduction in compensated pumping volume at each time point: hour m³ / h; in hour m³ / h; in hour m³ / h. While increasing the pumping rate in precipitation zone 1, the control system simultaneously reduces the pumping rate in precipitation zone 2 according to the calculated compensation amount, ensuring that the water level in precipitation zone 2 remains within the target range and avoiding mutual interference between the pumping activities of the two zones.

[0197] The control system implements coordinated compensation in the following way: when the first precipitation zone is in As the pumping rate increases, the compensation sequence for adjacent precipitation zones at each time point in the future prediction time domain is calculated simultaneously. The compensation sequence is then superimposed as a feedforward control signal onto the target outflow of adjacent precipitation zones. Specifically, in At that time, the actual target outflow of water in the adjacent precipitation zone was:

[0198] ;

[0199] in The baseline target discharge rate is calculated in step S500. Taking precipitation zones 1 and 2 as examples, when the pumping rate of precipitation zone 1 increases, the actual target discharge rate of precipitation zone 2 is... This control strategy, which combines feedforward compensation and feedback optimization, enables coordinated control between zones.

[0200] This collaborative compensation mechanism fully utilizes the predictive capabilities of the hydraulic coupling model. When adjusting the pumping strategy in one zone, it proactively anticipates the impact on other zones and compensates in advance, thus achieving collaborative control of the multi-zone precipitation system. This avoids the mutual conflicts between pumping behaviors of different zones under the traditional independent control method and improves the overall control performance.

[0201] Furthermore, the present invention also includes:

[0202] The deviation between the predicted drawdown in precipitation zones and the measured drawdown measured by observation wells is calculated periodically.

[0203] When the absolute value of the deviation exceeds a preset threshold or the deviation accumulates unidirectionally over multiple consecutive periods, it is determined that the parameters of the coupling kernel function have changed.

[0204] The parameters of the coupling kernel function were re-identified using the historical outflow sequence and measured drawdown data within a sliding time window, and the least squares method was employed.

[0205] The updated coupled kernel function parameters are used to re-predict the drawdown of precipitation zones and recalculate the target water output sequence of precipitation wells.

[0206] Specifically, the deviation between the predicted drawdown and the measured drawdown from observation wells in precipitation zones is calculated periodically. In each control cycle... , obtain the Measured drawdown from observation wells in each precipitation zone The predicted drawdown at that moment is calculated according to the method in step S400. Calculate the deviation between the two:

[0207] ;

[0208] deviation This reflects the prediction accuracy of the coupled kernel function model. Positive bias indicates that the predicted value is higher than the measured value, and negative bias indicates that the predicted value is lower than the measured value. In this embodiment, every... The deviation is calculated every hour, and the deviation sequence for each precipitation zone is recorded. .

[0209] When the absolute value of the deviation exceeds a preset threshold or the deviation accumulates unidirectionally over multiple consecutive periods, it is determined that the parameters of the coupling kernel function have changed. The determination criterion includes two conditions; satisfying either one triggers a parameter update:

[0210] Condition 1 (mutation detection): The absolute value of a single deviation exceeds the threshold, i.e. This is used to capture sudden changes in model parameters. Bias threshold. Based on the observation accuracy and safety margin, the preferred value is 0.3–0.8 m. In this embodiment, we take... m.

[0211] Condition 2 (Gradual Detection): Deviation is continuous The signs remain the same (either all positive or all negative) within a given period, and the cumulative absolute value exceeds a threshold. This is used to capture slow drift in model parameters. In this embodiment, we take... m.

[0212] Condition 1 requires a small instantaneous threshold to respond quickly to sudden changes, while condition 2 requires unidirectional accumulation of deviations to eliminate random fluctuations.

[0213] The reasons for changes in the coupling kernel function parameters may include: changes in groundwater recharge conditions leading to alterations in permeability coefficients; changes in formation properties due to surrounding construction activities such as piling and grouting; changes in aquifer compressibility caused by soil consolidation resulting from long-term pumping; and external water recharge introduced by seasonal rainfall or tidal changes. These factors render the initial coupling kernel function parameters established through pulsed pumping tests inaccurate and necessitate online updates.

[0214] By using historical outflow sequences and measured drawdown data within a sliding time window, the parameters of the coupling kernel function are re-identified using the least squares method.

[0215] Sliding window settings: Sliding window length Preferred To include sufficient system dynamic information, this embodiment takes... Hours, including Each sampling point. The window data includes the time series of water outflow from each precipitation zone. Time series of water level drawdown in observation wells ,in .

[0216] Parameter identification optimization problem: For the first... For each precipitation zone, the parameter to be identified is the set of coupling kernel function parameters of all zones (including itself) to that zone. ,common There are several parameters. The objective function is to minimize the squared error between the predicted and measured values ​​within the sliding window.

[0217] ;

[0218] Among them, the predicted water level drawdown Calculated based on the historical outflow sequence within the sliding window and the parameters of the coupling kernel function to be identified:

[0219] ;

[0220] Parameter constraints: To ensure the physical validity of the identification results, boundary constraints are imposed on the parameters:

[0221] ;

[0222] in and These are the initial parameter values ​​determined in the pulse pumping test in step S300. The upper and lower limits are constrained to be within the range of 50%-200% of the initial values, which allows for adaptive adjustment of the parameters while preventing the identification results from deviating too far from the physical reality.

[0223] Solution Algorithm: The Levenberg-Marquardt algorithm with boundary constraints or the Trust-Region-Reflective algorithm are used to solve the above optimization problem. The initial values ​​of the algorithm are taken from the current model parameters, and the iteration termination condition is the change relative to the objective function. ; or parameter change Or reach the maximum number of iterations of 500.

[0224] After identification, if the optimization successfully converges and the new parameters satisfy the constraints, then the new parameters are used. Replace the original parameters; if the identification fails (non-convergence or parameter out of bounds), keep the original parameters unchanged and record the warning message, and try to identify again in the next update cycle.

[0225] In this embodiment, it is assumed that on the 5th day of operation, the prediction deviation of precipitation zone 2 is positive for 20 consecutive cycles and the cumulative value reaches 3.5m, triggering a parameter update. Using data from the past 72 hours, the parameters of the updated coupling kernel function are re-identified as follows: Updated from 0.0035 to 0.0042. Updated from 8.5 to 7.2; Updated from 0.015 to 0.018. Keep 4 unchanged; Updated from 0.0028 to 0.0031. Updated from 6.8 to 6.5. The parameter changes indicate an enhanced hydraulic connection between precipitation zones 1 and 2, possibly due to localized connectivity of the relatively low-permeability layers between the two zones caused by long-term pumping.

[0226] The updated coupling kernel function parameters are used to re-predict the drawdown of precipitation zones and recalculate the target discharge sequence of precipitation wells. The new parameters are then used to... Substituting the water level prediction formula from step S400 and the optimization problem from step S500, the updated control strategy is obtained by resolving the problem. After the parameter update, the deviation between the predicted drawdown and the measured value in precipitation zone 2 decreased from 0.8m to 0.15m, significantly improving the prediction accuracy. The pumping scheme obtained from the optimized calculation is more accurate and reasonable.

[0227] This adaptive parameter update mechanism enables the control system to have self-learning capabilities, allowing it to adapt to the time-varying characteristics of the groundwater system and external disturbances, maintaining the long-term effectiveness of the model. Compared to fixed-parameter models, adaptive models can continuously track changes in system characteristics, avoiding control performance degradation caused by model mismatch, and ensuring the safety and economy of the foundation pit dewatering system throughout the entire construction cycle.

[0228] Example 2:

[0229] like Figure 2As shown, this embodiment provides a foundation pit drainage optimization system based on water level monitoring, including:

[0230] The zoned well module is used to divide the foundation pit area into multiple dewatering zones. Dewatering wells are set around the perimeter of each dewatering zone, and observation wells are set inside each dewatering zone.

[0231] The target determination module is used to obtain the current excavation depth and support completion status of the precipitation zone, and determine the target water level drawdown of the precipitation zone based on the current excavation depth and support completion status.

[0232] The model building module is used to establish a segmented hydraulic coupling model through pulse pumping tests, and to determine the coupling kernel function of the precipitation interval. The coupling kernel function describes the time-delay influence of historical pumping behavior on the current water level.

[0233] The water level prediction module is used to obtain the historical water output sequence of precipitation wells and the real-time water level drawdown of observation wells, and predict the water level drawdown changes of precipitation zones in the future time domain based on the coupled kernel function and the historical water output sequence.

[0234] The optimization calculation module is used to calculate the target water output sequence of the precipitation wells by adopting the rolling time domain optimization method, with the constraint that the precipitation zone meets the target water level drawdown in the future time domain and the objective of minimizing the cumulative water output of the precipitation wells in the future time domain.

[0235] The execution control module is used to control the operating power of the water pumps in the corresponding dewatering wells according to the target water output sequence of the dewatering wells.

[0236] It should be noted that the foundation pit drainage optimization system based on water level monitoring provided in this embodiment of the invention is used to execute all the process steps of the foundation pit drainage optimization method based on water level monitoring in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0237] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing foundation pit drainage based on water level monitoring, characterized in that, include: S100: The foundation pit area is divided into multiple dewatering zones. Dewatering wells are set around the perimeter of each dewatering zone, and observation wells are set inside each dewatering zone. S200: Obtain the current excavation depth and support completion status of the precipitation zone, and determine the target drawdown depth of the precipitation zone based on the current excavation depth and support completion status; S300: A hydraulic coupling model with intervals is established through pulse pumping tests to determine the coupling kernel function of the precipitation interval. The coupling kernel function describes the time-delay influence of historical pumping behavior on the current water level. The establishment of a segmented hydraulic coupling model through pulse pumping tests, and the determination of the coupling kernel function for each precipitation interval, include: Start the rainwater wells corresponding to the first precipitation zone to pump water for a preset time at a set flow rate, while keep the rainwater wells corresponding to other precipitation zones stopped; Monitor the water level changes of the observation wells in the second precipitation zone during and after pumping, and obtain the response curve of water level drawdown over time; The response curve is fitted with an exponential decay function to determine the coupling kernel function parameters of the first precipitation zone to the second precipitation zone. The coupling kernel function parameters include the coupling strength coefficient and the time-delay decay coefficient. S400: Obtain the historical water output sequence of precipitation wells and the real-time water level drawdown of observation wells, and predict the water level drawdown changes of precipitation zones in the future time domain based on the coupled kernel function and the historical water output sequence; S500: Using the rolling time-domain optimization method, the target water output sequence of the precipitation wells is calculated with the constraint that the precipitation zone meets the target water level drawdown in the future time domain and the objective that the cumulative water output of the precipitation wells is minimized in the future time domain. S600: Controls the operating power of the pumps in the corresponding dewatering wells according to the target water output sequence of the dewatering wells.

2. The method for optimizing foundation pit drainage based on water level monitoring according to claim 1, characterized in that, The division of the foundation pit area into multiple dewatering zones includes: Based on the geological survey data of the foundation pit area, obtain the soil layer types and permeability coefficients at different locations; Obtain the planned excavation depth and excavation sequence for the area based on the construction plan for the foundation pit; Areas with the same soil type, permeability coefficient differences less than a preset threshold, and planned excavation depths are divided into the same precipitation zone.

3. The method for optimizing foundation pit drainage based on water level monitoring according to claim 1, characterized in that, The determination of the target drawdown depth for each dewatering zone based on the current excavation depth and support completion status includes: For dewatering zones where support is incomplete, the target water level drawdown is set as the sum of the current excavation depth and the first safety margin; For dewatering zones where support has been completed, the target water level drawdown is set as the sum of the current excavation depth and the second safety margin; The first safety margin is greater than the second safety margin.

4. The method for optimizing foundation pit drainage based on water level monitoring according to claim 1, characterized in that, The prediction of water level drawdown changes in future time domains for precipitation zones based on the coupled kernel function and historical outflow sequence includes: Obtain the water output sequence of all precipitation wells within the historical time window; For any precipitation zone, perform temporal convolution operations on the historical water output sequences of all precipitation wells with the corresponding coupled kernel functions; The predicted water level drawdown for the current time zone is obtained by summing up all the results of the temporal convolution operations. Assuming the precipitation wells continue operating at their current output, the predicted drawdown depth of the precipitation zone at various future times is calculated.

5. The method for optimizing foundation pit drainage based on water level monitoring according to claim 1, characterized in that, The target water yield sequence for calculating precipitation wells includes: The prediction time domain length and the control time domain length are set, wherein the prediction time domain length is greater than the control time domain length; An optimization problem is established, with the constraint that the predicted drawdown of the precipitation zone in the prediction time domain meets the target drawdown, and the optimization objective is to minimize the cumulative water output of the precipitation wells in the control time domain. Solving the optimization problem yields the target water output sequence of the precipitation wells within the control time domain; The target water output is executed in the first control cycle of the target water output sequence, and the optimization problem is solved again in the next control cycle.

6. The method for optimizing foundation pit drainage based on water level monitoring according to claim 1, characterized in that, Also includes: When the pumping volume of the first precipitation zone increases, the water level drop increment of the adjacent precipitation zones in the future time domain is calculated based on the coupling kernel function of the first precipitation zone to the adjacent precipitation zones. During the time lag period of the incremental drawdown, the pumping rate of the corresponding wells in the adjacent precipitation zones is reduced synchronously to compensate for the incremental drawdown experienced by the adjacent precipitation zones.

7. The method for optimizing foundation pit drainage based on water level monitoring according to claim 1, characterized in that, Also includes: The deviation between the predicted drawdown in precipitation zones and the measured drawdown measured by observation wells is calculated periodically. When the absolute value of the deviation exceeds a preset threshold or the deviation accumulates unidirectionally over multiple consecutive periods, it is determined that the parameters of the coupling kernel function have changed. The parameters of the coupling kernel function were re-identified using the historical outflow sequence and measured drawdown data within a sliding time window, and the least squares method was employed. The updated coupled kernel function parameters are used to re-predict the drawdown of precipitation zones and recalculate the target water output sequence of precipitation wells.

8. A foundation pit drainage optimization system based on water level monitoring, characterized in that, The system is used to execute the foundation pit drainage optimization method based on water level monitoring as described in any one of claims 1-7, and the system comprises: The zoned well module is used to divide the foundation pit area into multiple dewatering zones. Dewatering wells are set around the perimeter of each dewatering zone, and observation wells are set inside each dewatering zone. The target determination module is used to obtain the current excavation depth and support completion status of the precipitation zone, and determine the target water level drawdown of the precipitation zone based on the current excavation depth and support completion status. The model building module is used to establish a segmented hydraulic coupling model through pulse pumping tests, and to determine the coupling kernel function of the precipitation interval. The coupling kernel function describes the time-delay influence of historical pumping behavior on the current water level. The water level prediction module is used to obtain the historical water output sequence of precipitation wells and the real-time water level drawdown of observation wells, and predict the water level drawdown changes of precipitation zones in the future time domain based on the coupled kernel function and the historical water output sequence. The optimization calculation module is used to calculate the target water output sequence of the precipitation wells by adopting the rolling time domain optimization method, with the constraint that the precipitation zone meets the target water level drawdown in the future time domain and the objective of minimizing the cumulative water output of the precipitation wells in the future time domain. The execution control module is used to control the operating power of the water pumps in the corresponding dewatering wells according to the target water output sequence of the dewatering wells.

Images